Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Genie 2: A large-scale foundation world model

deepmind.google

Post reply on HN
Print view View original

381–390 of 436 posts

Re: Genie 2: A large-scale foundation world model

#381

Post by rjrdi38dbbdb »

fwip wrote: post #319
rjrdi38dbbdb wrote: post #303

Earlier quoted context omitted.

How could creativity in AI not get better? Sure, progress will likely not be linear or without challenges, but we already have the human brain as proof that it is possible.

Mountains exist, but that doesn't mean we'll ever build a structure the size of Everest.

If you compare the historical rate of improvements in computing power and algorithms vs rate of improvements in building scale, you'll find one is a whole lot more likely to reach its goal, even if the rate of progress slows significantly.

Re: Genie 2: A large-scale foundation world model

#382

Post by netdevphoenix »

Workaccount2 wrote: post #77
netdevphoenix wrote: post #46

Earlier quoted context omitted.

Progress is not linear. For all we know, in 2027 things will slow down to a virtual halt for the next 30 years. Look at how much big science progressed in the first 20 years of the 19th century/20th century and look how little it has progressed in the first 20 years of this century. We are on the downlow compared to the last centuries and even if you look at crisp or deep learning, they are not as impactful NOW as le…

>Look how little it has progressed in the first 20 years of this century This is naivete on the scale of "Cars were much safer 70 years ago".

Can you please elaborate further? My point is that truly world shattering groundbreaking scientific progress has slowed down significantly this century compared to the previous ones (by comparing just the first 20 years of each century)

Re: Genie 2: A large-scale foundation world model

#383

Post by woctordho »

baalimago wrote: post #374

To me, this is a bit like web3: Can't we already do this? What's the benefit?

We can already program Minecraft. Also we can already program GTA6. But imagine interpolating Minecraft and GTA6 such that all buildings are destroyable. It may be easier to achieve using AI rather than traditional programming

Re: Genie 2: A large-scale foundation world model

#384

Post by halflings »

nine_k wrote: post #347

While cool, this also seems utterly wasteful. Video games offer known "analytical" solutions for the interactions that the model provides as a "statistical approximation", so to say. I would consider a different approach, when the training phase watches games (or video recordings) and refines the formulas that describe its physics, the geometry of the area, the optics, etc. The result would be a "map" that is "playab…

> I cannot be the first person to think about such possibilities

Differentiable Rendering [1] is the closest thing to what you are describing. And yes, people have been working on this for the same reason you outline, it is more data/compute efficient and hence should generalize better.

[1] https://blog.qarnot.com/article/an-overview-of-differentiabl...

But also: > While cool, this also seems utterly wasteful. Video games offer known "analytical" solutions for the interactions that the model provides as a "statistical approximation", so to say.

A bit of the same debate as people calling LLMs a "blurry JPEG of the web" and hence useless.

Yes this is a statistical approximation to an analytical problem... but that's a very reductive framing to what is going on. To find the symbolic/analytical solution here would require to constrain the problem greatly: not all things on the screen have a differentiable representation, for example complex simulations might involve some kind of custom internal loop/simulation.

You waste compute to get a solution that can just be trained on billions of unlabeled (synthetic) examples, and then generalize to previously unseen prompts/environments.

Re: Genie 2: A large-scale foundation world model

#385

Post by shadowmanif »

hbn wrote: post #207
erulabs wrote: post #128

It’s interesting to me that we continue to see such pressure on video and world generation, despite the fact that for years now we’ve gotten games and movies that have beautiful worlds filled with lousy, limited, poorly written stories. Star Wars movies have looked phenomenal for a decade, full of bland stories we’ve all heard a thousand times. Are there any game developers working on infinite story games? I don’t ca…

Creativity is the one area where LLMs are completely unimpressive. They only spit out derivative works of what they’ve been trained on. I’ve never seen an LLM tell a good joke, or an interesting story. It doesn’t know how to subvert expectations, come up with clever twists, etc. they just pump out a refined average of what’s typical.

Claude can make some interesting guitar tabs if you prompt it to transcribe an instrument/music that wouldn't normally be something a rock guitar player would be influenced by.

It is like saying the paint brush and canvas lack creativity. Creativity is not a property of the tool, it is a property of the artist.

We also have a very poor understanding of human creativity from selection bias.

Last weekend I found a book at the library that was Picasso's drawings 1966 to 1968. There must have been 1000-1500 drawings in this book. Many were just half finished scribbles.

The average person seems to believe though that the master artist only produces masterpieces because they didn't bother to look at all the crap.

Re: Genie 2: A large-scale foundation world model

#386

Post by hulium »

erulabs wrote: post #128

It’s interesting to me that we continue to see such pressure on video and world generation, despite the fact that for years now we’ve gotten games and movies that have beautiful worlds filled with lousy, limited, poorly written stories. Star Wars movies have looked phenomenal for a decade, full of bland stories we’ve all heard a thousand times. Are there any game developers working on infinite story games? I don’t ca…

Unexplored comes to my mind, it differs from other games with procedural generation in that it generates a graph for the gameplay first and builds the levels around it. It's not necessarily ground breaking, but it has a special feel to it as objects are placed with purpose.

Re: Genie 2: A large-scale foundation world model

#387

Post by tomaskafka »

erulabs wrote: post #128

It’s interesting to me that we continue to see such pressure on video and world generation, despite the fact that for years now we’ve gotten games and movies that have beautiful worlds filled with lousy, limited, poorly written stories. Star Wars movies have looked phenomenal for a decade, full of bland stories we’ve all heard a thousand times. Are there any game developers working on infinite story games? I don’t ca…

“The inhabitants kept addressing the player and begging him to not shut down their world, so this patch raises punishments for breaking the 4th wall to eternal torment for all the perpetrator’s descendants and that seems to fix the problem for now.”

Re: Genie 2: A large-scale foundation world model

#388

Post by TexanFeller »

empath75 wrote: post #171
erulabs wrote: post #128

It’s interesting to me that we continue to see such pressure on video and world generation, despite the fact that for years now we’ve gotten games and movies that have beautiful worlds filled with lousy, limited, poorly written stories. Star Wars movies have looked phenomenal for a decade, full of bland stories we’ve all heard a thousand times. Are there any game developers working on infinite story games? I don’t ca…

It only looks like a video game because video game footage is plentiful and cheap. Now, imagine training it on thousands of hours of PoV drone footage from Ukraine, and then using that to train autonomous agents.

I’d prefer we trained AI politicians by watching Team America World Police and statements made by George Bush and the neocons after 9/11. Maybe AI politicians could learn from their mistakes and stop involving us in foreign proxy wars in the first place. Especially ones that could escalate to nuclear armageddon.

Re: Genie 2: A large-scale foundation world model

#389

Post by mdtrooper »

beeflet wrote: post #309

These game-video models remind me of the dream-like "Mind Game" game described in Ender's Game , because of how it has to spontaneously come up with a new environment to address player input. The game in that book is also described as an AI.

Yeah, I think the same it.
Post reply on HN