Live data from Hacker News

John Carmack: I’m going to work on artificial general intelligence

facebook.com

91–100 of 930 posts

Re: John Carmack: I’m going to work on artificial general intelligence

#91

This doesn't surprise me at all. He went on a week long cabin-in-the-middle-of-nowhere trip about a year ago to dive in to AI (that's all this guy needs to become pretty damn proficient). (edit: I'm not claiming he's a field expert in a week guys, just that he can probably learn the basics pretty fast, especially given ML tech shares many base maths with graphics) As recent as his last Oculus Connect keynote, he exto…

> He went on a week long cabin-in-the-middle-of-nowhere trip about a year ago to dive in to AI (that's all this guy needs to become pretty damn proficient).

You must be joking, right? I'm as much of a Carmack fan as anyone here, but overstating the skills of one personal hero does no good to anyone.

Re: John Carmack: I’m going to work on artificial general intelligence

#93

This doesn't surprise me at all. He went on a week long cabin-in-the-middle-of-nowhere trip about a year ago to dive in to AI (that's all this guy needs to become pretty damn proficient). (edit: I'm not claiming he's a field expert in a week guys, just that he can probably learn the basics pretty fast, especially given ML tech shares many base maths with graphics) As recent as his last Oculus Connect keynote, he exto…

> He went on a week long cabin-in-the-middle-of-nowhere trip about a year ago to dive in to AI (that's all this guy needs to become pretty damn proficient). You must be joking, right? I'm as much of a Carmack fan as anyone here, but overstating the skills of one personal hero does no good to anyone.

To be fair, a whole uninterrupted week of highly focussed work can get you pretty far (considering that you have the necessary background, which Carmack has, i.e. related to linear algebra, stats, programming, etc.)

Re: John Carmack: I’m going to work on artificial general intelligence

#94
After maybe 10-15 years of being an AGI enthusiast who never dared to pursue it "seriously", something has prompted me in the last few months to decide to make my next side project into a "real" AGI effort.

I think maybe its just the number of people who are talking publicly about working on it that is making me want to "work on it" "seriously"?

I mean, when I get my current side project "out the door" to some degree, I plan to spend at least a few months where the weekend (or sometimes nights) project that I actually admit to working on is "AGI research". Previously I have occasionally spent a few hours here or there mainly passively trying to learn about some deep learning or AGI thing by skimming papers or watching videos. But the plan now is to actually work on active learning projects/experiments for several hours every weekend. For at least two or three months (or longer if I don't give up before then).

Theoretically at least some of it could be of practical use, although I am thinking that I may avoid trying to become a deep learning expert because it seems like people have that covered and it might take me five years. Lol. So I am trying to think of GPU programming approaches that are new. Which most likely will turn out to be a waste of time but will certainly be interesting for me.

Re: John Carmack: I’m going to work on artificial general intelligence

#95

Earlier quoted context omitted.

Unfortunately VR turned out to be an intensely nauseating puke fest for a lot of people.

Same as the last go round. In ten years, a new generation will think that this time, they know how to do it right. Maybe the hardware will even. have caught up.

I'm still convinced we'll have "good enough" barebones yet spatially aware AR first. It's just less intensive on the front end side, which is the limiting factor now as I understand it.

VR is a whole other thing, and I think the "uncanny valley" stretches quite far; ie you need quality really close to what we see in movies to pass the acceptable threshold beyond a few hours of novelty.

Re: John Carmack: I’m going to work on artificial general intelligence

#96
post #28

John Carmack seems like a guy that would've made fundamental contributions to science, had he been born a century before. Computer science now occupies the place physics once did, in its impact on moving the world forward. Best of luck to him, I look forward to seeing what he produces!

You guys really get high on your own supply don't you? It's a toss-up these days whether something from computer science or a programmer will either enhance or erode the human experience and quality of life.

I don't have the same grandiose notions of CS that OP has, but you could say that about any research field.

Physics gave us airplanes and bombs to drop from them.

Biology and chemistry gave us pharmaceuticals and pollutants.

You can do this for any field of study really.

Re: John Carmack: I’m going to work on artificial general intelligence

#97
It's as if Carmack is simply paid to shill for Zuckerberg, like some pet under glass.

Every time Carmack opens his fucking mouth, and it's the facebook domain in the URL bar, there's no accident in that. Carmack transmits this nerd signal, knowingly.

Kind of disgusting.

Re: John Carmack: I’m going to work on artificial general intelligence

#99

Earlier quoted context omitted.

You guys really get high on your own supply don't you? It's a toss-up these days whether something from computer science or a programmer will either enhance or erode the human experience and quality of life.

Hehe, I love CS, but give me significant advances in physics, chemistry or medicine over some new "breakthrough" in machine learning.

A general AI would give you all 3

Re: John Carmack: I’m going to work on artificial general intelligence

#100

This doesn't surprise me at all. He went on a week long cabin-in-the-middle-of-nowhere trip about a year ago to dive in to AI (that's all this guy needs to become pretty damn proficient). (edit: I'm not claiming he's a field expert in a week guys, just that he can probably learn the basics pretty fast, especially given ML tech shares many base maths with graphics) As recent as his last Oculus Connect keynote, he exto…

> He went on a week long cabin-in-the-middle-of-nowhere trip about a year ago to dive in to AI (that's all this guy needs to become pretty damn proficient). You must be joking, right? I'm as much of a Carmack fan as anyone here, but overstating the skills of one personal hero does no good to anyone.

I'm not saying he's LeCun, I'm just saying he gets up to speed absurdly fast. So it's not unreasonable to suppose that by now, he's learned enough to start seriously contributing to this kind of problem.

edit: to be clear, all I'm saying is he can catch up to the body of research already out there quicker than the average bear, and he's shown a real knack for designing solutions and being crazy productive. I'm not pretending he's gunna be publishing insane novel research anytime soon, just that I wouldn't be surprised if he ends up being a real voice in the field.

Post reply on HN