[clarionbell]

He would be among those who lack "healthy inclination to skepticism" in my book. I do not doubt his brilliance. Personally, I think he is more intelligent than I am.

But, I do have a distinct feeling that his enthusiasm can overwhelm his critical faculties. Still, that isn't exactly rare in our circles.

[iLoveOncall]

It's not about that, he just will profit financially from pumping AI so he pumps AI, no need to go further.

[stephc_int13]

I have the same feeling.

Everything Karphathy said, until his recent missteps, was received as gospel, both in the AI community and outside.

This influencer status is highly valuable, and I would not be surprised if he was approached to gently skew his discourse towards more optimism, a win-win situation ^^

[runlaszlorun]

What are his recent missteps?

I'll confess I try to ignore industry chatter to a fair degree.

[Verdex]

I think many serious endeavors would benefit from including a magician.

Intelligent experts fail time and again because while they are experts, they don't know a lot about lying to people.

The magician is an expert in lying to people and directing their attention to where they want it and away from where they don't.

If you have an expert telling you, "wow this is really amazing, I can't believe that they solved this impossible technical problem," then maybe get a magician in the room to see what they think about it before buying the hype.

[runlaszlorun]

Ha, great analogy.

[whycome]

CMO?

[lawn]

Intelligent people are very good at deceiving themselves.

[sdf2erf]

Im gonna go against the grain and say he is an elite expert on some dimensions, but when you take all the characteristics into account (including an understanding of people etc) I conclude that on the whole he is not as intelligent as you think.

Its the same reason why a pure technologist can fail spectacularly at developing products that deliver experiences that people want.

[newyankee]

More like people know where to hype, whom to avoid criticising unless measured etc. I have rarely seen him criticising Elon's vision only approach and that made me skeptical.

[CamperBob2]

Unfortunately my understanding is that he originated the vision-only approach while at Tesla. As in, he's the one who sold Musk on the idea in the first place.

I don't have time to dig up the citation that someone pointed me towards, but it's out there and can be found. Which is a bummer, because I've learned a lot from his videos and writing and have a lot of respect for his work.

[sdf2erf]

I personally dont believe he is trying to profit off the hype. I believe he is an individual who wants to believe he is a genius and his word is gospel.

Being picked by Elon perhaps amplified that too.

[aleph_minus_one]

> Im gonna go against the grain and say he is an elite expert on some dimensions, but when you take all the characteristics into account (including an understanding of people etc) I conclude that on the whole he is not as intelligent as you think.

Intelligence (which psychologists define as the g factor [1]; this concept is very well-researched) does not make you an expert on any given topic. It just, for example, typically enables you to learn new topics faster, and lets you see connections between topics.

If Karpathy did not spend a serious effort of learning to get a good understanding of people, it's likely that he is not an expert on this topic (which I guess basically nobody would expect).

Also, while being a rationalist very likely requires you to be rather intelligent, only a (I guess rather small) fraction of highly intelligent people are rationalists.

[1] https://en.wikipedia.org/wiki/G_factor_(psychometrics)

[sdf2erf]

" Karpathy did not spend a serious effort of learning to get a good understanding of people

This does not come from spending effort in learning people - its more innate. You either have it or you dont. E.g. you cant learn to be 'empathetic'.

It always boggles my mind when people dont consider genetic factors.

[autoexec]

> you cant learn to be 'empathetic'.

I know it was just an example, but there's research suggesting otherwise. There are things you can do to increase/decrease empathy in yourself and others. If you're curious, it might be worth looking into the subject.

[hnfong]

There is the autistic spectrum, and there is understanding of people and psychology. Autistic people might have a hard time understanding people, but it's not like everyone else is magically super knowledgable about human psychology and other people's thought patterns. If that were the case, then any non-autistic person could be a psychologist, no fancy study or degrees required!

Unless your point is to claim that Karpathy is autistic. I don't know whether that's really relevant though, the original issue was whether/how he failed to recognize the alleged hype.

[pluralmonad]

You are what you do. If you want to develop your empathy, spend time/energy consciously trying to put yourself in the shoes of others. Eventually, you will not have to apply so much deliberate effort. Same way most things work.

[bwfan123]

> you cant learn to be 'empathetic'.

I would tend to disagree. The tech types have a strong intellectual center, but weaker emotional and movement centers. I think a realignment is possible with practice. It takes time, and as one grows older, the centers begin to integrate better.

[aleph_minus_one]

Being empathic is different from "understanding people".

Psychopaths and narcissists often have a good understanding of many people, which they use to manipulate them, but psychopaths and narcissists are not what most people would call "empathic".

[sdf2erf]

They dont understand people. They understand how to control people, which is completely different from the context of building products that people want - which requires an understanding of peoples tastes and preferences.

[aleph_minus_one]

> which is completely different from the context of building products that people want - which requires an understanding of peoples tastes and preferences.

Rather: it requires an understanding how to manipulate people into loving/wanting your product.