Hey Reader,

A lot of people think that February 2026 will mark a major turning point in how the world thinks about AI.

I have some doubts about their specific claim. Before I get to those, here are the news items that inspired it:

1. Major new releases from Anthropic are causing investors to sell tons of stock in old-school software companies. The result is major losses in the value of companies like Salesforce, Intuit and DocuSign.
   ​
   Since these are software-as-a-service companies, the term "SaaSpocalypse" has been coined 😱.
   ​
2. A viral essay encouraged everyone to come to terms with the fact that AI is going to be able to do a great job with pretty much all computer-based work in the near future.
   ​
   Specifically, OpenAI's Codex 5.3 and Anthropic's Claude Opus 4.6 are major leaps forward in quality.

Longtime readers of Innovating with AI will be familiar with all the big ideas presented in both links above. We've talked for years about the potential for AI to do complex office tasks and build software – in fact, students in our programs build stuff like this (and use these AI tools for work and business) every day.

What makes the essay I linked above really spicy is the comparison to the early days of the spread of Covid. The author, startup founder Matt Shumer, sets the stage in a way that makes you feel like we are currently in the moment where we started to watch the disease spread from China into Europe, but we hadn't yet fully realized what was coming.

The idea here is that we're on the edge of a really big shock – but instead of worldwide lockdowns, it's going to be AI becoming better than humans at complex computer work.

The corporate adoption barrier

The author and I both run small companies, so we are able to move fast. (I also just found out on LinkedIn that we went to the same college, so we must both be really smart. 🙂)

However, roughly 60% of American white-collar workers – which I'll use as a synonym for "knowledge workers" or "people who mostly work on computers" – work at large enterprises with 500+ employees. There's a similar stat showing that large businesses (500+ employees) account for a bit over half of all employment.

Some large companies will adopt AI quickly, but on the whole, they are very slow-moving. This is not a "we are locking down the whole city overnight" situation. It's also not a "Lehman Brothers goes bankrupt and sends all its employees home crying" situation.

Instead, the most likely outcome is that companies continue to slowly and somewhat haphazardly adopt new AI tools, experimenting in fits and starts. The tools will get better over the long term, but I don't think we'll look back at February 2026 and say "that was the month it all came crashing down."

In other words, there is a lot of inertia and status quo bias, especially in larger companies. I would also argue that many small businesses (the bulk of which have 1-4 employees) that are not staffed by tech-savvy people will have a similar status quo bias.

This is a human psychology thing that a lot of Silicon Valley folks regularly struggle to comprehend. If you are really excited by technological change, this is a wonderful moment for you, because you can do what Matt Shumer, the IWAI team and our 200,000+ readers are doing: get ahead while AI is still relatively new.

But the vast majority of people have no strong desire to do this – in fact, they dislike change and prefer normalcy and stability.

Likewise, business owners (especially owners of large companies) generally benefit from stability. They want to see things like a 10% annual growth rate and healthy dividends. Volatility is bad. Most corporate boards would prefer a stable situation with a predictable payoff over a volatile situation where they might come out a big winner or a big loser. (This is also why businesses dislike seemingly random and ever-changing tariff rules.)

Startups are very specifically built around disruption of the status quo, but most businesses disagree.

Fast improvements in tech, slow change in society

Anthropic, OpenAI, Google and the rest are going to continue to rapidly roll out new and better AI models. I think this is generally positive, and especially good for people who want to be early adopters like you and me.

But no matter how good Claude Opus 5 ends up being, I don't think it can force a faster pace of adoption among the vast majority of business leaders who prefer the status quo. Even the most ruthless corporate baron is generally going to understand that firing all your employees and replacing them with AI is an extreme and risky maneuver.

Paradoxically, the smart play for most big businesses is going to be to adopt AI relatively slowly. Benefitting from the status quo for a while longer is almost a sure bet. Replacing a large number of human workers with AI is comparatively risky.

There is also a risk that rapid movement toward AI-related mass layoffs in any sector (e.g. software as a service) could create volatility in all sectors (e.g. trucking and logistics) even if they are seemingly unrelated. This is a scenario that any large company would want to avoid (but which Robinhood options gamblers would love) since it is inherently unpredictable.

So, I think the disruption enthusiasts are ultimately going to be frustrated by the psychology and risk-reward calculations of the vast majority of business leaders. The AI labs might even release Artificial General Intelligence and find that the vast majority of humans would rather take their time.

Status quo bias gives society a buffer against catastrophic change

Even though advancements in AI are generally good for me, my team, and my readers, I don't want to benefit in a way that results in catastrophic change for most people.

(This is why I enjoy paying my taxes – it's narrowly bad for me but broadly good for my community, and I benefit a lot from living in a community full of people who are happy, healthy and safe.)

Status quo bias creates a natural speed limit on major societal change. When you want things to change faster, this can be frustrating. But because of how dramatic a change AI is likely to create, I think spreading that change out over a decade instead of a year seems like a far healthier way to operate.

Until next time,

– Rob
CEO of Innovating with AI