Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

Ok?

1. That’s his job as an investor. Getting one big outlier prediction right makes up for a hundred wrong ones.

2. These predictions tend to provoke interesting debates(that’s probably Sama’s intent here).



> Ok? That’s his job as an investor. Getting one big outlier prediction right makes up for a hundred wrong ones.

But he's not acting as an investor here. He's acting as an influencer on Twitter. And when a public figure consistently makes poor predictions, I think they should be called out for it.


I mean that it’s what VCs always do. Just as macro investors and political scientists throw out wild prognostications about global affairs all the time.

It’s the equivalent of a painter doodling in his notebook. Some of them will turn into paintings.


1. If it’s his job at being able to predict, he seems to be doing a bad job by being wrong so many times.

2. There’s so many better ways to spark an interesting conversation than throw bs predictions out on Twitter. If his opinion was informed with facts, he could write a blog or (heavens forbid) a Twitter thread detailing what lead him to the conclusion.


To be fair, his post does lead into a Twitter thread. Unfortunately it's a mess of the usual reactionary takes: white people are asked to feel guilty; Asians are discriminated against; nobody wants to hear whiny conservatives on campus. And so: "nearer to collapse than it appears." That's not a falsifiable prediction, of course: how long does it "appear" to take to collapse? A decade, a century, never?


no matter how close you think it is, it is closer than that!

- altman's paradox


>That’s his job as an investor

however unlike in his job as an investor when he makes stupid takes on twitter he isn't going to lose 100k each time. If he did he would probably vet his tweets a bit more.

The reason markets work and Twitter doesn't is because there's no negative consequences for posting without thinking.


Why should we pay any attention to someone who needs get only one prediction right? That suggests a probability of correct predictions that approaches zero. Let me know when a prediction comes out true. I would be foolish to pay any attention before then.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: