“LLMs are literally the midwit of the center of the bell curve.”
Sharp Tech with Ben Thompson · Tech Strategy & Big Tech · September 2026
Thompson is talking about tech's habit of mistaking more data for certainty. He calls LLMs consensus mechanisms that sample from the middle of a probability distribution, and says they avoid the edges because the edge that holds brilliance also holds the crazy.
Transcript
But there's a reason those rockets never ran up. By the way, all things considered, they are underrated. We're a great team. They ran up against one of the greatest teams of all time. And this is why it was so devastating. That's a round one. That was the
ultimate problem for Daryl, but it is what it is.
But you always go back to Steve Jobs and making fun of complaining about technology being obsessed with feeds and speeds. And there are things that can't be measured. And there obviously are. And a consequence of increased transparency, of increased tracking, of increased data is when there was no. Data, just tracking a little bit gave you this big advantage. But there's this false sense of certainty that arises from more data. Just because you have more data doesn't mean you have all the data. And tech has always been susceptible to falling in love with data. Data definitionally is backwards looking. Like the LLMs want me to cite something when I'm saying something that's happening right now or prediction about the future because that's how they work. They are consensus mechanisms. Like LLMs are literally the midwit of the center of the bell curve. Like that's what they are. It is a probability distribution. They're going to pull from the center of the distribution. They're not going to pull from the edges. And why? If you pull from the edges, you might pull from the crazy side as much as you pull from the brilliant side.
You can get things wrong and you're sort of engineered to be accurate.
So I said in that article, like this is clearly like half of the political divide. And it didn't like that because it's like, you're going to make people upset. Like, but look, that's literally what's happening. And the reason it's happening is because this is unfalsifiable. It's something that can't be proven. The overall precept of actually government needs to do more. There needs to be this, an entire bureaucracy around controlling this. There needs to be people that have jobs to monitor, to do things. Sorry, it's just a reality. There is a one side of the political divide that leans towards more government, that leans towards more bureaucracy. And this is not a commentary about whether that's good or bad. I'm obviously fairly skeptical of that and concerned about its impact on long-term innovation and all those sorts of things. But we can't just, I have to point that out. It's just a reality.
Not only that, it's going to be a reality that informs how policy is crafted here. And I do have some insight because I'm in Washington, D.C. and have a bunch of friends who are in politics. And I remember, I think it was like nine months ago, I met with somebody who is consulting with a potential 2028 presidential candidate. And she was asking me, like, what should our AI policy be? I don't know what we should be pushing. And I think the entire Democratic Party has been sort of casting about looking for what principle