I marathoned it blind, and I wasnt a Halo lore nut or anything, and I love me some space opera, yet it was just... a soap opera in a bad way? I couldnt believe it was a Halo game.
I heard even worse things about 5. And Infinite was such a waste of potential given the massive budget and the giant open world they barely did anything with.
Whats the gameplay plan? Swordplay? Magic? Multiplayer, companions? Roguelike?
The setting? Like, what kind of fantasy? Are there any lore details? Will there be a lore standard?
What kind of graphics? Voxel, retro, modern? Top down? 1st, third person?
Which engine? What language? Scripting?
I get much of this is probably in planning, but I feel the basics are pretty important. Many devs, for instance, may want a Morrowind-like, others may want a top down retro game. Those images look like modern graphics, I suppose, but that raises more questions than it answers.
Its sleazy software, and IMO its defaults provide a really terrible 1st time experience. It convinces 1st time users that locally run LLMs are terrible, and goes out of its way to be incompatible with other stuff.
I think LM Studio would be a much better "1st time" recommendation. Its not perfect either, but its lightyears better, not an active con like ollama, and its smoother to "graduate" from it to croco.cpp or llama.cpp or something.
Well biological structures can be replicated mechanically. So at risk of being pedantic, that's not strictly true.
I view "artificial" as just a term for conscious, purpose construction, whereas "natural" does not involve any conscious intelligence, as in "naturally occurring in nature." But even that is a spectrum, I guess, with genetic manipulation, BCIs, all sorts of intermediate possibilities.
I'd call a constructed, purely biological brain "artificial" because, even if it was grown naturally, it was built with conscious intent.
Well, theres a lot of experimentation with actual neural nets mimicing biology (or literally using biological neurons), and I think that could start to raise the spectre.
World model research could be a significant component of that.
In principle, OP is right; humans are just machines. All I constest is that transformers LLMs are even in the same ballpark, or possibly a path there... they are not.
But I suppose the anguish is on the "liker," as the undecided is the one who gets to ponder their feelings and decide the fate of the date.
...So if I was being selfish, I'd want to be the "undecided." Generous, and I'd want to be the "liker," so the heartache would be on me instead of my date.
If I had knowledge of this cosmic situation, I guess I'd be the liker. I don't want my date to have to sweat while I decide; I want them to have fun. And I'd classify myself as "looser than average," so I wouldnt mind and just enjoy the time with the attractive person.
Thats what I'm saying. They cannot do this, not even close. Even a fruit fly is somewhat adaptive and can respond to novel stimuli, but an LLM effectively cant.
There are tiny experimental models closer to a fruit fly than large LLMs now.
...What about future world models with some adaptive learning loop, sampling stripped out, and so on?
Yeah! That would be fascinating. Then sentience starts to become more of a question.
But thats not what LLMs are.
People like Altman and Modi have grossly overexaggerated what current LLM architectures are capable of, they trained the LLMs to reinforce the illusion as much as they can, and they arent to inclined to change it. Hence all the good researchers have fled to research world models, and are saying transformers LLMs are not a viable path forward.
I dont even know where to begin. Maybe with their state being fixed in time; LLMs do not change. An sci fi analogy might be the "no timers" in orion's arm, who are but a single thought frozen in an infinite loop in time:
This becomes (to me) very obvious if you ever use an LLM in raw completion mode. It very smart, but at the end of the day its no different than a weather prediction model spitting out probabilities for a storm system.
Chat finetuning is meant to get humans to anthropomorphize them by training on human preferences, and the interface further reinforces this. Its all a trick, albeit a very elaborote one.
Will future architectures be closer to "thinking?"
Honestly, it gets real yellow outside in Texas. I imagine it must in Mexico too.