BREAKING: OpenAI Launches Deep Research

I truly hope this new AI tool wipes out the entire industry of useless Wall Street sell-side "analysts" and corporate healthcare "consultants."

Welcome to AI Health Uncut, a brutally honest newsletter on AI, innovation, and the state of the healthcare market. If you’d like to sign up to receive issues over email, you can do so here.

I truly hope OpenAI’s new Deep Research feature, introduced tonight, wipes out the entire Wall Street sell-side analyst industry. These folks are the worst at their jobs relative to how much they’re paid and have contributed to trillions in losses for their clients over the past few decades.

I also hope OpenAI’s Deep Research wipes out the absolutely useless industry of healthcare-focused consulting and market research firms like McKinsey, BCG, and IQVIA—companies that charge millions and take months to produce mediocre research.

That said, I have my fair share of skepticism about this new feature. Since it still relies on GenAI (Generative Artificial Intelligence) models, hallucinations are inevitable. The biggest issue are references and citations—they are often either wrong or nonexistent. That’s been a problem with ChatGPT. That’s been a massive problem with every other research tool that relies on GenAI, including Stanford’s Storm. In any high-quality research, references are absolutely essential.

The Deep Research team mentioned in their presentation tonight that one could wait “up to 30 minutes” for an answer. So it better be good. But if o1 pro is any indication, waiting 3.5 minutes made very little difference—at least for my tasks—compared to ChatGPT-4o.