John Carmack: I’m going to work on artificial general intelligence • Hacker News Board

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

#591

Post by jack_pp » Thu, Nov 14, 2019, 9:36 AM UTC

MaximumYComb wrote: post #579

Animats wrote: post #437

Earlier quoted context omitted.

I've long taken the position that intelligence is mostly about getting through the next 10-30 seconds of life without screwing up. Not falling down, not running into stuff, not getting hurt, not breaking things, making some progress on the current task. Common sense. Most of animal brains, and a large fraction of the human brain, is devoted to managing that. On top of that is some kind of coarse planner giving goals…

I was thinking about the manipulation issue tonight. I'd been throwing a tennis ball in the pool with my kids and I realised how instinctual my ability to catch was. A ball leaves my kids hands and I move my hand to a position, fingers just wide enough for a ball, and catch it. All of it happens in a fraction of a second. The human brain can model the physics of a ball in flight, accurately, and quickly. As the ball…

You mean the brain has evolved over millennia to model the psychics of the world and specialize in catching and throwing things

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

#592

Post by srge » Thu, Nov 14, 2019, 9:39 AM UTC

mkl wrote: post #585

srge wrote: post #558

Earlier quoted context omitted.

John Carmack is a genius in at least 3 different things and probably more: business, tech and expression (written or oral). He can explain very complicated things in such a graceful way. I personally consider anything he does to be newsworthy.

Carmack is amazing at computer graphics and related technologies, but I fail to see evidence of the other two areas. He's never been much of a businessman - he's always focused on the technical side and left the selling/operations to others. I'm not aware of much he's done in terms of educational explanations, and the talks of his I've seen have been rambling and convoluted (but fascinating!).

I would recommend that you listen his recent podcast with Joe Rogan. He's such a great person altogether. Regarding business, I would not say he's only a technical guy. In the podcast he explains how he came to give his first Ferrari as the prize of the first Quake competition. He also explained (somehow to your point) how his rocket company failed.

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

#593

Post by jack_pp » Thu, Nov 14, 2019, 9:39 AM UTC

defterGoose wrote: post #448

nikki93 wrote: post #422

Earlier quoted context omitted.

Yeah I believe in this game / simulated world NPC idea too. To get the kind of complexity we want we either need sensors in the real world or interfacing in a virtual world that humans bring complexity to (probably both -- the humans are part of the sensing technology to start). Things like AlphaZero etc. got good cuz they had a simulatable model of the world (just a chess board + next state function in their case).…

Perhaps the first sentient program will be born in an MMORPG?

https://www.fimfiction.net/story/62074/friendship-is-optimal

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

#594

Post by josh2600 » Thu, Nov 14, 2019, 9:43 AM UTC

yitchelle wrote: post #590

voxl wrote: post #388

Earlier quoted context omitted.

What is actually being argued is the cult worship of a particular engineer. I am against cult worship, so find the fact that Carmack is entering this space simply not news worthy. Yet here it is. Do I wish him the best of luck and hope he cracks the problem? Of course, all the same I would wish that of an upstart PhD student. Yet, the announcements of a brilliant PhD student attending a university to work on AGI is s…

The difference between John and the PhD student is their history of Getting Stuff Done. I imagine that the PhD student list of Stuff Done is not as accomplished as John's. Also the reason why guys like John and Elon are much admired as engineers. They Get Stuff Done. The chance of something remarkable occurring with these folks are a lot higher than it is with the PhD student.

Having seen academics fail to convert their knowledge into production grade systems time and again, I can only say that there are light years of difference between theory and practice.

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

#595

Post by mr_mitm » Thu, Nov 14, 2019, 9:45 AM UTC

rfhjt wrote: post #487

mr_mitm wrote: post #474

Earlier quoted context omitted.

Generating papers is research. I don't understand why you dismiss all papers as pointless.

The need to generate publishable papers means that a researcher can only participate in activity that leads to such a paper. He can't try to work on that idea for 5 years, because if no big papers follow, he's toast /he'd probably lose funding long before that).

When you have tenure, you can work on whatever you want for as long as you want. Nobody works on an idea for five years without publishing anything, though. Progress is made step by step.

Take Albert Einstein as an example, who arguably made one of the largest leap in physics with his theory of general relativity. He never stopped publishing during that time.

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

#596

Post by simonh » Thu, Nov 14, 2019, 9:46 AM UTC

Animats wrote: post #313

This is encouraging. If you're going to work on artificial general intelligence, a reasonable context in which to work on it is game NPCs. They have to operate in a world, interact with others, survive, and accomplish goals. Simulator technology is now good enough that you can do quite realistic worlds. Imagine The Sims, with a lot more internal smarts and real physics, as a base for work. Robotics has the same issue…

I don't think that approach is going to work. For any clearly bounded and delineated task, such as a game, the most optimal, lowest energy and lowest cost solution is not AGI but a custom tuned specialist solver. This is why I don't think Deep Blue or Alphago are paths towards AGI. They are just very highly advanced single-task solvers.

Now Alphago and it's implementation framework are much more sophisticated than Deep Blue. It's actually a framework for making single-task solvers, but that's all. The fact it can make more than one single-task solver doesn't making it general in the sense we mean it in the term AGI. AlphaGo didn't learn the rules of Go. It has no idea what those rules are, it's just been trained through trial and error not to break them. That's not the same thing. It's not approaching chess or Go as an intelligent thinking being, learning the rules and working out their consequences. It's like an image classifier that can identify an apple, but has no idea what an apple is, or even what things are.

To build an AGI we need a way to genuinely model and manipulate objects, concepts and decisions. What's happened in the last few decades is we've skipped past all that hard work, to land on quick solutions to specific problems. That's achieved impressive, valuable results but I don't think it's a path to AGI. We need to go back to the hard problems of “computer models of the fundamental mechanisms of thought.”[0]

[0]https://www.theatlantic.com/magazine/archive/2013/11/the-man...

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

#597

Post by randomsearch » Thu, Nov 14, 2019, 9:49 AM UTC

randomsearch wrote: post #480

carlosdp wrote: post #68

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…

What scientific work has Carmack done?

This was downvoted, so I figured it must be obvious. I googled but as far as I can tell, Carmack is an engineer not a scientist. No formal scientific training, no scientific work.

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

#598

Post by BuckRogers » Thu, Nov 14, 2019, 9:50 AM UTC

Zod666 wrote: post #341

carlosdp wrote: post #68

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…

"With the Quest, he (in my opinion) solidified VR's path to mobile standalones" Yes and I really wished he hadn't. Before he joined oculus they were working on the rift2, he steered them away from that to focus on mobile efforts. I do see the appeal of mobile vr but at the end of the day it is basically an android phone in a vr headset. PCvr is already 2 big steps back in graphical quality from desktop games. Mobile…

I don't see it that way, but rather as the best of both worlds, even if the Index is better. I want it to tether to a PC (or console) as the Quest can, but have hardware onboard so I can take it off by itself and watch Netflix on it. I could never figure out why everything wasn't like the Quest from the start. The Oculus Link and hand tracking (good for video controls without having to use a controller), are what's pushing me over the edge to buy one. In my opinion it's the first VR headset compelling enough to actually purchase. I can recommend it to everyone whether they have a gaming PC or not, and frankly at $400+ for these headsets, people should get a Snapdragon attached for basic gaming and video.

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

#599

Post by nathias » Thu, Nov 14, 2019, 9:51 AM UTC

visarga wrote: post #457

drongoking wrote: post #318

Earlier quoted context omitted.

> (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) This may be his biggest impediment. ML has gotten very far with looking at problems as linear algebraic systems, where optimizing a loss function mathematically yields a good solution to a precisely defined (and well circumscribed) classif…

Humans are not AGI's. We're specialised in human survival, not general intelligence. We're pretty limited in intelligence in many ways, actually, and the environment doesn't support generality. Without a proper challenge an agent would not become super intelligent. The cost of developing such an intelligence would conflict with the need to minimise energy for survival.

We are AGI, the general part comes from language and specialization of brains for language use.

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

#600

Post by vegarab » Thu, Nov 14, 2019, 9:58 AM UTC

tastroder wrote: post #528

vegarab wrote: post #210

Earlier quoted context omitted.

You are talking about AlphaGo. AlphaZero was not given any prior knowledge of the game and is trained exclusively through self-play -- and it outperforms Monte Carlo tree search-based systems such as AlphaGo and Stockfish in chess 100-0 with a fraction of the training time. AlphaZero is also capable of playing Chess, Shogi and Go at a super-super-human.

Not working in this area but the abstract of the AlphaZero paper [0] seems to disagree about your /any prior knowledge/ point: "Starting from random play, and given no domain knowledge except the game rules, AlphaZero achieved within 24 hours a superhuman level of play in the games of chess and shogi (Japanese chess) as well as Go, and convincingly defeated a world-champion program in each case." [0] https://arxiv.or…

This is my point exactly. The model is trained without any prior domain knowledge at all. It only has access to a game world where the constrains in the world is a representation of the game's rules.