Why Humans Are Still Powering AI [Sponsored]

There's a dirty secret, isn't there? In in Silicon Valley and in the tech world that there is and I don't know if people realize the extent to this. There is an absolutely huge importance on human data and human expertise and human understanding, And that is completely glossed over. Yeah. I mean, fundamentally, artificial intelligence is founded in human intelligence. And I think in the the stack of data, algorithms, and compute, I think the the human data element is often the the least spoken about, maybe the least least glamorous. People want to imagine that there's a a simple kind of input output equation, but ultimately, there's a messy layer in the in the stack of human beings who are providing their data to, you know, either, label data, provide all HF post training data, and then ultimately the evaluation and the assessment of the the model model performance all ultimately is kind of has an element of of human human data in it. I I can't tell the difference between a doctor and someone who pretends to be a doctor. The only way that you can actually tell the difference is if you have this deep abstract understanding and you know that they're breaking the rules. It's increasingly clear that Frontier models as a as a platform are going to be centralized and controlled by a relatively small number of of players, which at the moment is almost exclusively these US tech companies. So I think there is a bit of a a wake up call. What the future will hold is basically a marketplace of intelligence. So in the past, we had a marketplace of, oil, for example, or electricity. Intelligence is gonna be the new traded thing. I'm the cofounder and CEO of Prolific, and, Prolific is a human data, infrastructure company. So we make it easy for people developing frontier AI models, and running research to get access to trustworthy, high quality participants for high quality online data collection. Prior to the chat GBT moment, the the primary modes of data collection, the people were fairly fungible, right, so you're optimizing for cost and scale, maybe offshore lower cost labor, which I think created this dynamic of human data for AI being a bit of a a dirty secret. Let's talk about your core technology. Now you you've solved an interesting problem that I've tried to solve in the past. So I I I started a company called Merge, and it was a code review platform. Mhmm. And it was exactly the same thing. I realized that code review has to be done by humans, and people talk about automation and software engineering life cycle. It's mostly bullshit. It's it's actually orchestration. You need humans involved in every single step, most importantly, code review. We had a skill matrix, and we could learn their skill, and and we could you know, a a pull request would come in from one of our customers, and …