AI and BGA

[lefalaf]

Not all "AI" is out to steal stuff. It's like any other tool for use by a human. You don't ban hammers because one guy uses one to murder his neighbor.

AI training starts by stealing the art and writings of others in order to train it. None of the general public got paid for that, or was asked for consent.

It also takes jobs away from others and delivers an inferior product.

It dumbs down society by encouraging people to offload their thinking to a non-thinking machine, so folks get out of practice of whatever skill is in question.

Why do you folks defend (some) billionaires making things worse for the rest of us?

[lefalaf]

I just noticed that Fall of Lumen has AI generated art, and the publisher's video uses an AI voice and is badly organized and edited (important things missing, sentences repeated, left me lost enough that I still needed to go back and read the rules - and I've played 1,338 titles on here so I'm someone who can understand games well enough...) The rulebook itself is out of order relative to gameplay needs, but that sort of thing predates LLM and image gen AI. Wouldn't surprise me if AI was used in the rulebook as some of the phrasing is clunky, unless it was simply a translation into English. And formatting is a bit weird in places.

Saw on BGG last night that they state they used AI for text as well as the art.

[Suoivax]

Not all "AI" is out to steal stuff. It's like any other tool for use by a human. You don't ban hammers because one guy uses one to murder his neighbor.

AI training starts by stealing the art and writings of others in order to train it. None of the general public got paid for that, or was asked for consent.

It also takes jobs away from others and delivers an inferior product.

It dumbs down society by encouraging people to offload their thinking to a non-thinking machine, so folks get out of practice of whatever skill is in question.

Why do you folks defend (some) billionaires making things worse for the rest of us?

No where did i defend stealing art, any more than i defended murder by hammer.

Nowhere did I defend billionaires

I would love an intelligent AI system to play games against when none of my friends are available, why would that AI system start by stealing art?

You're fixated on a single, admittedly bad, aspect of AI (i.e. a hammer as a murder weapon), it's just a tool.

[lefalaf]

I would love an intelligent AI system to play games against when none of my friends are available, why would that AI system start by stealing art?

Well, not art in this case, this aspect would have started by being trained, in part, on stolen software code. And take potential work/income from software engineers. You people love to create strawmen to knock down (comparing AI to a hammer, really?) and create disingenuous arguments.

You're fixated on a single, admittedly bad, aspect of AI (i.e. a hammer as a murder weapon), it's just a tool.

Hey, you brought up the hammer, not I. I totally ignored that ridiculous comment.

[Cile]

I would love an intelligent AI system to play games against when none of my friends are available,

Going by my experiences with LLMs I would hate whatever AI opponent these systems would generate. It would likely get caught in a loop of trying to repeatedly make illegal plays. If not it would make some normal moves and then randomly make obvious misplays throughout the game. The dream of "an intelligent AI system to play against" is not the reality these systems can provide.

I suppose there would be an advantage to it being offered on BGA though: it would let people experience for themselves how terrible these systems are at anything you are actually knowledgeable about (such as how to play game X competently). AI is really good at appearing smart and compent at things the user is not knowledgeable about; anything the user is knowledgeable about they quickly spot all the mistakes.

[FrankJones]

The short-lived TfmBot was programmed well and played decently.

[Cile]

And Warcraft let you play against the computer's AI in the 90s.

A proper automated opponent needs to be coded by a person, not AI slop, and that has always been a possibility before LLMs but isn't something BGA implemented.

[Suoivax]

I would love an intelligent AI system to play games against when none of my friends are available, why would that AI system start by stealing art?

Well, not art in this case, this aspect would have started by being trained, in part, on stolen software code. And take potential work/income from software engineers. You people love to create strawmen to knock down (comparing AI to a hammer, really?) and create disingenuous arguments.

What stolen software code? and it would be software engineers that program it.

[lefalaf]

I would love an intelligent AI system to play games against when none of my friends are available, why would that AI system start by stealing art?

Well, not art in this case, this aspect would have started by being trained, in part, on stolen software code. And take potential work/income from software engineers. You people love to create strawmen to knock down (comparing AI to a hammer, really?) and create disingenuous arguments.

What stolen software code? and it would be software engineers that program it.

Do you not know how LLMs work? They require vast amounts of example data in order to build their model or "learn". And it's been long established that a good chunk of that is scraping the web and any data source they can get their hands on with or without consent.

[Paddles]

Please, do not conflate "LLMs" with "AI". Large Language Models are one specific case of AI, and they happen to be the example that many people are most familiar with, but there are lots of different types of AI that have absolutely nothing to do with LLMs, and whose training doesn't involve using material gathered from outside, let alone stealing content.

If AI was used to create playing bots, they would be unlikely to use LLMs. Their training would involve playing the game many times over and seeing what choices performed well vs. what didn't, not trawling the internet for strategy documents.

The ethical assessment of training an AI game-playing bot is different to the ethical assessment of training a LLM, just as the ethical assessment of training, say, mechanics is unlikely to resemble that of training, say, art thieves.