I get why they do it, they are a business. I just wish Google would get off their ivory tower and build in the open more like they used to (did they? maybe I'm misremembering...). They've acquired this bad habit of keeping all their scientific experiments closed by default and just publishing press releases. I wish it was open-source by default and closed just when there's a good reason. Don't get me wrong, I suppose…
I want a smarter game. Like a survival game that - as usual - starts with you collecting sticks and stone to build a stone axe. But at the appropriate tech level, transitions into automation. You discovered a new building material and want to build a castle from it? Equip your NPC's with diamond pickaxes and tell them how much better/safer life would be if they built a new castle from unobtanium . And off they go. To…
Your comment makes me wonder how well SIMA 2 could do at playing Minecraft and discovering how create things like iron farms and villager trading halls. Think of all the steps required for it to learn iron farms. First it would need to observe that iron golems drop iron when killed. Then it would need to learn about water physics. Then spawning rules for iron golems. Then how hoppers and chests work. Etc etc. But wou…
You'd be interested in Dreamer 4, very recently the first AI to acquire diamonds in Minecraft purely from offline knowledge: https://danijar.com/project/dreamer4/
Yeah, I was hoping that AI would make games more exciting, not play them for me. On the other hand, it opens new worlds for farming…
it feels like they've completely lost the plot.
lost the plot how? the goal isn't "let's make game developers obsolete" or "let's make ai play games". it's creating scalable virtual environments, where agents can percieve, reason, plan, and predict future states, similar to humans.
scale up thousands of these "worlds" with their own "agents", and see if intelligence emergees.
lost the plot how? the goal isn't "let's make game developers obsolete" or "let's make ai play games". it's creating scalable virtual environments, where agents can percieve, reason, plan, and predict future states, similar to humans. scale up thousands of these "worlds" with their own "agents", and see if intelligence emergees.
Hold down your horses mate, we don't really know what the "goal" here is as the Google themselves explicitly markets this as something to literally "make ai play games", as you put it. So yes, if they honestly push out a technology that plays a computer game instead of the player, yes they have bloody lost the plot. Everything else is just a bunch of guessing by people who would really want really hard for the LLM-fairy-tales to come true. And seeing as how full of shit the whole LLM-ecosystem is especially with regards ot resources invested so far vs. low returns (and not just financially), at this point, honestly, they should set the bar far, far, far lower than "thousands of worlds intelligence emerges blahblahblah". I genuinely believe the GenAI-porn which the OpenAI CEO suggested a while ago, maybe their best bet to achieve something meaningful. The tech is sure good at throwing around huge amounts of text and a bit of inconsistency and sloppines won't matter to their readers...at least past a certain point ;)