Box founder Aaron Levy made headlines this week for a social media post suggesting that tech company CEOs are “prone to AI psychosis.”
On the latest episode of TechCrunch’s Equity podcast, Kirsten Kolosek, Sean O’Kane, and I did our best to clarify Levy’s comments. First, we noted that he is not denying AI tools, only that CEOs need to actually use them to understand them.
This is relatively mild skepticism compared to other signs of widespread backlash, such as graduating college students booing the mention of AI, bad vibes around layoffs in the tech industry, and the apparent spike in installs of search engine DuckDuckGo after Google announced it would bring AI to its search experience.
Kirsten suggested that Google faces the dilemma of “chasing what it feels like it has to do to catch up, but it’s undermining what people are most attached to a brand, and that’s not getting better.” More broadly, she wondered, “could this anti-AI moment be an opportunity for startups and other areas of business?”
Continue reading for a preview of the conversation, edited for length and clarity.
Anthony Ha: AI is incredibly polarizing. And that’s part of what’s hard to talk about. You might feel a little crazy because (at the same time) everyone uses it and everyone loves it, but at the same time no one uses it and everyone hates it. There is a large contingent where both of these apply.
On the user side, one of the things that was really impressive was Google’s announcements around search and how AI is (already) becoming a big part of search. However, it was interesting to see Google trying to back off on that a bit, or at least add some nuance in that if you want the 10 blue links experience, there’s still a way to get it. It doesn’t go away completely.
But I think a lot of people aren’t excited about the direction Google is going. So, for example, DuckDuckGo just announced a 30% increase in installs, which is a huge jump. Of course, DuckDuckGo is a much smaller product than Google. I don’t think Google is in immediate trouble, but I think this is a sign that there are many users who don’t like the current direction of AI.
Sean O’Kane: When I look at the big AI labs and technology companies that are really pushing AI capabilities and products forward, there’s one thing I always look for. It seems to me that we’re collapsing towards the Anthropic approach, which is the idea of really trying to understand what you want to offer people and sticking to it.
And Google is actually moving in a different direction right now. They’re trying to do different things, but it’s not good for them to be too vague about it.
What I mean is that when Google gets on stage at IO and talks about how it’s going to change search, most of what they’re talking about is something that ends up in shopping and commerce. And I think what we collectively think of Google, especially many of us who have been using Google for 20 to 30 years, is an information retrieval system.
Google can have a really hard time with this problem, and there’s a reactive fear that it’s hurting aspects of information retrieval. The reaction then is, “Yeah, but it still exists. Let’s focus on how it helps us book things like airline tickets.”
And they also shoot themselves in the foot by releasing them. Stress testing these systems should be very difficult, but they’ve released, released this, and are running into the same problems they’ve been running into for years.
Kirsten Korosec: There’s a great article published about Google doesn’t know how to spell your name. If you ask Google, “How many Ps are there?” it will say 2.
It’s the tension between Google chasing what it feels it has to do to catch up, but ruining what people are most attached to a brand and not improving it.
What I’m wondering about is that we’re already seeing some early evidence of people’s fingers voting and walking on their behalf by literally going to another service. But it is questionable whether this anti-AI moment represents an opportunity for startups and other areas of business that have been less thought of so far, or culturally for other startups.
Anthony: Of course. Again, it’s probably difficult because there are so many different opinions. And if you build something for a group of people who are skeptical about AI, you’re probably going to alienate other users who are more evangelical and enthusiastic about AI. But I think that’s exactly the moment we’re living in.
And if you look at how DuckDuckGo promotes themselves, you can see that they’re very emphatic about the idea that they’re anti-AI. As I mentioned earlier, this is very impressive as I am moving away from Google (myself) and trying other search engines. And a year ago when I started that quest, even these alternative search engines were still trying to experiment with AI features and were emphasizing AI to some degree. Because they thought they had to do it too.
And I think now they understand that there’s a lane to say, “No, we just weren’t interested in that kind of thing at all. Or, as long as we’re doing it, we’re putting it in a separate sandbox so it doesn’t impact the core search experience.”
Kirsten: I think we sometimes unfairly categorize all the CEOs of tech companies as force-feeding people AI. And at least one tech CEO has come out and said, “I think some of the other tech CEOs are a little bit psychotic when it comes to AI.”
I’m talking about Aaron Levie, the founder of Box. He’s been to Disrupt many times and is a friend of TechCrunch. He makes these comments about how CEOs are so capable that AI psychosis is endemic, and I’m reading this article, “that we’re far from the last mile work that still needs to be done to create the most value with AI.”
I thought that was really interesting. And I’m wondering if there are other CEOs who agree with this. I also think you’re thinking about how your workforce is changing as part of a shift in thinking about what needs to happen to create the most value. This is another topic for today. () We’re not just thinking about the AI divide, but also about how AI is changing work. And certainly, we’ve seen some of the bad news sides of that. That’s a lot of layoffs.
However, I also believe that major changes are occurring in the way people work. I’m wondering if there’s evidence of that in the areas you both cover, because I don’t think it’s just the quote-unquote “AI startup sector” or the big tech companies.
Sean: In terms of the companies that I cover, a lot of them tend to be working on things that are adjacent to, if not physical transportation. And, of course, on the software side it looks much slower than it actually is.
We are beginning to see some of that change. We talked a little bit on the show about Rivian CEO RJ Scaringe’s spinout, Mind Robotics. And you know, we’re definitely seeing more and more AI being applied to physical infrastructure, manufacturing, robotics, and autonomous driving.
I think what’s really changing is on the software side. There are people whose jobs are directly related to writing code.
Anthony: I think part of your question involves both the adoption of AI in companies and the subsequent AI layoffs. To what extent is it top-down or bottom-up?
Because I think a lot of the other transformations in the workforce over the past few decades have been bottom-up, at least to some degree. People actually want to use these tools, people adopt them, and at some point executives and IT managers embrace them.
Many of the productivity gains that AI is believed to have are being embraced by executives, or, in the case of startups, perhaps the VCs funding them, who love this dream that small teams can be as effective as companies with larger teams.
I don’t think that’s necessarily impossible, but I think what Aaron is essentially saying is, if you’re not actually working on the final work, how do you know? And while he’s not someone who says we should throw away all AI tools, he does say we should actually use these tools and understand what they’re doing. You can’t look at a slide and think, “Yes, that’s incredibly efficient. Let’s get started.”
Kirsten: Well, I think there’s a lot of real evidence that these companies are using these tools and that it’s directly impacting workers in the form of layoffs. It’s also affecting the way they work. Two truths are accurate here.
If you buy through links in our articles, we may earn a small commission. This does not affect editorial independence.
