The AI Acceleration Gap

0:00Today on the AI Daily Brief, we are talking about the AI acceleration gap, what it is, why it matters, and what you should do about it. Before that, in the headlines, what we learned from a recent town hall at OpenAI. The AI Daily Brief is a daily podcast and video about the most important news and discussions in AI.

0:22All right, friends, quick announcements before we dive in. First of all, thank you to today's sponsors, Robots and Pencils, Optimizely, Zen Coder, and Super Intelligent. To get an ad-free version of the show, go to patreon.com slash ai daily brief, or you can subscribe on Apple Podcasts. And if you are interested in learning about sponsoring the show, send us a note at sponsors at AIDailyBrief.ai. Welcome back to the AIDailyBrief Headlines Edition, all the daily AI news you need in around five minutes. Over the weekend, Sam Altman announced that on Monday afternoons last evening, they would be hosting what they were calling a town hall for AI builders at OpenAI.

0:58In his announcement post, Sam said that this was an experiment and a first pass at a new format. He framed the live stream event as an opportunity to gather feedback as OpenAI begins building their next generation of tools. Ultimately, it's sort of played out as a Q &A about the state of the company and the industry. One of the big points of discussion was the performance of GPT-5 too. Altman acknowledged, for example, that the latest model has a writing style that can be unwieldy and difficult to read. He said, I think we just screwed that up. We will make future versions of GPT-5.x hopefully much better at writing than 4.5 was.

1:29Now, Altman noted that their focus hadn't been on writing, saying, we did decide, and I think for good reason, to put most of our effort in 5.2 into making it super good at intelligence, reasoning, coding, engineering, that kind of thing. And we have limited bandwidth here, and sometimes we focus on one thing and neglect another. Now, of course, rumors suggest that the next model, Codenamed Garlic, is weeks or even days away. So for those of you who find GPT-52's writing clunky, you presumably won't have to deal with it much longer. Altman also discussed a hiring slowdown at OpenAI. Responding to a question about how AI had changed the interview process, he commented, we are planning to dramatically slow down how quickly we grow because we think we'll be able to do so much more with fewer people.

2:08He assured the crowd that this was not a hiring freeze and that headcount reductions are not on the table, but did suggest that AI developments could rapidly shift staffing needs over the short term. What I think we shouldn't do and what I hope other companies won't do either is hire super aggressively then realize all of a sudden AI can do a lot of stuff and you need fewer people and have to have some sort of very uncomfortable conversation. I think the right approach for us will be to hire more slowly, but keep hiring. In other comments, Altman said that he expects the cost of AI to continue to hyper-deflate, forecasting that OpenAI will be able to deliver, quote, GPT-5.2 level intelligence by the end of 2027 for at least 100 times less.

2:44Reflecting something that we've talked about a bunch on this show, Altman said that another big goal of 2026 is to push, quote, super hard on memory and personalization. Altman said that he's personally ready to give ChatGPT complete access to his computer and internet history, allowing it to, quote, just know everything. And while he acknowledged that security and privacy were still major concerns, he said that the company will focus on building a system that has, quote, such a deep understanding of the complex rules and interactions of my life that it knows what to use when and what to expose where.

3:10As part of that goal, Altman explained that login with ChatGPT is coming soon, which will in the short term enable token budgets to be shared across various apps, with the long-term vision being allowing portable memory to function across different AI products. We even got some little hints about their hardware plans, with Altman saying that the vision is now a collaborative multiplayer experience. He framed it as five people gathered around a table with what he called a little robot to help the group do better. And highlighting the massive shift that we have talked about extensively and is in fact the theme of the main episode today, Altman commented that, quote, what it means to be an engineer is going to super change.

3:43He said there will probably be far more people creating far more value getting computers to do what they want. He noted at the same time, however, that demand for software seems not to be slowing down at all. So what to think about this overall? There were quite a few snarky responses to this on Twitter slash X. Some people didn't like that comments were off on the live stream. Others thought that the vibes and the atmosphere were just really weird. Some thought that Altman himself looked kind of tired and run down, which others interpreted as OpenAI's competitive struggles, but which also could easily be explained by being father to an infant.

4:12Overall, what I would say is this. I think it kind of doesn't matter if their A wasn't all that much revealed on this, and B, people have critiques about the vibes. I think that if OpenAI regularizes this, it could actually be really valuable and build a lot of trust. Having a predictable, regular place to have these sort of conversations could go a long way to making things that need to be explained not always feel like they're a PR response. So in that regard, I think it was successful and they should do more of it. Now, another discussion from this weekend around OpenAI had to do with their advertising plans.

4:43The information reports that pricing sheets are starting to circulate with a premium price tag for OpenAI ads. OpenAI appears to be selling on a CPM basis, with an offering at$60 CPMs or$60 per thousand views, which is around three times the cost of placing an average ad on, for example, Meta's platforms. The only data available during the early stage will be total ad views and clicks, which is obviously a lot less information than they're going to get from other advertisers, but presumably that will change soon. OpenAI does pledge not to sell personal data to advertisers, and they appear to be taking that stance to the extremes, and the premium pricing, at least at the beginning, probably won't be a deterrent to the early advertisers that are clamoring to get on the platform.

5:20Studies have generally shown that AI users have high intent relative to other types of internet users, which could end up easily justifying that premium over other digital ad units. Now, advertising isn't the only place where OpenAI plans to charge a hefty premium. Fintech reporter Simon Taylor recently noted that Shopify merchants are being charged a 4 % fee for sales conducted through ChatGPT, which is a fee on top of existing Shopify charges. Shopify CEO Toby Lutke filled in some further details, commenting, This is ChatGPT charging 4 % and we collect the fees on their behalf. Everyone gets a free trial that starts after the first sales.

5:52Not saying that's good or bad. Ads definitely cost more for most. Taylor acknowledged that this is pretty close to fees from buy now, pay later services and added, for what it's worth, I think 4 % is very defensible if conversion is there. Moving over to chips today, Microsoft has unveiled the second generation of their in-house AI chip, taking aim at custom silicon from Google and Amazon. Called the Maya 200, Microsoft claims the chip is the quote most performant first-party silicon from any hyperscaler. The chip is optimized for inference, and Microsoft says it's the most efficient silicon in their fleet, outperforming the next best by 30 % on a performance-per-dollar basis.

6:26The accelerator was built using TSMC's latest 3-nanometer process. Google is also using the 3-nanometer process for their 7th generation TPUs, but NVIDIA chose to stick with 4-nanometer manufacturing on their latest generation Blackwell chips. The chip also features enough memory to easily run the latest models with plenty of headroom for the next generation. Now, whenever a new chip is released, the immediate chatter is all about whether this will end NVIDIA's dominance of the industry, but Tom's hardware pointed out that comparisons to NVIDIA's leading chips are a little spurious as they do very different things.

6:57Microsoft's hardware won't be available for outside sales, so the Maya 200 will only be able to move the needle internally, and Tom's also pointed out that the Blackwell 300 Ultra flagship vastly outperforms on raw power and of course integrates with NVIDIA's highly developed software stack, but the Maya 200 does beat the Blackwell chip in efficiency, operating at nearly half the total power draw. Ultimately, with the Maya 200, though, Microsoft is staking their claim as a player in the custom silicon race. Staying in chip land but moving over a bit, NVIDIA has invested a further$2 billion into CoreWeave to kickstart the deployment of AI factories.

7:29Now, NVIDIA CEO Jensen Huang has been discussing the concept of AI factories for the past year. The language describes the idea that data centers will need to be deployed on a much greater scale to supply the AI tokens that will drive economic outcomes in the future. Essentially, it reframes data centers from large cloud storage and compute providers to the producers of the core commodity of the AI age. With the next leg of their CoreWeave partnership, NVIDIA will support a scaling up of CoreWeave's infrastructure. The goal is to deploy 5 gigawatts in capacity by 2030, with NVIDIA using its financial might to help procure land and power for the rollout.

7:59NVIDIA already owned a 6.6 stake in CoreWeave, so this new investment brings their ownership to around 10%. Now, one last note before we move over to the main episode. On Monday, Anthropic CEO Dario Amadei released a new 21 ,000-word essay called The Adolescence of Technology. In some ways, it's a more critical and concerned complement to his Machines of Loving Grace from a couple of years ago. Originally, I had planned on focusing on that. Given how dense and deep this thing is, people are still just wrapping their heads around it. And so rather than dive all the way into it today, I wanted to give it a couple more days for takes and reactions to marinate.

8:30We will discuss it at some point this week. But in today's main, we are instead going to talk about something which I am calling the AI acceleration gap.

8:43Today's episode is brought to you by Robots and Pencils, a company that is growing fast. Their work as a high-growth AWS and Databricks partner means that they're looking for elite talent ready to create real impact at velocity. Their teams are made up of AI-native engineers, strategists, and designers who love solving hard problems and pushing how AI shows up in real products. They move quickly using RoboWorks, their agentic acceleration platform, so teams can deliver meaningful outcomes in weeks, not months. They don't build big teams. They build high-impact, nimble ones. The people there are wicked smart with patents, published research, and work that's helped shape entire categories.

9:20They work in velocity pods and studios that stay focused and move with intent. If you're ready for career-defining work with peers who challenge you and have your back, Robots and Pencils is the place. Explore open roles at robotsandpencils.com slash careers. That's robotsandpencils.com slash careers. Most marketing teams aren't short on ideas, but what they are short on is time. And that's exactly what Optimizely Opal gives you back, with AI agents that handle real marketing workflows. You know, like creating content and checking compliance, generating experiment variations, personalizing user experiences, analyzing pages for GEO, even tasks like approvals and reporting.

9:58It's your AI agent orchestration platform for marketing and digital teams, plugging seamlessly into the tools you already use, handling the boring busy work, and keeping everything on brand. That leaves you marketers with more time to do your actual job. See what Opal can automate for your team by signing up for a free enterprise agentic AI workshop with Optimizely. Learn more at Optimizely.com slash the AI Daily Brief. If you're using AI to code, ask yourself, are you building software or are you just playing prompt roulette? We know that unstructured prompting works at first, but eventually it leads to AI slop and technical debt.

10:32Enter Zenflow. Zenflow takes you from vibe coding to AI-first engineering. It's the first AI orchestration layer that brings discipline to the chaos. It transforms freeform prompting into spec-driven workflows and multi-agent verification, where agents actually cross-check each other to prevent drift. You can even command a fleet of parallel agents to implement features and fix bugs simultaneously. We've seen teams accelerate delivery 2x to 10x. Stop gambling with prompts. Start orchestrating your AI. Turn raw speed into reliable production-grade output at zenflow.free. Today's episode is brought to you by Superintelligent.

11:09Superintelligent is a platform that, very simply put, is all about helping your company figure out how to use AI better. We deploy voice agents to interview people across your company, combine that with proprietary intelligence about what's working for other companies, and give you a set of recommendations around use cases, change management initiatives that add up to an AI roadmap that can help you get value out of AI for your company. But now we want to empower the folks inside your team who are responsible for that transformation with an even more direct platform. Our forthcoming AI Strategy Compass tool is ready to start to be tested.

11:39This is a power tool for anyone who is responsible for AI adoption or AI transformation inside their companies. It's going to allow you to do a lot of the things that we do at Superintelligent, but in a much more automated, self-managed way and with a totally different cost structure. If you are interested in checking it out, go to aidailybrief.ai slash compass, fill out the form, and we will be in touch soon.

12:02Welcome back to the AI Daily Brief. Today we are talking about the AI acceleration gap. This is a new phenomenon that I've been thinking about a lot lately, and which I think has some pretty significant consequences for both individuals and companies, but which I also think is at risk of being subsumed into broader AI conversations in a way that isn't all that helpful. And I want to talk first about the acceleration side of the acceleration gap. Perhaps the biggest thing that we've been tracking here at AIDB in January is this sense and realization among many of the most enfranchised users of AI that something fairly meaningful has shifted, that some inflection point has been reached recently, which really has changed what we can do.

12:41You started to see this around the holidays, with a great example of it being this viral tweet from OpenAI co-founder Andre Karpathy who wrote, I've never felt this much behind as a programmer. The profession is being dramatically refactored as the bits contributed by the programmer are increasingly sparse in between. I have a sense I could be 10x more powerful if I just properly string together what has become available over the last year. And a failure to claim the boost feels decidedly like a skill issue. Clearly some powerful alien tool was handed around except it comes with no manual, and everyone has to figure out how to hold it and operate it.

13:11while the resulting magnitude 9 earthquake is rocking the profession. Roll up your sleeves to not fall behind. And so obviously why this was so resonant is one, that many people were feeling like this, but two, the source of it. This is someone who can claim to be in the very top list of people who have actually built this technology, not just use it, and they are saying they are feeling behind relative to what's possible. However, there was a positive side of this as well. On January 3rd, MidJourney founder David Holes wrote, I've done more personal coding projects over Christmas break than I have in the last 10 years.

13:43It's crazy. I can sense the limitations, but I know nothing is going to be the same anymore. Now, we have talked extensively about the combination of models, Opus 4.5, 5.2 Codex, as well as the harnesses like Cloud Code in which they operate. What's more, since the beginning of the month, there has been a continuous set of additional updates. For example, Cloud Cowork and most recently, Cloud Bot, which everyone has been talking about, including me on yesterday's episode, that just continue to extend this discourse of acceleration. Over the weekend, New York Times columnist Kevin Roos wrote about the increasing chasm of experience and impression of AI between the most enfranchised users and everyone else.

14:19He wrote, I follow AI adoption pretty closely, and I have never seen such a yawning inside-outside gap. People in San Francisco are putting multi-agent clawed swarms in charge of their lives, consulting chatbots before every decision, wireheading to a degree only sci-fi writers dared to imagine. People elsewhere are still trying to get approval to use Copilot and Teams, if they're using AI at all. It's possible the early adopter bubble I'm in has always been this intense, but there seems to be a cultural takeoff happening in addition to the technical one. Not ideal. Adding a little bit more, Kevin continues, I want to believe that everyone can learn this stuff.

14:56But in the same way that the AI companies that took scaling seriously started stockpiling GPUs, etc. before 2022 had a virtually insurmountable head start over latecomers, it's possible that restrictive IT policies have created a generation of knowledge workers who will never catch up. So Kevin here is talking about a natural outcome of this acceleration experience that we've been discussing in a way that is, of course, concerned, that sees this understanding gap as a problem. Many people chimed in to say that this also resonated with their experience. AEI fellow John Bailey wrote, this captures it exactly.

15:30In late December, my feed was full of people using Claude Code and declaring AGI. The same week, a DC consulting exec told me AI was mostly hype because of hallucinations. And then at a holiday party, most of my mom's friends still hadn't tried ChatGPT. It felt like living in three different realities. Now what's fascinating and builds on this idea of living in different realities is that the responses to Kevin's post were basically a Rorschach test for how people feel about AI, almost as a social or political issue, not just as some new technology category. Lots of people were determined to imply that AI was NFT's 2.0.

16:06Link and Michelle shared an old post from the Rare Candy NFT marketplace that said, a lot of y 'all still don't get it. Ape holders can use multiple Slurp juices on a single ape. So if you have one Astro Ape and three Slurp juices, you can create three new apes. Tonight's Slurp Juice mint event is essentially a minting event for both lab monkeys and special forces. Point being, of course, that this sounds absurd, and all this hype around AI will sound just as absurd a few years from now. Vinny Truvati did what many people did, which is throw old Kevin Roos posts and articles in his face, like this one from 2021 where he wrote about pudgy penguins, in a piece titled I joined a penguin NFT club because apparently that's what we do now.

16:39Some were even angrier. John Rappetti paraphrased Kevin's post as this, All the evil morons who can't do anything are using AI for everything. Normal people aren't. How can we explain this? Dr. Andrew Naber summed up the people who insist that all of this is ineffective bluster. He writes, new tech does not exist outside of culture. This sounds the same as any other hustle culture optimizing life hack grift. These fussy little bits of AI software give dopamine hits with marginal or negative impact on productivity, lifestyle, or mental health. Now, it should be noted that it is not just the AI haters who have critiques to be levied, even if they are not being pointed so personally at Kevin himself.

17:14Notebook LM co-creator Reza Martin wrote,

17:44time for early adopters, but really grading to anyone else. In other words, Reza is saying, in addition to us AI people being kind of annoying with how we talk about all this, the products themselves aren't all that great to use, especially when compared to the capabilities. Others pointed out that identifying this as a San Francisco thing is probably incorrect. Professor Ethan Malik wrote, this isn't just a San Francisco thing. There are people in a range of professions who found absolutely breathtaking uses of current capabilities, like using agent swarms to do real work in crazy ways, but they are often more isolated because of a lack of unifying community.

18:16Kevin Warbach said this is real and notable, framing it as SF versus the world is misleading. The real question is whether companies, which tend to be risk-averse, will lose out as startups and individuals capitalize on the new capabilities of agentic AI. MIT Sloan's Matt Bean wrote, the gap is huge, consequential, and growing. Many of the consequences are wonderful, but generally this gap is unnecessary, driven by privilege, and, if prior science is any guide, will blunt the gains from the tech and concentrate power even further. Summing it all up, Dean Ball wrote, The gap between the early adopters and everyone else, both in terms of their AI use but also in their ways of thinking, has never been wider and appears to be widening at an accelerating rate.

18:54Even most of my followers clearly don't get it. Slightly worrisome. So I think what all these folks are identifying is actually a real thing. And for the sake of having a simple, memorable name, I'm calling it the acceleration gap or the AI acceleration gap. I recently shared this slide in a corporate presentation, and basically I argued that for much of the last few years of AI, while there were certainly some groups that had a real capability advantage versus everyone else, the gap between the early adopters and the other types of users was fairly consistent. In other words, there was some correlation in the rate of progress between all the different categories of users.

19:30Recently, however, it feels to me that we've seen a major uptick in what is capable at the frontier, and that, in the context of enterprises, that meant that we were going to see an increasingly wide divergence between whatever the median of enterprise AI usage was and the frontier of the most successful users of AI. The challenge, of course, is that as that frontier accelerates, the gap between the frontier and everyone else compounds. The AI capabilities themselves beget more and more advanced use cases, which allow the deployers of those use cases to have more advantage relative to their peers who are not deploying those particular use cases and capabilities.

20:05And so now I'm exploring the idea of this acceleration gap, but not just as a company phenomenon, but as an individual phenomenon as well. The risk of this is that linear growth in an exponential environment is ultimately a compounding disadvantage and could, if we are being doomy about it, lead, as Kevin suggests, to the creation of a generation of knowledge workers who will never fully catch up. Now, I don't want to definitively say that this is the point that we're at. I'm presenting the acceleration gap more right now as something that I'm exploring than that I feel that I have my head fully wrapped around.

20:35There are plenty of smart people, even who aren't anti-AI zealots, that shared plenty of reasons on Kevin Roos' post that things won't play out this way. Bloomberg's Joe Weisenthal, the host of Odd Lots, and someone who has been going very deep with Claude recently, responded, I doubt late adopters will be impaired very much. For most AI tools, the learning curves aren't very steep, and the interfaces keep getting more intuitive. This is basically the Claude Cowork argument. If it's an Anthropics incentive to launch Claude Cowork to make Claude code-type capabilities available for everyone without having to figure out how to use the terminal, maybe there's a sense in just waiting for those capabilities to come online in a user interface that is, for most people, actually usable.

21:12But why I wanted to talk about it is, one, I do think that this compounding gap is a real possibility with some fairly serious implications for an individual's career. And two, I think that the discourse surrounding AI gets more and more fraught and fracturous every day in ways that I worry will very much not serve the vast majority of people who are just trying to figure this stuff out. Ever since the beginning of this show, I've had the feeling and I've shared my feeling that while the loudest voices are those on the extremes, the people who are incredibly excited about everything that AI has to offer, almost determined to love it no matter what, even in some cases who refuse to see any of the bad sides, and then on the other end of the spectrum, the absolutely determined detractors, the people who are determined to hate it no matter what.