News for hackers. Stuff that matters.
Re:Microsoft and OpenAI Working on ChatGPT-Powered Bing in C... by virtual_void on Wednesday January 04, 2023 @05:33AM (#34244074)
The parent seemed to suggest this was at an individual level, rather than everyone being forced to get information from some consensus trusted source.

There’s no need, with this idea, for you to have to get your info directly from the Vatican. You can decide you trust someone different.

Re:Microsoft and OpenAI Working on ChatGPT-Powered Bing in C... by jhoelzel on Wednesday January 04, 2023 @05:43AM (#34244119)
and how are you going to set up those experts for "your search ai"?`

Because its really going to be different from the one your neighbour has.

How do you select turstworthy people on a new topic? Does every search start with "hmmm i found 6521 people blogging about this, please click on the trustworthy avatars you like"?

The concept you are trying to apply here, will just lead to your own bubble. Im not even going to go into detail of "orange men knows best" but this example alone should shock you.

Re:Microsoft and OpenAI Working on ChatGPT-Powered Bing in C... by afterburner on Wednesday January 04, 2023 @06:22AM (#34244339)
You find out about a person, evaluate them, and subscribe to their newsletter.

This isn't about getting quick answers.

Re:Microsoft and OpenAI Working on ChatGPT-Powered Bing in C... by ranger47 on Wednesday January 04, 2023 @07:21AM (#34244706)
It's interesting that people don't have the patients to vet or even start to question the veracity of their information sources. I think this part of the "how do you expert?" problem is the one that needs to be more carefully examined and solved. We can probably come up with 10 different ways for our algos and AIs to present "experts," but is everyone going to trust those methods? Not likely, so we fall back to the problem of "my expert is better than your expert" regardless of the talent pool or selection process. So, how do we solve that? Would people make better choices if the were given better options, or does human irrationality always win, in the end?
Re:Microsoft and OpenAI Working on ChatGPT-Powered Bing in C... by virtual_void on Wednesday January 04, 2023 @05:47AM (#34244136)
No idea - it’s quite hard. I didn’t suggest the original idea and I’m not trying to apply the concept. I was merely pointing out that your argument was based on a false premise.

Now you’re arguing about implementation, which is different.

Re:Microsoft and OpenAI Working on ChatGPT-Powered Bing in C... by koonsolo on Wednesday January 04, 2023 @08:59AM (#34245545)
Well, my solution is not trying to solve humanities problem you know, it's just trying to solve my own problem. And I'm just guessing I'm not the only person who has this problem.

So how to vet people? Well, Paul Graham started a VC company and has seen a lot of startups, so my guess is that he knows a lot about the topic. More so than anyone who hasn't come close to any startup.

On (game) development, I would go for Carmack and the likes.

And of course just like PageRank, if they have vetted for certain people, those also get scored.

Does it lead to your own bubble? Of course! But if you're going to get a religious person to dump his own religion in favor for another, good luck to you. But that's not the kind of problem that I want or need to solve.

I just want reliable information. And I get to decide what "reliable" is. If you're a flat-earther, you can vet any crazy person you wish. I don't care, because it won't affect me at all.