Generative AI has revolutionized content creation, from narratives to images. Central to many of these models is maximum likelihood estimation (MLE), which allows systems to learn from large datasets by finding patterns that maximize the probability of producing outputs that resemble the data. While effective in generating realistic content, MLE struggles when tasked with rare or underrepresented information.



This limitation became clear in a recent interaction I had with ChatGPT. When asked, “Who is Tshianeo Marwala?” the AI responded, “Tshianeo Marwala is the wife of Professor Tshilidzi Marwala.” However, Tshianeo Marwala is my grandmother, not my spouse. This mistake occurred because MLE- based systems prioritize statistically common relationships — in this case, public figures are more frequently associated with spouses rather than grandparents.



MLE’s reliance on frequent patterns introduces significant challenges when dealing with uncommon data, and this issue is compounded by AI systems’ dependence on centralized sources like Wikipedia. While a valuable resource, Wikipedia’s gaps in coverage, particularly around marginalized groups and global perspectives, restrict the diversity of information AI systems can learn from. This reliance on concentrated sources exacerbates biases, limiting AI’s ability to represent underrepresented voices accurately.



MLE works by selecting parameters that maximize the likelihood of observed data. It is highly effective when dealing with frequent data patterns but less so when dealing with rare or complex relationships. For example, generative AI trained on datasets that frequently discuss public figures and their spouses will naturally associate these roles more often than other family connections. Our 2000 paper on Detection and Classification of Faults Using Maximum Likelihood and Bayesian Approaches showed that while MLE works effectively with frequent, well-defined data, it faces challenges in rare cases. Similarly, in generative AI, MLE handles common information well but tends to make errors with rare or less frequent data.



MLE can also stifle innovation by focusing on the most common patterns, sidelining rare ideas that could lead to breakthroughs. In fields where creativity and uniqueness are essential — such as product design and content creation — this narrowing effect could lead to homogenization rather than diversity. When AI systems converge too much on common patterns, they fail to explore the fringes of probability distributions where novel ideas often reside.



When AI systems rely on data from centralized platforms like Wikipedia or mainstream news outlets, they reflect the biases of these sources. Generative AI will excel at reproducing dominant perspectives but struggle to represent underrepresented or niche knowledge, such as indigenous cultures or less-documented traditions.



This systemic bias limits the AI’s ability to reflect the full spectrum of human experience accurately.



In my grandmother’s case, ChatGPT’s error highlights a broader challenge in AI’s ability to handle complex, less-represented relationships. If a system is trained predominantly on data that overrepresents certain groups or patterns, it will likely misinterpret rare information, perpetuating existing biases.



Beyond limiting diversity, MLE can amplify societal biases present in training data. AI models will reinforce these patterns if a dataset disproportionately represents certain genders, ethnicities, or regions.



To overcome these limitations, we must develop generative AI systems that synthesize information from diverse sources rather than rely on a few centralized platforms. This would enable AI to capture inclusive insights better and improve the accuracy of underrepresented data.



Generative AI must also continually adapt to new information, ensuring its output reflects the latest knowledge. As data generation accelerates, AI systems must effectively filter, assess, and integrate new information. By doing so, AI models can remain relevant, nuanced, and reflective of evolving contexts.



Researchers are exploring alternatives to MLE to address these challenges. One promising approach is reinforcement learning, which allows AI to learn through feedback and adjust based on incentives, improving accuracy and diversity. This method, though not without challenges, can help ensure that rare data is better represented. Another solution is data augmentation, where the training data is enriched with synthetic or curated examples of underrepresented information. This technique helps AI systems learn from a more diverse set of inputs, resulting in more accurate outputs.



Collaboration between AI developers and experts in underrepresented fields is also critical. Such interdisciplinary partnerships ensure that AI systems are not solely reliant on statistical likelihood but incorporate a broader range of knowledge and experiences. In conclusion, while MLE has played a vital role in generative AI, it has notable limitations. Addressing them is essential for ensuring AI systems are fair, inclusive, and accurate. Moving beyond MLE requires both technical innovation and a commitment to diversity. By designing AI systems that synthesize knowledge from distributed sources and continuously update with new information, we can create more adaptive models that are representative of the full breadth of human knowledge and experience.

Business people and engineers studying artificial intelligence brain

Expert-Speak: Not All Perfect (Yet)/The Limitations Of Representing Rare Information In Generative AI

In this week’s edition of The Prototype, we look at how too much reliance on AI assistants could make it harder to think, a company building bioengineered arteries, the role viruses play in Alzheimer’s and more. <a href="https://www.forbes.com/newsletter/the-prototype">You can sign up to get The Prototype in your inbox here</a>.



<abbr>N</abbr>ew AI tools&nbsp;are slowly becoming ubiquitous, being added to the software and hardware we use every day (sometimes whether we like it or not). But if we’re using artificial intelligence to perform tasks, search for information and solve problems, what does that mean for the intelligence we’re born with?



To figure this out, a team of researchers conducted a study involving 666 individuals ages 17 and up, representing a diverse population. It first evaluated the extent to which each of them made use of AI tools, then tested their critical thinking skills. The results of the study,&nbsp;<a href="https://www.mdpi.com/2075-4698/15/1/6" rel="noreferrer noopener" target="_blank">which were published</a>&nbsp;in the journal&nbsp;Societies, found that those who used AI tools a lot showed worse critical thinking abilities than those who didn’t use them often or at all. Whether someone used AI tools was a bigger predictor of a person’s thinking skills than any other factor, including educational attainment.



The reason for this is a phenomenon called “cognitive offloading” – where people’s thinking and problem-solving are essentially delegated. Frequent cognitive offloading reduces a person’s ability to independently think and solve problems. “This relationship underscores the dual-edged nature of AI technology,” the study authors wrote. “While it enhances efficiency and convenience, it inadvertently fosters dependence, which can compromise critical thinking skills over time.”



These findings are consistent with other studies that have shown a similar negative impact from AI tools on critical thinking skills. The authors note, however, that other studies show AI tools can be beneficial when they complement critical thinking, rather than offloading it. “Future research should explore strategies to integrate AI tools in ways that enhance rather than hinder cognitive engagement,” they wrote. “Ensuring that the next generation is equipped with the skills necessary to navigate an increasingly complex digital landscape.”



Stay tuned.



<h2 class="wp-block-heading" id="h-bioengineering-off-the-shelf-arteries-for-trauma-patients">Bioengineering Off-The-Shelf Arteries For Trauma Patients</h2>



<abbr>E</abbr>very year, thousands of Americans receive traumatic injuries that damage the arteries that crucially carry blood to their limbs. When these occur, patients basically have three options, Laura Niklason, CEO of regenerative medicine startup Humacyte, told me: they can have a vein grafted from another part of their body, which takes extra time and causes additional trauma; the surgeon can replace the artery with a synthetic plastic one, which creates a high risk of infection; or the limb can be amputated.



Niklason has a fourth option: her company has developed a bioengineered artery called Symvess which can be surgically implanted in a patient’s body, helping to both preserve the limb and reduce risk of infection. Symvess is derived from human donor cells that are grown in a lab on a degradable scaffold, in batches of 200 at a time. Over time, the cells recreate the physical structure of a blood vessel. Once that’s complete, the living cells are then processed away, leaving the artery structure behind, which can then be implanted in a new patient without risk of rejection. Cells in the patient&#8217;s body then occupy the implant, essentially regenerating a new artery.



Last month, Symvess was&nbsp;<a href="https://investors.humacyte.com/news-releases/news-release-details/humacyte-announces-fda-approval-symvesstm-acellular-tissue" rel="noreferrer noopener" target="_blank">approved by the FDA</a>&nbsp;for use in patients with traumatic injuries. That approval was based on studies that showed that the product, which has been implanted in over 600 patients so far, is significantly less likely to cause infection or lead to limb amputation compared to conventional synthetic arteries made from plastic.



For its next steps, the company will be working with Medicare and private insurers to ensure that they will pay for transplant procedures using Symvess. It’s also conducting studies to use Symvess with patients who have kidney dialysis and patients who have pulmonary arterial hypertension. It’s also developing a new, smaller version of its product that could potentially be used for heart bypass patients.



“This is truly a first-of-its-kind technology, and the FDA had to see a lot of data before they could get comfortable and say yes,” said Niklason. “So I’m very glad to be here.”



<h2 class="wp-block-heading">DISCOVERY OF THE WEEK: INACTIVE VIRUSES MAY CAUSE NEURODEGENERATION</h2>



It’s long been known that repeated head injuries can lead to&nbsp;neurodegenerative diseases&nbsp;like Parkinson’s, Alzheimer’s and CTE. But it wasn’t clear exactly how that happened. New research suggests the answer may be&nbsp;dormant viruses, particularly&nbsp;herpes simplex 1, which is found in the brains of over 80% of the population. These findings were reached by growing&nbsp;small tissue models&nbsp;of the brain, some with the herpes virus and some without, which were then struck repeatedly with a piston to mimic concussion.



The study, which was&nbsp;<a href="https://www.science.org/doi/10.1126/scisignal.ado6430" rel="noreferrer noopener" target="_blank">published in&nbsp;Science Signalling</a>, found that the brain tissues with the virus developed signs of&nbsp;Alzheimer’s disease, while those without did not, and that the damage was markedly increased with repeated blows. The researchers plan to conduct future studies that explore whether exposure to&nbsp;antiviral or anti-inflammatory drugs&nbsp;shortly after an injury could help prevent the onset of Alzheimer’s.



<h2 class="wp-block-heading">FINAL FRONTIER: WATER-BASED PROPULSION TEST</h2>



Japanese startup&nbsp;Pale Blue&nbsp;<a href="https://pale-blue.co.jp/news/543/" rel="noreferrer noopener" target="_blank">announced</a>&nbsp;it has signed a deal with Italy-based company&nbsp;D-Orbit&nbsp;to conduct two demonstrations in-orbit of its&nbsp;new ion thruster, which allows satellites both to maneuver and to take themselves out of orbit at the end of their life. Pale Blue’s thruster uses&nbsp;water&nbsp;as a propellant, which potentially offers a cleaner and more sustainable fuel for small satellites.



<h2 class="wp-block-heading">WHAT ELSE I WROTE THIS WEEK</h2>



In my other newsletter,&nbsp;InnovationRx, my colleague&nbsp;Amy Feldman&nbsp;and I&nbsp;<a href="https://www.forbes.com/sites/innovationrx/2025/01/08/innovationrx-medical-debts-will-be-erased-from-credit-histories/">covered a number of healthcare stories</a>, including a drug that significantly improves survival for&nbsp;lung cancer&nbsp;patients, the controversy over&nbsp;pulse oximetry&nbsp;technology and a new study that finds that while&nbsp;LLMs&nbsp;are good at answering medical exam questions, they’re much less proficient at making diagnoses from patient conversations.



<h2 class="wp-block-heading">SCIENCE AND TECH TIDBITS</h2>



My colleague Jeremy Bogaisky&nbsp;<a href="https://www.forbes.com/sites/jeremybogaisky/2025/01/09/amazon-alibaba-fiber-optic-drone-ukraine/">reports</a>&nbsp;that online retailers like Amazon and Alibaba are selling technology that&nbsp;protects drones&nbsp;from having their signals jammed or being remotely taken over.



Ai startup Anthropic’s new funding round will make all seven of its founders&nbsp;billionaires,&nbsp;<a href="https://www.forbes.com/sites/alexkonrad/2025/01/08/anthropic-60-billion-valuation-will-make-all-seven-cofounders-billionaires/">write my colleagues</a>&nbsp;Alex Konrad and Phoebe Liu.



Researchers&nbsp;<a href="https://www.nist.gov/news-events/news/2025/01/novel-quantum-refrigerator-great-erasing-quantum-computers-chalkboard" rel="noreferrer noopener" target="_blank">developed a technique</a>&nbsp;for&nbsp;refrigerating quantum computers&nbsp;in a way that reduces errors and could potentially make them more reliable.



Eating&nbsp;grapefruit&nbsp;can interfere with the action of several kinds of medications, including anti-anxiety drugs and statins. But new research suggests it may be possible to&nbsp;<a href="https://nph.onlinelibrary.wiley.com/doi/epdf/10.1111/nph.20322" rel="noreferrer noopener" target="_blank">grow a variety of the fruit</a>&nbsp;that doesn’t interfere with prescription drugs.



Biologists successfully transplanted&nbsp;chloroplasts–the cells that power photosynthesis– from algae into&nbsp;hamsters, where they&nbsp;<a href="https://www.jstage.jst.go.jp/article/pjab/100/9/100_pjab.100.035/_article" rel="noreferrer noopener" target="_blank">successfully functioned</a>&nbsp;for about three days.



Scientists at Northwestern University&nbsp;<a href="https://news.northwestern.edu/stories/2025/01/trash-to-treasure-industrial-waste-battery/?fj=1" rel="noreferrer noopener" target="_blank">built a battery</a>&nbsp;that uses triphenylphosphine oxide, a common&nbsp;industrial waste product, to store energy instead of metals.



<h2 class="wp-block-heading">PRO SCIENCE TIP: ADD WATER TO YOUR COFFEE BEANS</h2>



Here’s an easy tip to enhance the flavor of your morning cup of&nbsp;coffee: add a little&nbsp;water&nbsp;to the beans before you grind them. This isn’t a new idea–among coffee professionals this is known as the “Ross Droplet Technique”–but new research gives scientific backing to the idea. When beans are ground, they build up a&nbsp;static electric charge, which causes clumping and inconsistent extraction during the brewing process. In new&nbsp;<a href="https://www.cell.com/matter/fulltext/S2590-2385(23)00568-4" rel="noreferrer noopener" target="_blank">research published</a>&nbsp;in the journal&nbsp;Matter, scientists found that the Ross Droplet Technique reduced the static charge during grinding, producing a&nbsp;better-tasting coffee.



<h2 class="wp-block-heading">WHAT’S ENTERTAINING ME THIS WEEK</h2>



I’ve been listening to the concept album&nbsp;Epic&nbsp;by&nbsp;<a href="https://www.epicthemusical.com/" rel="noreferrer noopener" target="_blank">Jorge Rivera-Herrans</a>. It’s a musical-style adaptation of&nbsp;The Odyssey, telling the tale of&nbsp;Odysseus&nbsp;from the end of the&nbsp;Trojan War&nbsp;until his return home to&nbsp;Ithaca. The album is divided into nine sagas of a few songs each, serving as chapters in the story. It’s an amazing version of one of humanity’s oldest stories, and it’s appropriate for it to be sung–that is, after all, how it was originally presented in the time of Homer. It’s streaming on most of the major music services.

Working robots

The Prototype: Study Suggests AI Tools Decrease Critical Thinking Skills

Scott Belsky Says AI Is Collapsing Boundaries Between Traditional Business Functions And Putting More Creative Ability Within Anyone’s Reach.



Tell me a bit about your role at Adobe.



Belsky:&nbsp;My responsibility is overseeing our emerging product groups. These are the groups that include 3D &amp; Immersive. They include the AI-first products, like&nbsp;<a href="https://blog.adobe.com/en/publish/2024/10/14/new-creative-beginnings-gen-ai-redefining-ideation-with-project-concept" rel="noreferrer noopener" target="_blank">Project Concept</a>&nbsp;and a whole roadmap of others. They include the Adobe Stock business, which is going through a re-imagination in the world of AI, and is also responsible for sourcing the data that we train our models on.



Then, of course, design. The design organization spans all of our products and services across the company, and it’s an important part of my role to make sure that the designers are aligned around the experiences we’re delivering to our customers. Upgrading our design system, ensuring that we don’t reinvent the wheel across different products when it relates to a particular component or function.



The Strategy &amp; Corporate Development Group also is in my organization. That’s a group that plots the strategy for Adobe going forward, and explores the edges that may someday become the center for the company, then also takes action on some of the inorganic opportunities that we have on the M&amp;A side.



When you’re looking forward, how large of a role does AI play in corporate strategy?



Artificial intelligence to me is reimagining and refactoring every function of an organization. Whether it be finance and accounting, or HR, or legal, or creative, or marketing, these functions are being re-imagined based on these tools. These tools are also collapsing the boundaries between some of these functions, which allow stakeholders who ordinarily were never able to access the power of these functions to be able to do stuff on their behalf.



When you look at the world of creativity and marketing, a few things become clear. First of all, you have a lot of marketers at the edge, like these social media marketers, and people who are engaging in real-time on behalf of the brand on social platforms, where a lot of the spend goes these days for marketing. Those folks need to be outfitted to be able to create in real time. They need to be able to create variations of marketing campaigns. They need to be able to create content that responds to the moment. So in some ways, the lines are blurry. These marketers are acting more like creators.



Similarly on the creator side, creators are starting to act as if they’re art directors and creative directors when they have the ability to assign tasks to AI. Instead of the three or five options that they have the time to explore, now they can explore 300 options leveraging the superpowers of AI. In some cases, that makes them more tastemakers, creative directors and art directors when they’re at that stage of the process.



When it comes to using tools to get exactly what’s in their mind’s eye reflected in some sort of content or creative asset, they have these new tools at their disposal. That’s one of the things that we’re most excited about at Adobe: Taking our creative pros from the prompt era of AI creativity to the controls era. Right now, the creative pros that I talk to are still tinkering with these models. They have fun putting in prompts and stuff like that. It’s a cool concepting and ideation process, but none of it is commercially ready for them to use for putting out something. They want to make all these changes, so they take it into Photoshop, they take it into Illustrator, they take it into our products. I think that control era, where AI becomes even more granular and they can take the reins, is an important step for creative pros.



It’s been two years since ChatGPT hit and every enterprise decided they wanted to have generative AI. How have you seen businesses change their perspective on AI? How do they look at it now?



I think there are two camps of businesses. There are the businesses that typically are the pragmatists. They’re always late to adopt new technologies, for all sorts of reasons. It could be regulatory, it could be deeply ingrained traditions, patterns and reflexes. And it could be the leaders are somewhat technophobic and they want to wait until they have to change something.



Then there’s the camp that’s always saying, we need to reimagine how we do what we do before someone else does. I think [a] characteristic of that camp that leans forward in new technology is they recognize that not only should they be optimizing their product, but they should also be optimizing how they work to create their product. They’re always saying: if we could have more efficient meetings, or if we could have more efficient usage of our data, or if we could optimize the way that we come up with marketing ideas, or if we could streamline this and streamline that.



In that camp, I feel like there’s a big lean in on this technology. They’re in most cases at least playing with it. Many of them are piloting it in some way, shape or form across specific projects. And those who are still playing and piloting are trying to protect the groups that are doing so by making the objectives more along the lines of learning, as opposed to a business outcome, because they want to make sure that they grasp the knowledge and learnings of these tools and their workflows.



And then there are some that have already started to achieve business outcomes and really refactor. When I look at Adobe, we are using products like Copilot and GitHub for our engineers. That has a material impact on how we work. We use Firefly for our branding and marketing work. We’re already creating variations of content and experiencing this kind of multiplier benefit of using those tools.



From where you sit and what you see, both at Adobe but also in the wider space, what do you see coming online and becoming available and accessible to businesses in the way of AI in 2025?



In 2025, there will be a few themes. Number one is that the memory that these AI tools have of our preferences is going to start to enhance the outcomes that we can achieve. On the content and marketing side, the AI remembering the character consistency of something you’re trying to generate content around—whether it be a product that you’re shooting marketing images of, or a character going through a commercial or a short piece of video that you’re developing. The remembering of your brand components, your preferences for brand tone and copy and all that sort of stuff.



On the personal consumer LLM stuff that we all use, the context windows are getting longer and more sophisticated. These AI agents remembering everything we’ve ever purchased, everything we’ve ever liked or disliked, those preferences are going to start to kick in to uplevel the quality of those experiences.



The rise of [AI] agents is going to surprise all of us. We’ve always had a learning curve when it comes to learning tools, software and processes. That’s the learning curve of starting a job or rising the ranks. I would imagine sometime this year, these agents are going to start to meet us where we are in the software tools that we use. They’re going to know what we know and don’t know about it, and they’ll allow us to leverage a lot of power with natural language.



I think it’s really exciting to know that anyone could be able to use a lot of tools that were out of reach. It [also] allows companies like us to not just target marketers and creative pros, but also the stakeholders of marketing and creative, which is almost everyone in an organization. And it also just drives better usage of the product. If an agent helps you and takes actions on your behalf and can explain and show you, it’s really a holy grail for helping people get value out of the products they use.



For creative and marketing tools at Adobe, what is on the horizon—in terms of new developments, ideas and concepts—that will impact what you do?



A couple come to mind. Number one is this new category of ideation and concepting. That typically is the top of the funnel of any creative process. It typically happens in a very offline or manual way, in the form of people making mood boards of images or bookmarks on Instagram of things that inspire them. As I’ve gone around and talked to a lot of creators, art and creative directors and brand managers, they’re constantly collecting all this stuff. For them to be able to use AI to rapidly expand the surface area of possibility they can explore to find great solutions to problems, that category of concepting and mood boarding in the age of AI is really interesting to us. I think that’s one new category that’s going to emerge that’s going to be important for Adobe, and also for the creative and marketing worlds.



The second thing that I think we will see is the maturing, development and innovation of our foundation models. Firefly is in the leading pack of models out there for media generation. The key difference is that Firefly was only trained on commercially safe content that we have licensed. I think we’re one of the only companies that’s really transparent about exactly how we’ve trained our model—as opposed to politely not answering the question. I think that’s important to a lot of our customers, especially in the enterprise, or [among] artists who want to use new tools like AI, but in the right way. I think these models will continue to get better, but that differentiation will continue to become more important.



While we want to make Firefly as great as possible and have it be the commercially safe model for commercial creativity and marketing, I also think that customers may want to use third-party models for various purposes. They might want to use a certain generative model that was trained to make explosions, or to help typography. I’ve been an advocate internally of us supporting third-party models in the workflows that people have. Adobe’s always been a platform-agnostic company, and whenever a platform like Apple or Windows gets better, so do our tools. Similarly, if a new video model comes out, and then another better one, and another better one, if all of those are accessible to our customers for various use cases, how could we not help our customers make use of them? That’s also something that we’re exploring.



You are one of the creators of the&nbsp;<a href="https://contentauthenticity.org/" rel="noreferrer noopener" target="_blank">Content Authenticity Initiative</a>. Where do you see that going in 2025?



This is one of the true passion projects of my career that I think [is] most important. The Content Authenticity Initiative started over five years ago when we started to see new features emerge in products like After Effects that allowed people to edit real video in a very fast way using AI. So this is before the advent of generative AI, but [it] allowed you to remove objects, mask and do things in video that made something very clear to us: The era ahead was one where we could no longer believe our eyes. In that sense, the question was: What do we do to help people demonstrate that something is real? Or if they’ve edited it, how do we help them show how it was changed so that they can still be trusted? It wasn’t developed to punish bad actors. It was developed to reward good actors who are willing to show the provenance and attribution of their work, which creators love. They want to get credit for their work, so they want to add credentials if they can.



Now, in the age of generative AI and the ubiquity of these technologies, the importance of the Content Authenticity Initiative became 10x, because suddenly all of this stuff, anything in your mind could be generated in a second. In some of the models out there that are trained on the open internet, you can make anyone be doing anything. We realized everyone’s looking for the same solution. The social platforms want a way to expose credentials of assets to show how they were edited or made. The enterprises of the world want to know which model was used to create assets, because if it’s a non-commercially safe model, they may not want to use it in a marketing campaign.



Over the last five years, this has become way bigger than Adobe. We are one of the leaders of a number of people. It’s an open-source, open protocol. It’s a nonprofit organization that runs this with constituents from other companies, including competitive companies. The adoption of content credentials is now north of 3,000. Organizations including all the major camera companies like Nikon, Leica, Canon, Sony. You name it, folks are getting on board.



The vision here is very simple. There’s a cryptographic signature that is put into every asset. It’s not a centralized service. This manifest of data, and how this asset was edited and made, travels with the asset. It’s done in such a way where even if you take a screenshot, it still preserves that information. The platforms that display it have the option—whether it’s YouTube, who’s on board, or Facebook, Meta’s on board—has the option of showing any degree of granularity over those credentials that are attached to that asset. And at the end of the day, us as end viewers can determine whether we can trust it or not.



If we were to be having a conversation a year from today, looking back on what’s happened in 2025 and what to expect in terms of AI for 2026, what would we be talking about?



I think we’d be talking about the new discoveries we’ve had on how to use these reasoning engines. I think we’re still at the infancy of how we leverage these state-of-the-art LLMs. As they gain more powerful and longer context windows, and they start to learn who we are, and in some ways empathize with us and where we are in our knowledge, and what our interests are, et cetera, these reasoning engines are going to play a really interesting role, kind of augmenting the way our brain works.



The thing that I find most interesting these days, personally, is when I talk to friends about how they’re using these tools: To go back and forth and make health decisions about themselves that they never had the resources to make. Or they’re putting a proposal together for a customer, and putting it into the LLM first and saying, what might give my customer pause? Why might my customer not want to do business with me based on this proposal? They’re getting such incredible, interesting responses increasingly based on who they are or what their company does. It’s really magical. It’s like having an instant focus group of exactly who you’d want to be in front of you at every given moment in time, and having those insights come back to you. I think we’re going to be surprised by the power of these reasoning engines for making decisions, for managing personal things, for being more effective in our jobs.



I also think they’re going to start to play a role in the commerce or purchase decisions that we make. The idea that we used to decide what restaurant to go to or what thing to buy based on how many stars it has from strangers is going to be replaced by the wisdom and guidance of an AI that knows us extremely well and can give us a recommendation based on our preferences, opposed to based on what the average person thinks.



If you could give some wisdom to CEOs that are looking forward and trying to map out an AI strategy for 2025, what would you say?



No. 1 is allow your teams to play and pilot this technology. Novelty precedes utility when it comes to adoption of new technology. You might not feel like it’s driving the bottom line immediately, but that play helps them discover the utility that ultimately makes a huge impact in the business.



No. 2 is constantly asking yourself a question. Whenever you’re reviewing the roadmap for a particular function of your organization, or the hiring plan, or the product strategy, you should always be asking yourself: Are we sufficiently refactoring how this works? And are we sufficiently re-imagining what’s possible? We’re in a platform shift now that happens probably once every decade—if not longer—where you actually can fundamentally change the way something operates. Over the coming years, roles will change. The tools that we use to run certain functions will change. And as a leader of an organization, you have to be asking yourself: Are we sufficiently refactoring? Are we sufficiently reimagining how we can achieve our mission?

In this photo illustration, the Adobe, Inc. logo is seen

AI Is Reimagining Every Aspect Of Business, Adobe’s Strategy Lead Says

<a href="https://www.forbes.com/"></a><a href="https://www.forbes.com/search/"></a>In 2025, advancements in AI and voice technology could fulfill a promise that tech giants have been making for more than a decade.



Voice assistants like SIri and Alexa could get a much needed boost by AI agents.NurPhoto via Getty Images



In 2016, when newly-minted Google CEO Sundar Pichai unveiled the Google Assistant as part of his new “AI-first” agenda, he touted the fledgling voice assistant as a tool to help people complete tasks.



“The Google Assistant allows you to get things done, bringing you the information you need, when you need it, wherever you are,” he wrote in a&nbsp;<a href="https://blog.google/intl/en-mena/product-updates/explore-get-answers/2016-google-assistant-en/" rel="noreferrer noopener" target="_blank">blog post</a>&nbsp;at the time.



It was a lofty goal that has, for the most part, fallen short. Too often, the software gets stumped by a request, defaulting to a web search and apologetically saying it can’t help. That led people to relegate voice assistants to simple tasks like setting cooking timers, playing music or controlling their lights. Amazon’s Alexa, released a decade ago, hasn’t fared much better. Siri, the earliest of the bunch, launched by Apple in 2011, has been&nbsp;<a href="https://www.reddit.com/r/SiriFail/" rel="noreferrer noopener" target="_blank">panned most of all</a>.



But as generative AI has gone mainstream over the last two years, it has paved the way for AI “agents”: software that’s specifically programmed to take action or complete tasks on behalf of a user, like booking a reservation or buying something online. And as the “agentic era,” as Pichai calls it, arrives in 2025, the technology has the chance to do something that has to date eluded the big tech platforms: make their voice assistants actually useful.



That means Google Assistant, Alexa and Siri could finally fulfill their promise to act like personal assistants. Instead of just reciting your meeting schedule for the day, like Google Assistant can do now, it might actually be able to book the meetings, reaching out to contacts and finding a time that works for both people. They might have the ability to book your flights and hotels for a big vacation like a digital travel agent, with little more info than trip dates and destination.



Agents are the latest frenzy in the tech industry, with more than 470 platforms devoted to the technology, according to Forrester research. That ranges from big tech giants to smaller startups like LangChain, CrewAI and Play.ai. Beyond consumer features, they can also potentially transform businesses, with agents for customer service or&nbsp;<a href="https://www.forbes.com/sites/rashishrivastava/2024/12/02/cognition-scott-wu-devin-ai/">software development</a>. Deal count for AI agent startups is up more than 81% over the last year, according to PitchBook, with more than $8 billion invested in the space.



“The race is on,” said Steve Jang, a&nbsp;<a href="https://www.forbes.com/lists/midas/">Forbes&nbsp;Midas List investor</a>&nbsp;and founder of the firm Kindred Ventures. “Startups will be competing with the established platforms on who can orchestrate this at much higher fidelity. And who can create much more humanistic and realistic voices and conversations, and access the data and actions that we all want.”



The big tech voice assistants are best poised for such an AI jump start. Google has its marquee model Gemini to beef up its voice searches. Apple earlier this year announced a partnership with OpenAI to use ChatGPT to power some Siri queries. And in the last year, Amazon has invested $8 billion in Anthropic, which makes the powerful Claude chatbot. Google declined to make any of its executives available for interviews. Apple and Amazon didn’t reply to interview requests.



<blockquote class="wp-block-quote is-layout-flow wp-block-quote-is-layout-flow">
“I only use Siri for trivial things that I know it&#8217;s not going to screw up.”Kanjun Qiu, Co-founder, Imbue
</blockquote>



Jang thinks the real innovations will be made in actual voice AI models. Unlike large language models, which underpin services like ChatGPT, voice models are not trained on text and then read aloud by the software. Instead, voice models are trained on actual voice audio, so they can pick up on subtleties in speech, like cadence or emotional cues. Jang has invested in Play.ai, which specializes in voice agents; it’s competing with companies like ElevenLabs, OpenAI and Google that are all working on voice models.



Some, however, are not so convinced that agents will make the big voice assistants exponentially better. Kanjun Qiu, founder of Imbue, which is building agents for coding software, thinks adding more AI to those products will only “incrementally” improve them. She said that new AI features still won’t be a big enough leap for people to trust them. “Delegation as a paradigm is actually really hard for people,” said Qiu. “I only use Siri for trivial things that I know it&#8217;s not going to screw up.”



But she thinks recent improvements in voice AI will help consumers in other ways. For example, more apps will integrate voice features, she predicts. With improved latency and natural language understanding, you’ll be able to give an app instructions and it will carry out that action, Qiu said — like telling an e-commerce app you’d like to return the pair of shoes that don’t fit quite right. (An engineer by training, she said she’s built an app for herself that turns rambling into a to-do list.)



Improvements in AI and voice technology could also unlock hardware ambitions that Silicon Valley has been attempting for years. More than a decade ago, Google infamously faceplanted when it unveiled Google Glass, a piece of smart eyewear that stoked privacy fears and wasn’t very useful. Earlier this month, the company teased a&nbsp;<a href="https://www.forbes.com/sites/richardnieva/2024/12/11/googles-gemini-ai-glasses/">new pair of prototype glasses</a>&nbsp;to be used with Project Astra, Google’s new platform for AI agents. In a demo, the glasses, which are voice-controlled, automatically pulled up a door code from the wearer’s email the moment he looked at the entry keypad. The tech could also conjure up route information about the bus in front of him or the art sculpture he walked by.



Meanwhile, Facebook’s Orion glasses, announced earlier this year, use a combination of voice and hand gestures to control AI tools, like looking at ingredients in your pantry and asking the tech to find a recipe that uses them.



Voice-based innovations also make technology more accessible. Not everyone can read or write or type, but more people have the ability to speak, Jang said. And it’s an increasing preference for young people: 42% of 18-to-29 year olds in the U.S. send voice messages in their chat apps at least weekly, according to a study by YouGov and Vox.



New advancements in AI could make voice tools even more widely used and change the way people interact with their technology. “It makes voice agents — and voice itself — this great new user interface that has been untapped so far in computing,” Jang said.

Internet technology and people’s networks use AI to help with work, AI Learning or artificial intelligence in business and modern technology, AI technology in everyday life.

AI Agents Could Finally Make Siri And Alexa Truly Useful

From the rainforest to the ocean, artificial intelligence is decoding nature’s most complex data patterns and is proving to be a game-changer in conservation.



For as long as there has been wildlife conservation, there has been a need for data. In order to understand ecosystem health, animal behavior and population trends, you have to be able to measure what’s going on in nature. And while there are many ways to do this, from drones to tracking devices, thermal cameras and even rangers on-foot, satellite imagery and camera traps have made study- ing wildlife movement a little bit easier. Not only can conservationists collect ample data, there’s minimal disturbance to animals. When it comes to analyzing the imagery, all that’s really needed is someone who can recognize an animal and there are many training programs to do just that.



But to really understand ecological trends, there are many creatures that are simply too small to be captured through pictures. “With very big animals like elephants, you can see them using satellite data which is great but a significant portion of the ecosystem is hidden,” says Dr Juan Lavista Ferres, who has worked in the field of artificial intelligence (AI) for over 20 years, and is the Chief Data Scientist of Microsoft’s AI for Good Lab. Luckily, those animals that are undetectable in satel- lite imagery make up for their size with sound, which is where acoustic data comes in – and in the case of ani- mal sounds, bioacoustic data. “From a noise perspec- tive, they are there,” laughs Ferres. “There is no corre- lation between the size of an animal and the sound it makes. Very small animals can make a lot of noise and can do very specific frequencies.”



<h1 class="wp-block-heading" id="h-nature-s-silent-signals">Nature’s silent signals</h1>



There’s huge potential in environmental recordings (bio- acoustics are important bioindicators —they are nature’s health signalers) and the good news is that a lot of the data is already available. At the AI for Good Lab, Ferres and his team have continuous surveillance recordings from the Amazon that go back to 1992. The problem is that there are only a handful of scientific experts who can understand what is going on when the data is visualized, and then interpret the frequencies to identify specific animals. “It’s extremely niche. In the case of bioacoustic data, you need a huge level of expertise and there are very few people in the world that can recognize these sounds and frequencies,” adds Ferres.



Another challenge is that going through all of this audio takes time. Ferres says that many of the record- ings he has run non-stop, day and night, for 365 days. “A person would need to be listening to these sounds for a whole year to get through just one year’s worth of data. Now imagine that you have multiple recordings in one forest?” There’s simply too much data for hu- mans to process manually, which is why AI is such a game-changer in conservation – it’s making research a much more efficient process. AI can learn from these experts to automate identifying and classifying animal sounds across large datasets. “We can go through all of the recordings and understand how the health of the ecosystem is changing,” says Ferrers. “Now that we have more compute and better models, we can work with these different types of data, whether it’s images, videos or acoustic data.”



<h1 class="wp-block-heading">Under the sea</h1>



One of the AI for Good Lab’s first projects using acoustic data was with endangered beluga whales in Alaska us- ing underwater sensors. “You can capture the sound of these whales, their vocalizations,” says Ferres. Because sounds travel much further than visual cues, with acous- tic data, researchers can detect (and even potentially identify) different types of whales over vast distances in the ocean, far beyond what would be possible with visual observation methods. “There are particular frequencies in the water that will be very difficult to detect without AI,” adds Ferres. “You can’t solve a problem if you can’t measure it. By doing this, we can actually start measuring the problem.”



Working with different industries and partners throughout the world, Ferres’ hope is that now that the information is in these AI models, more people will be able to use them to track animals over time. All the work that comes out of the AI for Good Lab is also open source which means all of the projects are freely available to re- searchers and organizations worldwide. “It’s a complete philanthropic effort. We give up the idea. We publish everything. All the work we do is for the societal good,” he says. “AI is not just the solution, but sometimes it’s the only solution we have.”

Botswana. Woman photographing giraffe in Chobe National Park with tablet.

Work On The Wild Side: How AI Is Helping Conservation In Africa And The Huge Potential Of Environment Recordings

With The Spurt In Healthtech, AI-powered apps Are Now All The Rage Offering Groundbreaking Interventions In Africa’s Battle against Rising Obesity Rates.



It’s an issue that affects a bulk of the continent’s population.



In Africa, where rising rates of obesity present a growing public health crisis, researchers and innovators are now turning to artificial intelligence (AI)-powered tools to combat the epi- demic. According to the World Health Organization (WHO), in 2022, 2.5 billion adults worldwide were classified as overweight, with over 890 million living with obesity. The problem is becoming especially urgent in low- and middle-income countries. In Africa alone, the number of overweight children under five has increased by nearly 23% since 2000, signaling a dramatic shift in public health needs across the continent.



One of the figures joining the battle against obesity through AI Richmond Essieku, a Ghanaian doctoral student at Texas Tech University. Essieku has reportedly identified 13 genes as potential biomarkers for obesity, with four of them already confirmed. His research is now being translated into a practical solution: an AI- powered mobile app, tentatively titled Generating Artificial Intel- ligence (AI) Based Algorithms to Model and Predict Genotype-by- Socioeconomic Status Interaction for Precision Obesity Detection. This app is designed to model the interaction between genetics and socioeconomic factors in order to provide a more accurate picture of obesity risk in individuals, particularly in Africa, where such nuances are often overlooked.



AI-powered tools like Essieku’s app are not the first examples of how AI can help address the complex factors contributing to obesity. Globally, AI has been used in various ways to combat obesity — from monitoring dietary habits to creating personalized fitness plans. In the United States (U.S.), platforms like Lark Health use AI to offer real-time health coaching, helping users monitor their eating habits and physical activity.



Similarly, Noom, a popular weight-loss app, uses AI to provide personalized advice and insights into user-behavior, which can lead to long-term lifestyle changes.



With the number of smartphone subscriptions in sub-Saharan Africa expected to reach 689 million by 2028, according to Statista, mobile technology offers a promising platform for AI-powered health interventions. AI-driven mobile apps can help individuals monitor their weight, track their food intake, and maintain healthy levels of physical activity. These apps can be tailored to include traditional African foods in dietary rec- ommendations, making them culturally relevant and easier for users to adopt.



The global applications point to the potential for similar systems tailored specifically to African contexts. Abideen Ola- supo, founder of MyAIFactchecker, an AI fact-checking platform based in Nigeria, highlights the importance of culturally-local- ized solutions. “We want to build people’s confidence in explor- ing how they can interact with AI to find safe and effective ways to combat obesity,” Olasupo tells FORBES AFRICA. “For a long time, many have turned to harmful concoctions that negatively impact their health. Africa’s unique challenges cannot be under- stood from the outside, which is why we’ve been researching the solutions people seek online. Obesity, in particular, is a sensitive issue that many feel ashamed to confront.”



MyAIFactChecker is one of several platforms working to ensure that AI solutions are tailored to Africa-specific chal- lenges. Olasupo explains that the platform sifts through different themes, including health and nutrition, to ensure that data is lo- calized and actionable. He notes that combating obesity in Africa requires integrating AI into the daily lives of Africans in ways that are culturally relevant and accessible.



AI’s ability to personalize interventions is especially critical when considering the unique challenges Africa faces in address- ing obesity. Many Africans live in ‘food deserts’ — areas where healthy food is difficult to find or too expensive to buy. Physical activity can also be limited by unsafe environments or long work- ing hours, making traditional approaches to weight-loss difficult to implement.



Africa’s healthcare infrastructure faces significant challeng- es, including a shortage of healthcare professionals and limited access to health facilities. AI is helping to alleviate these issues by offering remote monitoring and telehealth services for obesity management. In countries like Nigeria and Ghana, where access to healthcare is often limited, AI-powered remote health services could provide essential support to individuals looking to manage their weight.



AI-powered health apps are gaining popularity in Africa, also offering personalized recommendations for diet and exercise.



In Nigeria, the startup WellNewMe is using AI to provide us- ers with customized health assessments based on their lifestyle choices. The platform uses machine learning algorithms to ana- lyze users’ eating habits, physical activity, and medical history. It then generates recommendations to improve their health and reduce the risk of obesity-related conditions. For example, users can receive tailored meal plans, exercise routines, and lifestyle tips, all customized to their specific health profiles.



A similar initiative is underway in Ghana, where mPharma, a healthtech company, has developed the mymutti Mobile app. While primarily used to offer affordable medications, the app has recently integrated an AI-powered feature that tracks users’ body mass index (BMI) and offers recommendations for healthier liv- ing. Users receive customized tips on maintaining a balanced diet, increasing physical activity, and managing weight through simple lifestyle changes.



In Rwanda, AI is being used to improve access to healthcare in rural areas where obesity is on the rise. The Babyl Health plat- form, which uses AI to provide remote medical consultations, has been instrumental in helping individuals in rural communities access healthcare services. Through the app, users can consult with doctors about weight-related issues, receive prescriptions, and get referrals to local health facilities for further treatment.



Farming innovations to support the health sector



AI’s role in combating obesity extends beyond healthcare to the agricultural sector, where it is being used to improve the avail- ability of healthy, affordable food. In Nigeria, the startup Zenvus is leveraging AI to help farmers increase crop yields and reduce food waste. By using AI-powered sensors, Zenvus collects data on soil conditions, weather patterns, and crop health. This data helps farmers make informed decisions about when to plant, irri- gate, and harvest their crops, resulting in higher yields and more efficient food production.



Another example is FarmPin, a South African agritech com- pany that uses AI and satellite imagery to monitor crop health and predict food shortages. By analyzing data from farms across the country, FarmPin helps farmers optimize their production and ensure a steady supply of fresh produce to local markets. This, in turn, makes healthy food more accessible and affordable, reducing reliance on processed foods that are often linked to ris- ing obesity rates.



Public health experts are also optimistic about the potential of AI to offer scalable solutions for obesity prevention in Africa. Opeyemi Adam, a public health expert based in Lagos, explains that AI could help overcome some of the challenges of limited healthcare resources in many African countries. “Obesity is often thought of as a personal issue, but it’s deeply systemic,” Adam says. “AI offers a way to analyze large amounts of data and iden- tify trends that would otherwise go unnoticed, which is crucial in places where healthcare resources are stretched thin.”



Adam also points out that many Africans lack access to per- sonalized health advice, something AI could help bridge. “With AI, we could create more personalized health plans that take into account not just the individual’s body weight or fitness level, but also their lifestyle, environment, and even their genetic makeup,” he says.

Scientists Working on Powered Exoskeleton Project in Lab.

Weighing In On The Use Of AI In Helping Combat Obesity

When it comes to controversial advertising campaigns, they usually don’t have to do with income taxes. But a short-lived campaign from enterprise software company Intuit, which makes TurboTax,&nbsp;crossed that line and raised the ire of accountants.



<a href="https://www.forbes.com/sites/kellyphillipserb/2024/10/16/intuit-backpedals-after-insulting-accountants-with-latest-ad-campaign/">Forbes’&nbsp;Kelly Phillips Erb writes</a>&nbsp;about the saga, in which a consumer-targeted message intended to be a bit cheeky and punny&nbsp;ended up insulting a group of professionals who actually rely on Intuit’s products. The centerpiece of the campaign is an advertisement in which a young woman announces to a bland-looking accountant that she’s leaving him for a younger and more fashionable man on her phone “who charges me less but gives me more” as part of a “tax break-up.” The younger man represents Intuit’s “TurboTax Full Service plan,” featuring human tax preparers who work with customers online or in person to complete their taxes.



Tax preparers, who have competed with TurboTax for many years,&nbsp;felt the campaign went too far. The unflattering portrayal of the accountant, plus the barrage of messaging telling consumers to “break up” with them, made many individuals and organizations upset. American Institute of Certified Public Accountants President and CEO Barry Melancon issued a statement saying, “The crucial role of CPAs, who serve as expert and trusted advisors to business owners and individuals, was not reflected in an unfortunate recent ad campaign launched by Intuit.” And the National Association of Tax Professionals said it will no longer accept Intuit’s sponsorships or exhibits at its conference because of the campaign’s “direct conflict with our members’ interests.” Predictably, Intuit decided to phase out the campaign.



While TurboTax is a well-known Intuit product, it’s only one. The company also owns QuickBooks and Mailchimp—financial management and email marketing platforms used by many businesses, including accountants—as well as consumer budgeting tool Credit Karma. In its last fiscal year, the bulk of Intuit’s growth came from those using its enterprise financial planning and email software. TurboTax online made up less than a quarter of the company’s total fiscal year revenues.



Intuit may be seeking more consumer tax business, but it may have damaged its larger customer base by trying to be funny. An average consumer may have giggled at the ad, but it seems unlikely that it would end up getting the desired results,&nbsp;especially since most consumers aren’t thinking about income tax prep in October. Instead, with this campaign, Intuit might have inspired more financial professionals to break up with&nbsp;them.



Finding the right emotional balance is an important tool to make a campaign successful. Reticle AI is using artificial intelligence to determine the emotions that are present in campaigns, helping marketers strike the proper tone to get their messages across. I talked to Vice President of Business Development Tricia Allen about how this works. An excerpt from our conversation is later in this newsletter.



MARKETING MATTERS


<div class="wp-block-image">
<figure class="aligncenter size-full is-resized"><img loading="lazy" decoding="async" width="1024" height="683" src="https://cms.forbesafrica.com/wp-content/uploads/2024/09/GettyImages-2169510342.jpg" alt="" class="wp-image-63122" style="width:608px;height:auto" srcset="https://cms.forbesafrica.com/wp-content/uploads/2024/09/GettyImages-2169510342.jpg 1024w, https://cms.forbesafrica.com/wp-content/uploads/2024/09/GettyImages-2169510342-300x200.jpg 300w, https://cms.forbesafrica.com/wp-content/uploads/2024/09/GettyImages-2169510342-768x512.jpg 768w" sizes="auto, (max-width: 1024px) 100vw, 1024px" /><figcaption class="wp-element-caption"> (Photo by Dilara Irem Sancar/Anadolu via Getty Images)</figcaption></figure></div>


Google is making a change to its Local Services Ads business next month that&nbsp;may impact roughly half of businesses advertising on the search engine,&nbsp;<a href="https://www.forbes.com/sites/brandonkochkodin/2024/10/22/googles-upcoming-ad-change-could-spell-disaster-for-millions-of-unprepared-small-businesses/">reports&nbsp;Forbes’&nbsp;Brandon Kochkodin</a>. Starting November 21, the local ads that pop up at the top of search results for users will be blocked if they aren’t linked to a Google Business Profile. This free profile indicates how businesses appear in search and is fully customizable by the business. But businesses need to be verified as legitimate through video or phone authentication, or by receiving a postcard. The authentication is an extra step, but it’s important that local businesses do it sooner rather than later. Studies have shown that 96% of people learn about local businesses online, and four out of five of those searches convert into business. In fact, nearly half of all Google searches are for local businesses.



SOCIAL MEDIA


<div class="wp-block-image">
<figure class="aligncenter size-full is-resized"><img loading="lazy" decoding="async" width="1024" height="662" src="https://cms.forbesafrica.com/wp-content/uploads/2024/05/GettyImages-1513152452.jpg" alt="" class="wp-image-61892" style="width:632px;height:auto" srcset="https://cms.forbesafrica.com/wp-content/uploads/2024/05/GettyImages-1513152452.jpg 1024w, https://cms.forbesafrica.com/wp-content/uploads/2024/05/GettyImages-1513152452-300x194.jpg 300w, https://cms.forbesafrica.com/wp-content/uploads/2024/05/GettyImages-1513152452-768x497.jpg 768w" sizes="auto, (max-width: 1024px) 100vw, 1024px" /><figcaption class="wp-element-caption"> (Photo by Hakan Nural/Anadolu Agency via Getty Images)</figcaption></figure></div>


Two social networks are making changes that may impact user safety. For as long as X was known as Twitter, users have been able to block other accounts—a step often taken when a user is bothered or harassed by someone. Previously, the blocked account couldn’t see or engage with posts from the user who blocked them. Now, users&nbsp;<a href="https://www.forbes.com/sites/antoniopequenoiv/2024/10/17/x-will-soon-let-users-see-tweets-from-people-who-block-them-the-controversial-change-explained/">can see all posts</a>&nbsp;from accounts that have blocked them, though they still can’t engage with them.&nbsp;X says this change allows “greater transparency,” so users can see if those who have blocked them are trying to “share and hide harmful or private information.” But critics have taken the other side of the argument: Stalkers and harassers can keep following accounts of those who have blocked them. X owner Elon Musk, who has made scores of controversial changes to the platform since he bought it in 2022, previously posted the blocking function made no sense. After this change was made, competing social media platform Bluesky gained 1.2 million users in two days,&nbsp;<a href="https://www.socialmediatoday.com/news/bluesky-gains-momentum-wake-xs-blocking-changes/730371/" rel="noreferrer noopener" target="_blank">Social Media Today reported.</a>



<a href="https://www.forbes.com/sites/lindseychoo/2024/10/17/instagram-sextortion-child-safety/">Changes to Instagram</a>&nbsp;intended to prevent “sextortion” scams and protect teens are much less controversial. Now, users are unable to take screenshots or recordings of disappearing images, which are only intended to be viewed once. The Meta-owned platform is also testing safety notices to let teenagers know when they are exchanging messages with someone out of the country. Images flagged to contain nudity will be automatically blurred for users younger than 18. And accounts with “scammy” behaviors will not be able to see users’ following and followers lists, a common way some bad actors get information about a target’s friends and family.



BRANDS + MESSAGING


<div class="wp-block-image">
<figure class="aligncenter size-full is-resized"><img loading="lazy" decoding="async" width="1024" height="683" src="https://cms.forbesafrica.com/wp-content/uploads/2024/07/GettyImages-1536268798.jpg" alt="" class="wp-image-62351" style="width:556px;height:auto" srcset="https://cms.forbesafrica.com/wp-content/uploads/2024/07/GettyImages-1536268798.jpg 1024w, https://cms.forbesafrica.com/wp-content/uploads/2024/07/GettyImages-1536268798-300x200.jpg 300w, https://cms.forbesafrica.com/wp-content/uploads/2024/07/GettyImages-1536268798-768x512.jpg 768w" sizes="auto, (max-width: 1024px) 100vw, 1024px" /><figcaption class="wp-element-caption">(Photo by Nikos Pekiaridis/NurPhoto via Getty Images)</figcaption></figure></div>


Netflix&nbsp;<a href="https://www.forbes.com/sites/lindseychoo/2024/10/17/netflix-earnings-streaming-beats-expectations-subscriber-growth/">reported earnings</a>&nbsp;last week, beating Wall Street’s expectations. The streaming giant made $9.8 billion in revenue in its most recent quarter, and gained 5.1 million subscribers. Now, nearly 283 million households have Netflix subscriptions, but the rate of new subscriber gain is slowing down.&nbsp;<a href="https://www.forbes.com/sites/dbloom/2024/10/17/netflix-added-51-million-subscribers-can-we-talk-about-what-matters/">Forbes&nbsp;senior contributor David Bloom writes</a>&nbsp;that&nbsp;for Netflix, this really shouldn’t be seen as a negative. After all, there is a finite amount of Netflix subscriptions that can be bought. What is more important, Bloom writes, is viewer engagement. Netflix reported that the average paid membership has two hours of engagement per day. It has an 8.2% viewership share in the U.S., according to Nielsen data. Netflix first became profitable in 2003, and to date only one other streaming service—Disney+—has posted a profit. With more emphasis on engagement, Netflix also seems to be the first of the streamers to start caring more about actual viewers and interaction.



<h4 class="wp-block-heading" id="h-on-message">ON MESSAGE</h4>



<h3 class="wp-block-heading" id="h-how-reticle-ai-uses-technology-to-determine-emotions-in-ads">How Reticle AI Uses Technology To Determine Emotions In Ads</h3>



Emotion is important to marketing. Nostalgia, joy, adventure, curiosity and stress can all inform how consumers make purchases, choose what to eat or drink, or select service providers or leisure activities. Reticle AI uses artificial intelligence to determine the emotions that are present in marketing campaigns, as well as the emotions that consumers are experiencing when they see them, and makes recommendations for an optimal outcome. I talked to Vice President of Business Development Tricia Allen about how their method works and how it can get results. This conversation has been edited for length, clarity and continuity.



How can you use AI to determine things like sentiment and nostalgia?



Allen:&nbsp;We’ve trained our AI technology to understand emotional sentiment within digital content. We’re almost curating content into various emotions. If a brand wants to be associated with nostalgia, [or being] uplifting, joyful, nurturing, they can align their message within content that has that emotional sentiment. We can control how the content potentially makes someone feel. If you’re consuming content that makes you feel nostalgic, nurturing, uplifting, chances are you’re in that frame of mind. If you get served an ad that is in that nostalgic, nurturing tone, you’re more predisposed to take action. That action is that connectivity and attention that the brand is getting because it’s being served at the right time.



[AI has] that ability to take the universe of digital content, really cull it down, and categorize it into these emotional categories because of its understanding of human emotion.



There’s a lot of talk about AI systems not always giving accurate responses. With something difficult to quantify like emotion, how do you make sure that what comes out of your system is accurate to how people actually would feel?



While AI is extremely relevant and helpful from a marketing standpoint, you can’t get rid of that human oversight. I really believe strongly in that for many reasons. There’s bias that’s involved because the technology is only as good as the person who is training the platform, and if they have inherent bias, then the tech has bias. You need more than one person to be vetting that. In addition, what nurturing means to me might mean something completely different to you.



We actually have human labelers. We have them go in and vet our data on an ongoing basis, just to have those human eyes on it. We’ll go through, and if the technology has said this content is deemed to be nurturing, we want to see what a human actually thinks. We have that human component to make sure, because AI is not perfect. Even when it does get better, I still strongly believe you’re going to need human oversight, especially when you’re talking about emotion.



In your research, what are some of the emotions that people are reacting to the most, and what are some more prevalent themes?



A lot of the themes that people are looking for from a marketing standpoint are around joy and that feeling of happiness, and conversely, that feeling of stress and pressure. [It’s] not necessarily a positive emotion. Maybe it’s a negative emotion, because maybe I’m a vacation company and I want to provide you with a solution to help you overcome that stress.



If you are able to tap into emotion—both in marketing and also what is being looked for—what would be the perfect marketing campaign?



To me, the perfect marketing campaign is [one in which] you first do the insights work to understand the emotional footprint of that brand. How do people think and feel about your brand right now? How do you want them to feel about your brand? Then ensure that creative or the ad itself is designed to hit those key emotions. We’re kind-of that end piece. You’ve established how people think and feel. You’ve established: Do I want them to feel this way? If no, where are there distinctions that we can draw on to differentiate from the competition and then make sure that creative is dialed in?



The biggest area where I see brands fall down is they try and tap into emotion, but there’s a disconnect between the creative strategy and the insights work that was done. The insights work is never being brought to life, and then the creative is a bit of a disconnect. When all of those pieces line up, and you take that emotionally designed creative [and] put it in the same environment from a digital placement perspective, that’s the trifecta of emotional marketing.



FACTS + COMMENTS



When the World Series between the Los Angeles Dodgers and New York Yankees begins on Friday,&nbsp;it will be Sho time—for Dodgers star Shohei Ohtani, that is. With the Japanese phenom in the playoffs, Major League Baseball has actually seen&nbsp;<a href="https://www.forbes.com/sites/mattcraig/2024/10/17/ohtani-mania-has-mlb-playoff-viewership-soaring-in-japan/">more October viewership</a>&nbsp;in his home country than in the U.S.



12.9 million:&nbsp;Japanese viewers of NLDS Game 5 on October 11, when the Dodgers beat the San Diego Padres to advance to the National League Championship Series. The U.S. had 7.24 million viewers, a record for a division series



$700 million:&nbsp;Value of Ohtani’s record-setting 10-year contract



‘I hope to continue to perform well as well for the fans in Japan’:&nbsp;Ohtani&nbsp;<a href="https://dodgersnation.com/shohei-ohtani-acknowledges-fans-in-japan-watching-dodgers-postseason/2024/10/21/" rel="noreferrer noopener" target="_blank">told Dodgers Nation</a>, after saying that the team has also been “playing really well”



STRATEGIES + ADVICE



Yes, there is such a thing as too much positivity. Here are some ways to&nbsp;<a href="https://www.forbes.com/sites/jackkelly/2024/10/11/managers-are-glossing-over-workplace-issues-and-pushing-toxic-positivity/">avoid glossing over</a>&nbsp;your employees’ challenges and support your team through hard times.



Time management is difficult for executives. Here are <a href="https://www.forbes.com/sites/jackkelly/2024/10/17/top-time-management-apps/">five apps</a> that can help you get more control over your work day.

Chat Bot Service Concept – Enhancing Customer Experience with Virtual Assistant

Can AI Capture The Emotional Side Of Marketing?

Google’s projections align with a previous study — the Google/IFC e-Conomy Africa 2020 report — which anticipates 167 million new mobile users in sub-Saharan Africa by 2025. This will bring the total number of mobile subscribers in the region to 623 million.



A new report released by Google estimates that artificial intelligence (AI) could contribute over $30 billion to sub-Saharan Africa’s economy in the coming years.



The report, The Digital Opportunity of Sub-Saharan Africa, was conducted independently by Public First, on the region with 70% of its population under the age of 30.



Internet access across sub-Saharan Africa varies significantly, with about 66% of people online in southern Africa, but only 24% in Central Africa, according to Statista. The World Bank reveals that internet usage in West Africa hovers around 30%. The report predicts that even a 1% increase in internet access could lead to nearly a 6% rise in gross domestic product (GDP).



Speaking on the report, Alex Okosi, Managing Director of Google Africa, explains AI’s potential for the region to FORBES AFRICA: “While the report doesn&#8217;t pinpoint specific countries as front-runners in AI-driven growth, it highlights sectors where AI&#8217;s impact is already evident or holds substantial promise&#8230; It emphasizes its [AI] value in the area of automating tasks, boosting efficiency and productivity; and creating new products and services.”



According to the data, 76% of online adults in sub-Saharan Africa identify as entrepreneurs, further highlighting that every dollar invested in digital technologies in Sub-Saharan Africa is projected to generate over $2 in broader economic value by 2030. This finding illustrates the high return on investment (ROI) for digital advancements, particularly in sectors like healthcare, agriculture, and environmental sustainability.



“For the first time, we’ve included an AI curriculum to help business owners understand how they can leverage AI technologies for growth,” Okosi notes. “To further address infrastructure needs, investments like the Equiano subsea cable and our new Cloud region in Johannesburg are critical steps. Additionally, public-private partnerships will play a key role in closing the skills gap by delivering AI training programs and fostering a supportive policy environment that encourages innovation and addresses ethical concerns.”



Entrepreneurs across the continent are already recognizing the potential of AI. Ifeoluwa King, co-founder of VendorStack, a Nigerian social commerce startup, believes AI will touch every sector. “AI will streamline processes, increase efficiency, and cut costs. It will spark innovations from Africa’s creative youths and improve productivity, which will lead to increased revenues,” King tells FORBES AFRICA.

Google

‘AI Curriculum To Help Business Owners’: AI Could Add $30 Billion To Sub-Saharan Africa’s Economy, Report Finds

Navigating the complexities of our future is big business – and big money. Here are some of the more eminent global futurists offering fresh insights into future tech and a better tomorrow.



At the South by Southwest (SXSW) festival in Austin, Texas, there’s one particular session where people start queueing around four in the morning to get a seat. It’s not one of the big-name music performances or a movie screening with awards buzz, it’s a keynote by Amy Webb.



Webb, who is the founder and CEO of the Future Today Institute (FTI), is a futurist. She’s a professor of strategic foresight, a data and research-driven field that unpacks what could be in an increasingly complex world. Webb’s 2024 Tech Trends Report, which is downloaded more than a million times every year, is made up of just under a thou- sand pages of insights businesses rely on to navigate the future.



<blockquote class="wp-block-quote is-layout-flow wp-block-quote-is-layout-flow">
IN HER TALK AT SXSW, AMY WEBB SAID THAT THE NEXT STAGE OF AI IS MERGING DIGITAL SUPERCOMPUTERS WITH HUMAN TISSUE. THIS IS SOMETHING WEBB REFERS TO AS ORGANOID INTELLIGENCE, OR O-AI. SHE PREDICTS THAT BY 2039, O-AI WILL BECOME THE FOREFRONT OF COMPUTATIONAL EVOLUTION, MERGING HUMAN-LIKE DECISION-MAKING CAPABILITIES WITH THE SPEED OF TRADITIONAL COMPUTERS. IN SIMPLE TERMS, WEBB SAYS THAT WE WILL BE ABLE TO MAKE BIOCOMPUTERS OUT OF HUMAN BRAIN CELLS – WE WILL GROW COMPUTERS RATHER THAN BUILD THEM.
</blockquote>



Predicting what comes next may be big business but it’s not necessarily new. While modern futurists use data to forecast pos- sible future scenarios and spot emerging opportunities, the science of probability – long-term planning – has been around since the 1940s. And before futurology became an academic field and job role, science fiction authors like H.G. Wells, Jules Verne and Arthur C. Clarke, were recognized as futurists with an uncanny ability to foresee the world to come. Today, there are a number of well-known futurists sought out by industry giants (and the government) to steer business strategy and adapt to disruptive change. Businesses are hiring futurists to tell them the things they don’t want to hear so that they can prepare for unfore- seen inevitabilities.



Meet the futurists



As the founder of Singularity University and the XPRIZE Foundation, Peter Diamandis is a Silicon Valley digital futurist and engineer who has started over 25 companies. Now in his 60s, Diamandis (whose net worth has been reported to be around $200 million) has many innova- tive projects underway. While tickets to attend his membership-only Abundance Summit begin at $10,000, Diamandis is attempting to change the world with his large-scale XPRIZE competitions that tackle domains like climate, space, biodiversity and deep tech. According to the Key Speakers Bureau, Diamandis’ speaking fee starts at $60,000 – and that’s for a virtual, not in-person, appearance.



As artificial intelligence (AI) becomes more and more popular, there’s a good chance that futurist Ray Kurzweil will be mentioned. Kurzweil, who has received 20 honorary doctorates is a computer scientist and futurist with a 30-year track record of accurate predic- tions – his written predictions, of which there are 147, have an 86% accuracy rate. Kurzweil’s first predictions around AI came out in 1999 and while many people felt that his views were radical when he said that computers would be as intelligent as humans, his estimated date of 2029 doesn’t seem so far off. Kurzweil’s net worth is reportedly around $30 million, most of which comes from the four companies he success- fully founded based on his research and academic prizes. One of his boldest predictions is that immortality will be achievable by 2030 with nanobots – tiny robots that can repair human bodies on a cellular level.



In an interview, he famously said: “By the time we get to the 2040s, we’ll be able to multiply human intelligence a billionfold. That will be a profound change that’s singular in nature. Computers are going to keep getting smaller and smaller. Ultimately, they will go inside our bodies and brains and make us healthier, make us smarter.”



Webb, who has been called ‘one of the five women changing the world’, recently wrote a book called The Genesis Machine with fellow futurist, Andrew Hessel. The book, which won a Gold Axiom Medal, explores synthetic biology, a controversial and emerging techno-science. Biotech, which is something both Webb and Hessel focus on, was a key part of Webb’s 2024 Tech Trends Report. In her talk at SXSW, she said that the next stage of AI is merging digital supercomputers with human tissue. This is something Webb refers to as organoid intelligence, or O-AI. She predicts that by 2039, O-AI will become the forefront of computational evolution, merging human-like decision-making capabilities with the speed of traditional computers. 



In simple terms, Webb says that we will be able to make biocomputers out of human brain cells – we will grow computers rather than build them. Hessel is a pioneer in synthetic biology, microbiologist, geneticist and futurist. He is also the co-founder of a company called Hu-mane Genomics Inc. that engineers synthetic viruses to target cancer cells. “Biology is a very different medium than computing but there are fundamental similarities,” he wrote in a blog post talking about synthetic biology in Af-rica. “Africa is geographically well-positioned to benefit from synthetic biology. Much of the country is in the biological power band of the world, blessed with ample sunlight, fresh water, fertile soils, and incredible biological diversity. But it is also vulnerable to changes in climate, particularly shifts in rainfall and heat waves,” he continues. Hessel has seen first-hand how AI tools are supercharging all aspects of digital biology and predicts that DNA printers are on the cusp of a major upgrade. He believes that in the next five years, we will be able to print and assemble microbial-scale genomes — at reasonable cost and turnaround times.



In the workplace



No matter whether you call them futurists, fu-turologists, foresight practitioners or horizon scanners, it’s becoming increasingly common for organizations to employ professionals with these titles. According to the Univer-sity of Houston, around 25% of Fortune 500 companies have futurists on staff. They’re also commonplace at government agencies as well as non-profits. Andrew Bolwell, for example, is HP’s chief disrupter. In his role, he unpacks megatrends like microfluidics – working with tiny amounts of fluid with precision. Even though printers already do this, Bolwell is talking about even smaller amounts, fluids the fifth of a size of a human cell (or a thousand times smaller than a raindrop). With the microfluidics market projected to be worth $64.94 billion by 2029, according to Mordor Intelligence, the possibility of 3D-printing vital organs may become a reality.



<blockquote class="wp-block-quote is-layout-flow wp-block-quote-is-layout-flow">
&#8220;BY THE TIME WE GET TO THE 2040S, WE’LL BE ABLE TO MULTIPLY HUMAN INTELLIGENCE A BILLIONFOLD. THAT WILL BE A PROFOUND CHANGE THAT’S SINGULAR IN NATURE. COMPUTERS ARE GOING TO KEEP GETTING SMALLER AND SMALLER. ULTIMATELY, THEY WILL GO INSIDE OUR BODIES AND BRAINS AND MAKE US HEALTHIER, MAKE US SMARTER.”&#8211; Ray Kurzweil
</blockquote>



When Dr Werner Vogels, Chief Technology Officer (CTO) and Vice President at Amazon, one of the globe’s most innovative companies, looks into the future, the world takes note.



The well-known futurist, who releases tech predictions for the year and beyond, also happens to be one of the highest-paid CTOs worldwide – it’s no wonder his live speaking fee reportedly starts at $100,000. One of Vo-gels’ predictions is that women’s healthcare, femtech, is on the rise. “Women’s healthcare is not a niche market,” he writes in his tech predictions report. “At AWS, we’ve been work-ing closely with women-led startups and have seen first-hand the growth in femtech. In the last year alone, funding has increased 197%.” 



With increased access to capital, technologies like machine learning, and connected devices designed specifically for women, Vogels believes that the world is at the precipice of an unprecedented shift, not only in the way women’s care is perceived, but how it’s administered. From women in rural areas who have been historically underserved medically to female athletes whose training was modelled on what worked for men, femtech will finally see women’s care move from the fringes to the forefront.



As companies strive to achieve aggressive sustainability targets, there are now climate and environmental futurists emerging as key strategic advisors, guiding businesses in navigating the complexities of climate change, resource management and eco-friendly practices to secure long-term viability and environmental stewardship. For Gerd Leonhard, a German futurist, green has become the new digital and sustainable is the new profitable. In a recent keynote, he said that he’s optimistic about climate change and global warming because we already have the tools we need to solve it. “We’re going to suffer for 15 to 20 years in the process of fixing this,” he predicts, “but we’re now getting ready to do this… technological advancements offer solutions that can help tackle these challenges. It will be essential to find a balance between economic progress and environmental preservation.”



Leonhard, like other consulting futurists, is working with companies to help them understand the next decade and develop foresight. “It’s not predictions, it’s understanding, intuition and imagination,” he explains in a podcast, adding that being a futurist is similar to being a therapist. His job is to essentially point his clients in the right direction; clients who happen to be tech giants, business leaders and government officials around the globe.



The future, today



There’s no question that history is happening now and futurists provide a way to understand current trends as well as anticipate what’s to come. Futurism is essentially risk management because every industry is experiencing dynamic change. In the last 50 years we’ve made more progress than in the previous 250,000 years and it’s been said that the next 10 years will bring more change than the previous 100. Of course, the future is unpredictable and futurists are not fortune tellers – they get things wrong. However, their value lies in their ability to analyze data and identify patterns. By incorporating foresight into strategic planning, businesses, governments and other organizations can adapt more effectively to emerging trends, mitigate risks and seize the moment. In a world characterized by rapid technological advancements, the role of the futurist has become increasingly indispensable in shaping a better tomorrow.

GUI (Graphical User Interface) concept.

The Fortune Of Foresight: Futurists On Tomorrow’s Tech

<h2 class="wp-block-heading" id="h-topline">TOPLINE</h2>



With companies like OpenAI, Google and Meta dropping increasingly sophisticated artificial intelligence products, crowdsourced rankings have emerged as a popular—and virtually only practical—way of determining which tool works best, and LMSYS’s Chatbot Arena has become possibly the most influential real-time gauge.



<h2 class="wp-block-heading" id="h-key-facts">KEY FACTS</h2>



While most organizations choose to measure their AI models against a set of general capability benchmarks that cover tasks like solving math problems, programming challenges or answering multiple choice questions across an array of university-level disciplines, there is no industry benchmark or standard practice for assessing large language models (LLMs) like OpenAI’s GPT-4o, Meta’s Llama 3, Google’s Gemini and&nbsp;<a href="https://www.forbes.com/sites/roberthart/2024/06/26/anthropics-claude-what-to-know-about-chatgpt-rival-after-latest-model-trounces-industry-giants/">Anthropic’s Claude</a>.



Even small differences to factors like datasets, prompts and formatting can have a huge impact on how a model performs, and when companies choose their own evaluation criteria, it can make it hard to fairly compare LLMs, Jesse Dodge, a senior scientist at the Allen Institute for AI in Seattle, told Forbes.



The difficulty in comparing LLMs is magnified given how closely leading models score on many commonly used benchmarks, with some companies and tech executives&nbsp;<a href="https://x.com/Ahmad_Al_Dahle/status/1780993216118210899" rel="noreferrer noopener" target="_blank">claiming</a>&nbsp;<a href="https://x.com/AnthropicAI/status/1803790676988920098" rel="noreferrer noopener" target="_blank">victory</a>&nbsp;over rivals with differences as narrow as 0.1%., so close it would likely go unnoticed by everyday users.



Community-built leaderboards deploying human insight have emerged, and in recent years their popularity has exploded in step with the steady boom of new AI tools like ChatGPT, Claude, Gemini and Mistral.



<a href="https://chat.lmsys.org/?leaderboard" rel="noreferrer noopener" target="_blank">The Chatbot Arena</a>, an open source project built by research group&nbsp;<a href="https://lmsys.org/about/" rel="noreferrer noopener" target="_blank">LMSYS</a>&nbsp;and the University of California, Berkeley’s Sky Computing Lab, has proven particularly popular and has built AI leaderboards by asking visitors to compare responses from two anonymous AI models and vote which one is best.



Its scoreboards rank more than 100 AI models based on nearly 1.5 million human votes so far, covering an array of categories including long queries, coding, instruction following, maths, “hard prompts” and a variety of languages including English, French, Chinese, Japanese and Korean.



<h2 class="wp-block-heading" id="h-what-s-the-best-ai-model-on-chatbot-arena">WHAT’S THE BEST AI MODEL ON CHATBOT ARENA?</h2>



The top five AI models on Chatbot Arena’s overall leaderboard are:



<ol class="wp-block-list">
<li>GPT-4o</li>



<li>Claude 3.5 Sonnet</li>



<li>Gemini Advanced</li>



<li>Gemini 1.5 Pro</li>



<li>GPT-4 Turbo</li>
</ol>



<h2 class="wp-block-heading" id="h-what-to-watch-for">WHAT TO WATCH FOR</h2>



Figuring out how to evaluate AI models is set to become increasingly important as more AI tools are rolled out and adopted across society. While benchmarks are important, Vanessa Parli, director of research at Stanford University’s Institute for Human-Centered AI, told Forbes they are also important as “goals for researchers to hit when developing models.” It is important to remember that “not all human capabilities are quantifiable” in a way that we can accurately measure but are nonetheless desirable to have in AI models, Parli said. There is also a clear need for benchmarks to assess traits like “bias, toxicity, truthfulness and other responsibility aspects,” especially for organizations dealing with sensitive information like healthcare companies, Parli said.



<h2 class="wp-block-heading" id="h-crucial-quote">CRUCIAL QUOTE</h2>



“The benchmarks aren’t perfect, but as of right now, that’s the primary mechanism we have to evaluate the models,” Parli told Forbes, cautioning that “researchers can somewhat easily game the system” today, with AI models quickly saturating benchmarks. “I think we need to get creative in the development of new ways to evaluate AI models,” Parli said. “



<h2 class="wp-block-heading" id="h-what-we-don-t-know">WHAT WE DON’T KNOW</h2>



Measuring intelligence is tricky when we do not know what it is we are supposed to be measuring. There is no universally accepted definition of intelligence in humans, let alone a way to measure it, and the possibility, nature and scope of animal intelligence has divided scientists for centuries. While AI benchmarks have typically focused on the ability to perform a particular task, more general assessments will be required in the near future as researchers make progress towards their goal of creating artificial general intelligence (AGI). AGI is capable of excelling and possibly matching humans across a broad set of domains rather than at just one task such as walking, moving boxes, identifying tumors on scans and playing chess.



<h2 class="wp-block-heading" id="h-how-useful-is-chatbot-arena-for-evaluating-ai-models">HOW USEFUL IS CHATBOT ARENA FOR EVALUATING AI MODELS?</h2>



“The rankings [Chatbot Arena] gives are something that I trust more than most other rankings,” Dodge told Forbes, “because it uses a real human to say whether they prefer one generation over another.” Parli suggested assessments like Chatbot Arena could “implicitly evaluate factors” we want in our AI but are less quantifiable than something like coding ability. But she did stress that something like Chatbot Arena should not be the only evaluation method used, saying there are “many factors that should be important to organizations when evaluating models” and it “does not cover all of them.”



.

Einstein And Musk Chatbots Are Driving Millions To Download Character.AI

What AI Is The Best? Chatbot Arena Relies On Millions Of Human Votes

The rapid advancement and implementation of artifi cial intelligence (AI) technologies have signifi cant economic, societal, and ethical consequences for Africa. Efficient governance is crucial to optimizing AI’s advantages while minimizing its risks. This article offers policymakers essential governance issues, suggests strategic methodologies for the appropriate oversight of AI technologies, and presents a framework for AI governance and contextualizes these within the setting of Africa.



Good AI governance must be based on good values. One such value is AI transparency which is essential for AI governance, as it fosters trust, accountability, and fairness. Transparent AI systems allow stakeholders to understand and evaluate decision-making processes, identify and correct biases and assure ethical and legal compliance. The other value is AI truth. Truthful AI prevents misinformation and promotes critical thinking in public conversation, strengthening AI systems’ legitimacy and building social trust needed for widespread adoption and benefi cial integration of AI technology into daily life. The other values are safety and security which are vital for AI systems, as AI interacts with people in different ways, building trust and adoption.



Prioritizing safety reduces accidents and errors in AI-driven industries, safeguarding lives, and money. The other value is ethics. Ethical design and deployment of AI technologies benefit society and minimize harm, including promoting justice, transparency, accountability, and human rights. The other AI value is privacy. Privacy is crucial for AI systems to secure personal data and retain trust in technology. By using strict privacy rules, AI systems protect personal data from unlawful access and exploitation.The AI governance framework proposed in this article entails human behavior, mechanisms for incentives and disincentives, institutional governance structures, policies and regulations, standards, and the law. AI governance involves adhering to AI values, using behavioral science to cultivate a culture that maximizes AI’s potential while minimizing associated risks, and implementing mechanisms to foster a culture that maximizes the potential benefi ts while limiting its associated risks. 



Establishing an institutional structure combining the best practices derived from the Intergovernmental Panel on Climate Change (IPCC) and the International Atomic Energy Agency (IAEA) is recommended to regulate AI effectively. This entity would set global AI standards, monitor compliance, foster international cooperation, address ethical guidelines, establish safety protocols, provide a platform for sharing best practices and promote transparency.Establishing policies and regulations to control AI involves crafting legal frameworks that ensure ethical AI use, focusing on transparency, accountability, and fairness while protecting privacy and preventing discrimination. Key measures could include mandatory AI audits and strict data protection requirements. 



Governments might also create dedicated bodies to oversee AI practices and enforce these regulations, ensuring AI benefi ts society and adheres to ethical standards.Establishing AI standards involves creating uniform guidelines for AI design, development, and deployment, focusing on technical quality, ethical considerations, and compatibility. These standards should cover data privacy, algorithmic transparency, security, and bias prevention. Lawmakers should collaborate with experts and the public to craft adaptable, informed regulations.



This model proposed above governs the following aspects of AI: data, algorithms, computing, and AI applications. Data governance ensures ethical, accurate, and secure use of AI systems, while algorithmic governance ensures ethical behavior and impartial results. Computing governance ensures safe, responsible, ethical, and sustainable AI technology, as computer technologies permeate every part of life. Well-defi ned governance frameworks build public trust, promote more egalitarian technology access, and incentivize good uses of AI. This governance should include specifi c regulations for AI applications in critical sectors like healthcare and fi nance.



AI governance in Africa is crucial to prevent AI from exacerbating existing inequalities that divide the Global South from the Global North and creating new ones. There is a need for robust governance frameworks to ensure that AI systems are used responsibly, minimizing risks, and empowering Africans to benefi t from AI’s potential especially given the fact that much of the AI technology is developed in the Global North. 



Issues such as the potential for AI to displace workers in key sectors in Africa, such as manufacturing and routine service industries, must be dealt with carefully, as AI might eliminate any hope of Africa industrializing. There is also a need to mitigate the negative infl uence of AI on political processes, such as deepfakes and misinformation campaigns. 



African nations must develop national AI strategies, establish independent regulatory bodies, invest in AI education and capacity-building programs, and play an active role in shaping global AI governance discussions through mechanisms such as the African Union and the United Nations. In conclusion, AI governance frameworks must be dynamic and adaptable to technological advancements. Policymakers should collaborate with technologists, businesses, academia, and civil society to create comprehensive governance strategies that ensure AI serves the public good.

AI, Machine learning, robot hand ai artificial intelligence assistance human touching on big data network connection background, Science artificial intelligence technology, innovation and futuristic.

Expert-Speak: An Effective Framework For Governance Of AI In Africa

Artificial intelligence (AI) has made significant strides in all areas of our lives. AI can now monitor patients, safeguard buildings, improve manufacturing processes and write poetry. These are good uses of AI. There are, however, bad uses and characteristics of AI.



For example, AI can discriminate against digital minorities such as people of African descent or ethnic minorities. It can be weaponized with devastating consequences for peace and security. Given that AI is both good and bad, what should be done? We must ensure that AI is used to maximize the good and minimize bad uses. Various strategies have been deployed to maximize the good uses of AI. Furthermore, multiple techniques have been deployed to reduce the bad uses of AI. For example, ensuring that data for training AI systems is representative can minimize algorithmic bias and discrimination.



However, fixing defective algorithms is much easier than improving a flawed human being who intends to use AI negatively. For example, it is far more challenging to change a human being that discriminates against ethnic minorities than to fix an AI machine that discriminates against ethnic minorities. To improve the AI machine, we must apply the best practices of handling data, designing algorithms, and using AI technology, which requires good human behavior.



In AI, discussions of its development, deployment and regulation frequently focus on technological and ethical concerns. However, this discussion’s significant and underexplored aspect parallels human behavior and AI governance. These two generate a complex confluence that considerably impacts how AI technologies are conceived and deployed and how they are regulated. As AI permeates every aspect of our lives, understanding and leveraging this confluence of AI and human behavior is advantageous and is also required for developing AI systems that are both effective and fair. 



Humans are at the heart of AI development, and their decisions, prejudices, and actions affect AI systems.Behavioral science, which studies human behavior   via    psychological, cognitive, and emotional lenses, provides essential insights into how developers and designers should approach the building of AI. It highlights the cognitive biases that can influence algorithm design and the ethical blind spots resulting from prioritizing technical efficiency over social consequences.



Incorporating human behavior into AI governance requires identifying and addressing these human issues. It entails developing frameworks enabling developers to reflect on the potential biases they bring to the design process and cultivating a culture that values ethical considerations and societal wellbeing over technical innovation.



End-user behavior must also be considered when governing AI. Behavioral science provides a lens through which to investigate how people engage with AI systems, such as responding to recommendations, making decisions based on AI-generated data, and building trust in AI technologies. These findings are critical for developing AI systems that are not only user- friendly but also promote beneficial behavioral outcomes, such as improving decision-making processes and preventing the reinforcement of negative biases.



Furthermore, behavioral science can inform public education campaigns on AI, enabling users to be more critical and educated in their interactions with AI systems. Governance frameworks based on behavioral science can raise an understanding of AI’s possible biases and limits, allowing people to interact ethically and effectively with AI.



The link between human behavior and AI governance is dynamic. As AI technologies advance, so will the societal norms and behaviors that shape and are shaped by them. Governance frameworks must be adaptive and ready to respond to new insights from behavioral science into human-AI interaction and the societal repercussions of AI breakthroughs.



This adaptability necessitates a participatory and inclusive governance strategy involving AI developers, users, behavioral scientists, ethicists, policy-makers, and other stakeholders in ongoing discourse. This strategy can help to keep governance frameworks current and thriving in the face of rapid technological development and shifting human behaviors.



Education is critical to guaranteeing the ethical governance of AI since it provides developers and consumers with the knowledge and skills needed to navigate the complicated moral environment of AI technology. Technologists can gain a thorough awareness of the ethical consequences of their work by participating in comprehensive educational programs that integrate ethics, data privacy, and societal effects into the core curriculum of computer science and AI courses. 



Education can help people understand AI technologies, promoting informed and critical engagement with AI systems. By incorporating ethical considerations into all stakeholders’ education in creating and using AI, we can foster a culture of responsibility and accountability. This strategy teaches individuals to foresee and address ethical difficulties and promotes the advancement of AI systems that emphasize human values and social wellbeing. As we set the road for the future of AI, let us not underestimate the value of understanding human behavior in guiding us toward more responsible and beneficial AI governance

AI (Artificial Intelligence) concept. Electronic circuit. Communication network.

The Dynamic Link Between Human Behavior And AI Governance

<h2 class="wp-block-heading" id="h-topline">TOPLINE</h2>



Apple’s new line of&nbsp;<a href="https://www.apple.com/apple-events/" rel="noreferrer noopener" target="_blank">iPad Pros</a>&nbsp;will include a processor the company claims is an “outrageously powerful chip for AI,” as the tech giant laid the groundwork for long-anticipated generative AI features across its products amid&nbsp;<a href="https://www.washingtonpost.com/technology/2024/05/02/apple-sales-ai-iphone-earnings-regulations/" rel="noreferrer noopener" target="_blank">investor concern</a>&nbsp;that it’s falling behind in the race to harness the new technology.



<h2 class="wp-block-heading" id="h-key-facts">KEY FACTS</h2>



The new M4 chip represents an upgrade over the M2 chips that powered the last generation of iPads and the M3 chips that power Apple’s current generation of MacBook laptops, the company said in an&nbsp;<a href="https://www.apple.com/apple-events/" rel="noreferrer noopener" target="_blank">announcement</a>&nbsp;Tuesday.



Tuesday’s announcement didn’t offer specifics about potential AI features that could be built into Apple tech, though Tim Millet, Apple’s vice president of platform architecture, gave one example, claiming that users would be able isolate the subject of a 4K video and remove the background with just a tap in Apple’s Final Cut Pro app.



Millet said the neural engine powering the AI capabilities in the M4 chip has been included in Apple chips for years, but this latest neural engine is 60 times more powerful than Apple’s earliest models and “more powerful than any neural processing unit in any AI PC today.”



On Tuesday, Apple announced a new line of&nbsp;<a href="https://www.apple.com/ipad-pro/" rel="noreferrer noopener" target="_blank">iPad Pros</a>&nbsp;that the tech giant called its thinnest yet, coming in 11-inch and 13-inch models with OLED screen technology starting at $999 for the 11-inch model.



It also announced its new&nbsp;<a href="https://www.apple.com/ipad-air/" rel="noreferrer noopener" target="_blank">iPad Air</a>, running on the M2 chip,which comes in a new, larger 13-inch option for $799—both the iPad Air and the iPad pro are available to order online Tuesday and in stores May 15.



Apple&nbsp;<a href="https://finance.yahoo.com/quote/AAPL?.tsrc=fin-srch" rel="noreferrer noopener" target="_blank">stock</a>&nbsp;was slightly up—less than a percentage point—following Tuesday’s announcement.



<h2 class="wp-block-heading" id="h-what-to-watch-for">WHAT TO WATCH FOR</h2>



Apple could unveil more of its plans for AI at its upcoming&nbsp;<a href="https://www.apple.com/newsroom/2024/03/apples-worldwide-developers-conference-returns-june-10-2024/" rel="noreferrer noopener" target="_blank">Worldwide Developers Conference</a>&nbsp;on June 10. CEO Tim Cook said Tuesday the company intends to “talk about the future of our platforms and share some exciting details about what’s to come” at the event. Cook has&nbsp;<a href="https://www.cnbc.com/2024/02/28/apple-annual-shareholder-meeting-2024-ai-investments.html" rel="noreferrer noopener" target="_blank">previously said</a>&nbsp;the company is “investing significantly” in AI and would have more announcements on the subject “later this year.”



<h2 class="wp-block-heading" id="h-contra">CONTRA</h2>



Apple’s plans for artificial intelligence could face immediate competition. The Verge reports Microsoft is&nbsp;<a href="https://www.theverge.com/2024/4/8/24116587/microsoft-macbook-air-surface-arm-qualcomm-snapdragon-x-elite" rel="noreferrer noopener" target="_blank">set to host</a>&nbsp;an event in Seattle this month where it plans to unveil its vision for the future of “AI PCs” and could launch new hardware featuring its Arm AI chips.



<h2 class="wp-block-heading" id="h-key-background">KEY BACKGROUND</h2>



Sales of Apple’s flagship iPhone have been slumping, down 10% year-over-year in the first three months of 2024, driving a 4% decrease in revenue last quarter. And while Apple is considered the most profitable American public company, its stock has been&nbsp;<a href="https://www.forbes.com/sites/dereksaul/2024/05/02/apple-earnings-come-in-hot-as-china-blow-less-painful-than-feared/?sh=5f71ff956903">trending downward</a>&nbsp;of late due to lower iPhone demand in China and concern over Apple’s lack of major plays in the field of artificial intelligence.



<h2 class="wp-block-heading" id="h-big-number">BIG NUMBER</h2>



38 trillion. That’s how many operations the new M4 chip’s neural engine can perform per second, Apple claimed on Tuesday.<a href="https://www.forbes.com/sites/jamesfarrell/"></a>

American multinational technology company Apple store seen

Apple Promises AI Features In New IPad Pro Chips—As AI Tech Race Heats Up

It’s an auditory crisis. The perfect mimicry of the human voice by artificial intelligence (AI) used by criminals to access organizations, discredit governments and leaders, steal data and money, and exploit people and companies. A woman was tricked into thinking her daughter had been kidnapped – the terrified voice on the end of the call was an AI fake. An employee in a multinational enterprise was conned into paying cybercriminals $25 million because he believed he was talking to the company’s Chief Financial Officer and Tom Hanks’ voice was cloned to sell dental plans.



If a fraudster can mimic someone’s voice, they can bypass security checks designed to use voice biometrics. They can fool people into believing that they’re talking to a trusted friend and colleague. They can slide past security barriers and connect with people because their fakes are so good, nobody can tell the difference.



“It’s the trust factor,” explains Adrianus Warmenhoven, cybersecurity expert at NordVPN. “As humans we have implicit trust anchors – I can read your body language when I see you, I can hear inflections and reactions in your voice on the phone and these anchor trust between us. These were developed by being in school, going to work, seeing people. Now they’re being used by criminals to make connections and successful scams.”



Voice is easy to start with. It can be trained easily and there are plenty of applications online. Anyone can download an AI and clone someone’s voice. The technology capable of converting voice to text, text to voice and then applying it to anything from fake bank calls through to scams designed to phish passwords or con family members is readily available.



“How accurately can you pick up someone’s voice? Are the slight tonal differences not just background noise? What if they’ve told you they’re not feeling well? The last two are the highest contributors, accounting for more than 80% of the threats we are seeing in terms of how voice recognition can be compromised,” says Vukosi Sambo, Executive – Head of Data Insights and AIat AfroCentric. 



“We are also seeing instances where someone records a person’s voice and then uses it to authenticate themselves as that person to gain access to information or systems.” 



Getting hold of a person’s voice isn’t difficult either. Cybercriminals simply turn to the one thing they’ve used to hone their craft for years – social media. As Ryan Mer, CEO of eftsure Africa, points out: “People post videos on social media, they share voice notes on WhatsApp, and in the business space, key people have some form of media, whether it’s an interview on TV, radio or a podcast.”



The technology is capable of generating a legitimate voice profile using only a few seconds of these videos and voice snippets. Then, that voice can be used to get in touch with people in the business, especially the finance team, to swindle them out of sensitive information or get them to perform certain tasks. It’s not always a case of getting someone to transfer millions, it can also be trying to access the system for data that can be sold for millions, or to infiltrate a business to commit long-term fraud from within.



Anyone can go onto the dark web and, for a small fee, ask for a deep fake video saying whatever they want. Nobody is going to ask how this will be used. Nobody will care about the repercussions of its use. It is also very likely the next creator of millionaires and billionaires who have used AI to achieve their financial goals, as nefarious as those may be.



Sambo adds: “It’s important to understand what the organization is entitled to in terms of collecting and storing voice data. If you’re not careful, you could infringe individual rights in light of the AI threat. Can companies that record calls for quality and security purposes really protect themselves from this perspective?”



“Considering how this technology is available to anybody these days for next to nothing, security has to take a multi- layered approach,” says Lance Fanaroff, co-founder of iiDENTIFii. “You need to manage access control based on who has access to what; focus on individual access tools such as two-factor authentication and biometrics; and then invest into technologies designed to identify and mitigate potential threats on a granular, individual, level.”



There is a positive side to the conversation – the technologies designed to tackle the threats are evolving just as rapidly. They’re more resilient and alert, capable of adapting to the potential deep fake risks with agility. However, there is only so much a company can do to protect its people and its systems. As Kabelo Makwane, Managing Executive: Cloud, Hosting &amp; Security at Vodacom Business Africa, concludes: “What has become important now is finding the right balance between leveraging the technology, not locking down systems, and still finding a way of layering security so there remains space for growth and productivity.”

Asian woman using smartphone surrounded by beams of light. Metaverse and Artificial Intelligence

AI Has Stolen Your Voice… And Why It’s An Auditory Crisis

Five years ago, I couldn’t imagine asking boards and business leaders whether robots would one day run their companies.



However, the massive increase in awareness of the potential of large language models is forcing executives to explore how to use robots and artificial intelligence in day-to-day decision-making.



In fact, the rise of executive robots has already started.



For example, in May 2023, Forbes reported that Dictador, a progressive premium brand rum distiller, had appointed an AI CEO called Mika.



It is gimmicky for sure, as humans are still very much controlling her AI strings, but as Mika reportedly said in an interview with Reuters, “My decision-making process relies on extensive data analysis and aligning with the company’s strategic objectives.” She went on to explain how she was devoid of personal bias, which helps ensure that strategic choices are prioritized in line with the organization’s best interests, not hers.



This sounds amazing. So, is Mika the forerunner of the mass adoption of AI for corporate governance and leadership roles?



For once, the likely answer is yes, absolutely.



For starters, the demand for a solution like this is apparent.



Humans regularly make bad decisions, and more often than not, this leads to someone losing a lot of money.



Why, for example, did OpenAI’s board sack Sam Altmann (before he returned as CEO in November last year)?



What led Elon Musk to rebrand Twitter to X?



And why didn’t Silicon Valley Bank diversify its risk away from startups?



Decisions like these wipe out trillions of dollars of value every single year. And these three examples are just the tip of the iceberg.



Further, those impacted aren’t just the super-wealthy. Often, it’s the poorest who see their life savings and pensions hit hardest.



All of which means there is a huge interest from everyone, from governments to investors, to find ways to improve decision- making and corporate governance.



We already spend billions of dollars for consultants and business schools to develop leaders and models to help companies maximize profits while minimizing risks.



But these billions will quickly switch to AI alternatives if there is even the smallest of probabilities that technology will help improve a stock’s potential returns.



So, the demand is there.



But how long will we have to wait for the likes of the already hugely-impressive ChatGPT to offer intelligence that can not only answer questions in narrow fields but actually analyze data from multiple sources, draw inferences, consider interdependencies, develop execution plans, draw up funding requirements, consider resource allocations, raise funding and build support along with all the other tasks and activities that are required in any modern-day business?



Mustafa Suleyman, founder of DeepMind and Inflection.AI, explores this question in his recent bestseller, The Coming Wave.



Specifically, he asks how long it will be before machines answer and act on prompts such as “Please design and execute a strategy that will allow me to make $1mill on Amazon in as few months as possible with just $100K of investment.”



His conclusion? “Aside from the legal requirements of registering as a business on the marketplace and getting a bank account, all of this seems to me eminently doable. I think it will be done within the next year and probably fully autonomously within three to five.”



When an industry insider as respected as Suleyman makes a comment like this, we should sit up and listen.



And it makes absolute sense when you consider that a potent combination of factors is driving development in AI. Breakthroughs in machine learning algorithms, dramatic increases in computing power, the abundance of extensive data sets, advancements in cloud computing, enhanced global connectivity, escalating demand across industries and the democratization of access through open-source initiatives are just some of the factors propelling it at exponential rates.



And, of course, the biggest factor of all is perhaps the simple fact that we are now using AI to build AI, as more and more code is being written by machines, not developers. So perhaps the real question isn’t whether or even when robots will run companies. Shareholders and boards might do better considering what types of leaders they need to employ to manage the transition from human to augmented to algorithmic leadership. This is not an easy question to answer. There is no template for designing leadership systems for robots. Perhaps ChatGPT has the answer?

Tag: artificial intelligence

FORBES AFRICA SOCIAL MEDIA