Instead of killing jobs, generative AI is fueling a surprising job boom, according to venture capitalist Marc Andreessen.
In a recent podcast, the a16z cofounder said tech companies are hiring thousands of highly educated workers to help them solve a major bottleneck in the development of AI models.
OpenAI and Google are struggling to significantly improve their next AI models, according to multiple media reports this week. That's despite sucking up all human data from the internet for free and building huge, expensive clusters of GPUs to train these models.
Part of the problem, according to Andreessen, is that there's not much new, quality human data left to use. That's what makes these AI models smart, and without that, OpenAI and others are hitting a worrying wall.
"These systems are a function of their data more than anything else. They're a function of the training data," Andreessen explained. "And basically, the big models are trained by scraping the internet and pulling in basically all human-generated training data, all human-generated text, and increasingly video and audio and everything else. And there's just literally only so much of that."
Ben Horowitz, Andreessen's cofounder at a16z, chimed in, saying, "We're running out of human knowledge."
One solution is to hire a bunch of humans to literally churn out new knowledge and then pump that into the AI model training process.
"They're so gated on data that they're literally going out and hiring thousands of programmers and doctors and lawyers to actually hand write answers to questions for the purpose of being able to train their AIs," Andreessen said. "It's at that level of constraints."
This is not the future that was feared when OpenAI's ChatGPT burst onto the scene almost two years ago. Since then, dire warnings of job destruction have been issued, with lists (published by Business Insider and other publications) predicting roles in areas such as coding and law would succumb to the AI job apocalypse.
At least for now, some of these professions are needed to help feed new AI models with fresh, accurate human data.
The irony of this was not lost on Andreessen and Horowitz, who chuckled during the podcast.
"Well, actually, this is also part of one of the big fears is AI driven unemployment, and the irony is what's happening today is an AI hiring boom," Andreessen said. "And a big part of the AI hiring boom is actually hiring the experts to actually craft the answers to be able to train the AI."