Hacker News Reader

Doctors were using Google to diagnose patients before. The thing is, it's still the doctor delivering the diagnosis, the doctor writing the prescription, and the doctor billing insurance. Unless and until patients or hospitals are willing and legally able to use ChatGPT as a replacement for a doctor (unwise), ChatGPT is not about to eat any doctor's lunch.

View on HN

Not OP, but I think this makes the point, not argues against it. Something has come along that can supplant Google for a wide range of things. And it comes without ads (for now). It’s an opportunity to try a different business model, and if they succeed at that then it’s off to the races indeed.

It makes one point: that LLMs are useful.

The other point is still suspect: that LLMs will ever scale to AGI.

Which specifically means reliability and explainability for higher-order thinking.

The writing is on the wall that LLMs are going to automate failure-tolerant work.

But the rub there is that failure-tolerant work is also tolerant of less than state of the art, cost-optimized LLMs.

Which leaves OpenAI where? AGI or bust.

And I wouldn't take that bet, when MS, Google, and Apple are alternative options.

Doctors weren't paying for Google either. If ChatGPT or other LLM AIs play that same role, the OP remains correct.

Reply on news.ycombinator.com