Thrive AI Health is useless.

Survive AI

OpenAI CEO Sam Altman and Arianna Huffington — yes, the publishing mogul who started the Huffington Post during the mid-Bush administrationthought they could improve healthcare with AI. But so far, their joint venture Thrive AI Health seems more useless than reading WebMD when you have a headache. 

A demo of Thrive's AI product, as uncovered by TechCrunch and which doesn't seem like it was supposed to be live online, suggests that the company's product will take the form of an interactive health-tracking tool. The Thrive AI Coach seems to operate like a clumsy version of OpenAI's ChatGPT, supplying users with unimaginative prompts like "create me a workout" and the typo-afflicted "what were [sic] my heart rate over the last week?"

But while it incorporates AI, Thrive's product doesn't currently appear to be functionally different from services like the iPhone's built-in health app, or even more cutesy trackers like Pokémon Sleep, which documents your slumber through a tubby Snorlax. Despite the gimmick, at least Pokémon Sleep is honest about what it is. 

From what we can see of Thrive AI so far, though, it looks like it overpromises a revolution when it's recycling ideas — a typical marketing strategy for overconfident companies peddling AI, and a familiar pattern for both the hype-fueled Altman and the bombastic Huffington.

Artificial Benevolence 

Many startups have tried to reinvent existing product categories using AI, but they tend to run into familiar roadblocks, like when a would-be office assistant AI went off the rails and started Rickrolling clients, or when a bot intended to council people struggling with eating disorders instead gave harmful dieting advice

In a joint op-ed published in Time over the summer, Altman and Huffington tried convincing readers that their product is uniquely innovative. 

"Health is also what happens between doctor visits," they write. "In the same way the New Deal built out physical infrastructure to transform the country, AI will serve as part of the critical infrastructure of a much more effective health care system that supports everyday people's health in an ongoing way." 

In a broad sense, they're probably right. But in reality, an AI chatbot currently can't bring people vitamins, or perform scans or surgeries, or provide people with the money necessary to pay for healthcare in the first place. 

More on AI: OpenAI Alarmed When Its Shiny New AI Model Isn't as Smart as It Was Supposed to Be.


Share This Article