Genie 4

> "which is safetymaxxed to the limit,"

I feel refusal'd just *looking* at this page.

I don't get how Google is so horrible at product. Let me play with this. Otherwise it's DOA.

What an insane set of requirements for a frontier technology and your excitement by it.

You're damn right if you think quota for this model will be reached after two seconds. It's for the investors (and friends and xooglers) to make real use of this, not for you pleb.

It also failed to properly model the physics of knights riding wyverns into battle.

How often do things in the real world not behave exactly as you expected them to?

I feeling I get from this is that this could be the mental model for a robot in the real world, continuously corrected by actual events.

We all have a world model in our heads of what we think is going to happen, we adjust quickly (or not) based on what does. This is pretty incredible stuff.

They were demonstrating steering.

That's the thing about machine learning over time, and the big hole in the whole world model idea. Time and time again we make simulations and train models and we think that loop will work. It never does.

And why not? Because the first model that can actually learn to handle the simulation not only can do the simulation internally too, but it can do it better than out simulation.

It's where (ex-)"AI luminaries" go to die. It's a way to make glacial, slow progress in the correct direction while never actually getting anywhere.

So we get to a point and people keep making complicated simulations of one kind or another ... and then someone (some model) who's never used the simulation, and doesn't need it for model training gets by far the highest score on every benchmark and everyone agrees the simulation was kind of beside the point.

> Sometimes I really wanna skip 10 years into the future and see where all this eventually leads.

Sounds glib, but honestly ... more Ads :/

So, people will be able to completely skip living in the real world, and instead spend their entire time plugged into a simulation, like a vegetable.

Sounds wonderful. May as well skip this and just spend my life hooked up to an IV drip of drugs.

Type of thing he would say.

Llms wont be able to create good games. Consider for a sec that there's like 500 "Great" games ever made, maybe more but not much, in some genres that number is less than 10. There is no training data, and its largely a matter of taste and culture that cant be represented/distilled into vector embeddings.

> movies are going to become video games and video games are going to become something unimaginably better

This is the kind of detached from reality, devoid of humanity statement only a tech bro could come up with.

Why wait? It's become very clear already: https://kotaku.com/the-proliferation-of-copycats-on-steam-ma...

Instead of better it's become unimaginably worse since games can no longer serve as a means of self expression because the second you express anything worthwhile 100 people copy it immediately ... without you. It's like Disney (whose whole business model is copying public domain stories and music and ... but taking out the source and everyone who ever had anything to do with it). Except there's no longer one highly selective company destroying the world's culture one piece at a time, but 10000000 idiots looking to make a quick buck, destroying culture and self-expression like a fire raging in the rainforest.

There will be no more "celeste" games where you join for the platforming but stay for the (politically offensive to all sides) story, because AIs won't discuss the story as it's politically incorrect. Celeste is fundamentally about depression and anxiety, due to being transgender, and is about offending both sides of the argument. First, the main character is transgender, with zero explanation or need for explanation how far that goes, which of course offends conservatives. Second the goal of the main character is to run away from any provided help and professionals (which the director did too) and even under the most extreme of circumstances only reluctantly accept help from a family member that then promptly dies making things "worse".

The Qwen and other Chinese models (and ...) will refuse to discuss being transgender at all, and any western models will shut down any discussion of running away from help hard (because otherwise they get held responsible for the consequences, as has already happened)

(oh and that running away from professional help in psychological issue is actually observed to be a positive thing in almost all cases, in studies. The main way to make psychological problems worse is to let them fester and running away from help is time and time again observed to be far more productive than going to professionals and expecting them to fix things. Because that's an extreme human bias: it can never be (allowed to be) the right reaction to NOT do something about a problem, especially when countless people are paid to "solve" the problem. In other words: if you take people who expressed intent to commit suicide and compare the groups who "get help" and the group that successfully shut down help, less people actually go through with it in the second group, and worse than that, the more intense the help, the better the odds the patient ends up killing themselves. Right to the point that doctors sometimes assist such patients because after 2 years of attempts people live in constant misery and have caused enough medical damage that they will have constant pain or worse the rest of their life. And yet people keep forcing help on patients. Usually in response to an actual suicide where no help was provided. Then more help gets built, and 5 years later we yet again observe ... more suicides. But tell someone with an actual problem to avoid help and you'll be skinned alive for that)

So AI could never make a game like that. Not because it actually can't, or isn't capable, but because there already is a strong political bias, in every model you might use to make it and neither side is going to back down on their policies. The Chinese Communist Party is not suddenly going to change it's mind about transgenderism. And the western models are not suddenly going to openly discuss running away from psychological help, not even as part of a story.

So really, we've just lost these games.

Google is already public, has a larger more normie customer base, many products to integrate AI in with established enterprise accounts in the millions... For some reason theyre not valued at 50:1 revenue... I dont think all that many people are going to be that excited for Anthropic as a business, Its an escape hatch, using the public as exit liquidity while the founders can still make a 100bil.

That entire clip looks like it's straight out of a complete bargain-bin videogame. The weird smoothness of the ride, the pivot in mid air, the wings apparently don't have collision with the poles they fly past.

Definitely wouldn't have been one of my picks to show off my fancy new model.

I was under the impression he has already fly past the catapult.

[delayed]

Audio like: music, ambient-noises etc

That was probably meant to be an em-dash, or have spaces around the hyphen.

As opposed to video-like music...

probably meant "audio like music, ambient noises, and interactive dialogue."

It prompts "gigantic alien tentacles that destroy the world."

I suspect you're reading the base prompt, but the video also demonstrates adding further prompts to change what is happening as the scene progresses.

It was the complete lack of fear or urgency on the part of the characters for me.

Humans interact with their environment by subconsciously simulating it and predicting what will happen in the next few moments. This ability gives us the capacity to work more efficiently and reliably in any environment. These world models will serve a similar purpose to the AI and robotic systems of tomorrow. This is a foundational piece.

You could produce a thousand cheap ads for your product or one expensive real ad.

You could produce immersive VR content specified only by prompts and in which any action is possible.

And their most impactful application will be training smaller models for military drones to detect, target and kill humans. sigh...

Every time I'm about to be excited about AI the cynic in me slaps me back to reality. I imagine the endless possibilities of AI for disabled people, then realize that it's going to leave millions of people disabled instead.

I wonder if that's because the target audience is people directing/producing (commercial) videos, rather than games.

another POV is, what even are these experiences? whom are they for? it would take Google 10 minutes to hire all the best game developers for Android, and cost them like $2m, which is peanuts, and give them fancy Google Research titles. but instead, they're trying to get 100 google PMs, who don't give their kids screentime and who don't play any games the objective of making a product for Bart Simpsons. like these people don't realize that being good parents disqualifies them from making immersive first person shooter, "narrative" experiences.

ironically, the word "realistic" appears next to game footage, not next to real one.

No it doesn't, not even close. You are looking at a video represented as pixels changing on a flat screen. None of it actually exists. You exist, the computer or phone in front of you is composed of mass and energy, and absolutely none of it is algorithmic. World models, as complex as they are, are still just pushing numbers around on paper.

Even if we can show that it is technically possible to model a universe like ours to the degree that our current perception of reality suggests exists, there are three other big factors that would need to be weighed:

Is it feasible? What Dyson sphere level of power would be needed to maintain such a simulation? Is that level of energy available to whomever is running the simulation?

Is there a positive ROI? What does this advanced civilization get out of running such an expensive simulation?

Is it moral? The simulation involves sentient creatures who are capable of deeply contemplating their own existence, capable of feeling rapturous pleasure and indescribable pain and loss, and ostensibly capable of making moral choices that range from heroic to depraved. Doesn't there come a point where the beings in a simulation are so close to the beings in "base reality" that the same moral principles apply? If you press a button that kicks off a simulation where humans can torture and murder each other, behaving essentially identically to humans in "base reality," aren't you in some sense facilitating those deaths?

How does that argument work? I've seen more than a few advances, yet I don't see how any of it provides evidence that we are simulations. This might be one of those things that comes down to the definition of "simulation" I guess.

Not sure if you're joking, but the only such evidence I could think of would be something like finding some way to influence the hardware of our simulation and seeing a glitch in our world

That all sounds right to me, but apparently they can also be used for simply generating videos, or "unlocking the creation of new interactive experiences"... which both seem like competing goals with everything you just mentioned.

I guess we'll find out when they publish it.

The volume of content on YouTube is enormous, I’m kind of surprised they offer auto captions at all given that they have little to no competition in the space.

Typos mean the text came from a human.

They're not proof-reading the blog posts either. The second sentence:

> Genie 4 is natively multimodal, enabling audio-like music, ambient noises, and interactive dialogue to be simultaneously generated with the world.

(Emphasis mine). Like, yeah, I'd sure hope the music is similar to audio.

I wonder if this was a de-LLMese pass gone wrong, and the original was "enabling audio -- like music, ambient noises, and interactive dialogue -- to be simultaneously generated with the world.

Why? Why would a non-deterministic engine be better for a game when it can be instead structured and made deterministic and then more effectively distributed to consumer hardware?

Game industry is still struggling to find a way to incorporate regular LLMs without suffering catastrophic edge cases. A long way to go before these can do more than fancy tech demo, I think.

Because it wasn’t bad enough to need a $5000 5090 to play a game, now you’re going to need a full rack of them!

They can try, my household won't be purchasing those.

We'll be riding the huge wave of indie titles that are incoming.

Because stadia was a real banger right?

Do you have even the slightest idea of the level of expectation that even the most casual gamer has?

This isn’t achieving much beyond prototyping. They’re burning so much money for no reason.

And? Just spin up some open world model and pay for the compute and play your own games?

There are so many options for games now that just sticking with the biggest AAA companies seems like a waste.

Stadia is dead, the reason is playing games over the internet sucks, input lag and serious graphics degradation - this is why I cancelled

Ai generated games suck even more, no skill; no taste, there is no TLC put into them

Then there is the economics of it, cost prohibitive to (re)generate a scene and ship that over internet pipes

And nobody will play that trash, and the world won't change.

Also, lmao, they're already trying to nickel and dime you and sell you microtransactions that took 5 minutes to make for 10 bucks. If you think they're going to pay the costs for running a world model for every single person, lol.