Hackworth

@Hackworth@lemmy.world

This profile is from a federated server and may be incomplete. Browse more on the original instance.

Hackworth, (edited )

Sure, thanks for your interest. It’s an incomplete picture, but we can think of LLMs as an abstraction of all the meaningful connections within a dataset to a higher dimensional space - one that can be explored. That alone is an insane accomplishment that is changing some of the pillars of data analysis and knowledge work. But that’s just the contribution of the “Attention is All You Need” paper. Many implementations of modern generative AI combine LLM inference in agentic networks, with GANs, and with rules-based processing. Extracting connections is just one part of one part of a modern AI implementation.

The emergent properties of GPT4 are enough to point toward this exponential curve continuing. Theory of mind (and therefore deception) as well as relational spatial awareness (usually illustrated with stacking problems) developed solely from increasing the parameter count describing the neural network. These were unexpected capabilities. As a result, there is an almost literal arms race on the hardware side to see what other emergent properties exist at higher model sizes. With some poetic license, we’re rending function from form so quickly and effectively that it’s seen by some as freeing and others as a sacrilege.

Some of the most interesting work on why these capabilities emerge and how we might gain some insight (and control) from exploring the mechanisms is being done by Anthropic and by users at Hugging Face. They discovered that when specific neurons in Claude’s net are stimulated, everything it responds with will in some way become about the Golden Gate Bridge, for instance. This sort of probing is perhaps a better route to progress than blindly chasing more size (despite its recent success). But only time will tell. Certainly, Google and MS have had a lot of unforced errors fumbling over themselves to stay in what they think is the race.

Hackworth,
Hackworth,

That’s how I communicate my intention to pay a parking ticket. “Bowing to regulatory pressure”

Hackworth, (edited )

This is another good use case for gAI. Copy/paste the comment into a GPT and tell it to re-write the content at the desired reading or technical level. Then it’s available for follow-up clarification questions.

Hackworth,

I’ve consulted the oracles, and we’re going with 2006.

Hackworth,

At least gay has some positive etymological history as well as negative. F-- only has two meanings, and the vastly more common one is incredibly violent. The only thing I’ve seen remotely close to trying to “take that word back” is maybe Martin in the Simpsons in a throw-away gag about his pure nerdy naivete. And that’s not particularly close.

Hackworth,

I literally read it to mean they’re starting to run out of women and children to kill.

Hackworth,

Y’all run out of Confederate flags AND No Step on Snek flags?

Hackworth,

PS Plus, cause I own like 3 PS games otherwise, Nintendo because I keep forgetting to cancel it, and I’ve never had an X-box, on account of the steam library.

Hackworth,

Equally fake and equally real are equal?

Hackworth,

If one counterfeit dollar buys a soda, does it matter who’s king?

Hackworth,

Peas don’t kill people. Peasple…

Hackworth,

Oh I was just making a dumb joke about calling a 22 a pea. I’m a videographer and don’t own a camera, haha.

Hackworth,

IP isn’t going to withstand generative AI, at least not in a recognizable form. I don’t know what that means for the market.

Hackworth,

What’s the deal with phones auto-correcting to unusual proper nouns? Why is that even an option, if it’s literally a word I’ve never typed/swiped?

Hackworth,

Why all the salt? There’s been AI in video games forever. Is this a line people are drawing?

Hackworth,

Pathfinding was an absolute dumpster fire for a long time. Remember dreading any gameplay where you had to lead an NPC somewhere? Things take time to get better. Gotta start somewhere.

Hackworth,

Why would you train an NPC LLM to know it’s an LLM?

Hackworth,

So if they don’t hallucinate, have object permanence, and are fun to talk to, you’re all good with gen AI NPCs?

Hackworth,

Now that makes sense to me.

Hackworth,

And do you believe paying for more writers will remain a better option for the foreseeable future?

Hackworth,

They’re both A.I. Yes, they work very differently, but the goal is the same - simulate intelligence.

Hackworth,

Which AI is the RC car?

Hackworth,

Is Doom running on the calculator?

Hackworth,

The salt runs deeper than I expected.

Hackworth,

I understand them both well enough to implement them in my projects. I don’t see why people are anything other than excited about the implementation of more capable AI in games. Are these initial implementations garbage? Probably, but that’s just growing pains, So what is it about gen AI that actually bothers people?

Hackworth,

So I work in a creative industry (video production), and have for like three decades. If A.I. can do a lot of the work I do just as well, no part of me wants to continue to do that work. Most of what I get paid for is not “art” in the sense that it expresses some fundamental drive in me. But I do love collaborating with A.I.s to create things that I would’ve never been able to do on my own (and that A.I. would have never been able to do without me). This is where things are going, and I totally grant that greedy corpos doing greedy corpo shit is not to be lauded. But that’s an Ubisoft problem, not a gen AI problem. People are the issue with A.I.

Hackworth,

Musta entered the forbidden numbers. 80085

  • All
  • Subscribed
  • Moderated
  • Favorites
  • fightinggames
  • All magazines