🔥 AGI Isn’t Slowing — Why the pace of AI progress is faster than it looks, and why most people are already falling behind

🧰 Your AI Mentor — A simple, high-leverage prompt to personalize your goals and stay on track (for real this time!)

🎯Sam’s Club ChatGPT rollout — How one enterprise is implementing AI at scale and removing headaches for their managers

Yes, AI does in fact continue to make very rapid progress. Despite headlines of it failing 95% of the time in a company setting, it has actually exceeded most expectations. It can be difficult to see in the moment, but collectively we are moving towards artificial general intelligence according to Zvi Moskowitz. In his recent substack article, Moskowitz says the recent GPT-5 rollout, while rocky, wasn’t representative of the progress that we’re seeing. Because we are on the cusp, these predictions can seem further out than they really are:

Timelines can be fuzzy, and when someone says “imminent” they don’t mean tomorrow or 6 months from now. Yet the competition within AI is real and it continues to scale with new techniques and technology. The big breakthroughs do take time to diffuse throughout the economy, and the companies that can incorporate those breakthroughs sooner will win. You can see the gains from the recent GPT releases below, and how they’ve increased significantly against benchmarks:

Model lock-in and first mover advantage isn’t guaranteed. Yes, OpenAI may be in the spotlight today, but given the exponential acceleration of AI, another company could eat its lunch. We also forget that the original Generative Pre-trained Transformer technology was invented by Google and should have given them a generational lead in AI, but they squandered it.

There is also the factor that isn’t spoken out loud:

We haven’t sat back and really asked ourselves the true root causes of slowing AI innovation. I haven’t heard this one discussed and it would make sense as we move forward. As AI becomes more complex, fewer and fewer people truly are able to apply it and find benefit. So the innovation itself isn’t really slowing down, but the application of that innovation is more challenging because it requires advanced cognitive skills.

So don’t fade AI innovation. You’ll see the Hollywood Effect, where a few people/companies accrue most of the benefit and everyone else gets the scraps. AI is an amplifier and will further divide the haves and the have nots. This is why it is essential to develop AI skills, regardless of your role.

The Bob Kernel — One engineer at OpenAI, nicknamed Bob, wrote an attention kernel so efficient it’s now referred to as “The Bob Kernel.” It runs trillions of times per day across OpenAI’s infrastructure. Just a reminder: the cutting edge of AI still runs on the work of a handful of world-class individuals. Read more →

Waymo’s Safer Than You — After 7+ million miles on public roads, Waymo’s self-driving AI shows 92% fewer pedestrian injuries and 80% fewer injury-causing crashes than human drivers in similar situations. Autonomous vehicles aren’t perfect, just much better. Read more →

Human Nature, Optimized — HumanBehavior.co is a living library of how people think, decide, and act, from mental models to behavioral science. Perfect for anyone designing AI prompts, workflows, or products that actually work with human psychology. Read more →

Sometimes goals, especially personal ones, can feel like a New Year’s Resolution. You do great in the month of January and then all momentum falls off a cliff the rest of the year.

Wouldn’t it be great if you had an AI partner that could not only keep you on track, but also find your individual opportunities that prevent you from achieving your goals? This prompt below helps you develop a simple system to do this, personalized for you:

Given my goal of [insert specific goal], act as a mentor who has already achieved the goal.

When companies implement AI at scale, that is where the real test occurs. Walmart Sam’s Club stores did this recently, releasing an enterprise ChatGPT tools to all of their store leaders.

A few hours saving here and there may not seem like much, but once again scale is the biggest factor. When you extrapolate this across their front line leaders, that could hundreds of thousands or even millions of hours saved. Any leader worth their salt wants to minimize the administrative time and maximize the value added time for their teams.

This will also improve the customer experience and free up time for their staff to focus on delivering better customer experience. No one likes to walk into a store and the item you’re looking for is out of stock. I would also guess that this could improve employee engagement, as managers spending less time searching for insights and information can be more engaged with their teams. I would expect to see other large retailers following in similar footsteps, as this will become the new standard.

Avoid doomers, especially ones that go by the name AI Prophet of Doom. The New York Times wrote an article on Eliezer Yudkowsky, who goes by that nickname. The opening sentence is gold:

Life pro-tip: if the first sentence of an article says that, you can stop reading it and immediately discount much of what it will talk about. No one has crystal ball where they can determine a specific probability like this. Yudkowsky also has written a book titled If Anyone Builds It, Everyone Dies. So you pretty much know what you’re going to get when you read it.

It is this type of fear mongering that does the AI industry a disservice. I see the same tactic deployed by the AI safety “experts” who have created an entire business models on scaring companies to apply their proprietary framework (which in most cases is a waste, and at best checks boxes). If you really wanted to focus on saving lives, maybe start with the millions that die each year from chronic diseases.

It is important to understand the doomer perspective, though, because many people’s first exposure to AI is some sensational headline or article like the one I referenced. It plays into people’s fear of the unknown, and has humans we’re bound to focus on what could go wrong versus what could go right.

AI is going to usher in a new Renaissance, and propel innovation in every field it touches. In many ways it already has, from scientific research to being available for use to anyone with an internet connection. What is important to remember is that while the original Renaissance was the rebirth and explosion of ideas, it was also a time of conflict and war. I think we’ll see something similar, and we already are with how polarized and violent politics are becoming.

-Ylan