![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
"Ace Attorney developer responds as AI research pits various models against original game with mixed results"
"It's supposed to be simple for a human."
And from one of the comments under the article:
"I mean, I get it and I do get what they're trying to do here... but did any of the AI models laugh at any of the jokes? Did any of the AI models catch any of the references to pop culture or other games? Did any of the AI models fall in love with the characters and want to follow them around as the series progressed and they even made cameos in other games? Do any of the AI models now knowingly yet jokingly shout 'Objection!' with finger extended when they disagree with something?
"To me regardless of how well or how poorly the AI models 'solved the crimes' it seems like they completely missed the point of the games."
This is sort of like that China robots marathon story from a few days ago. A bunch of robots ran in a half-marathon alongside humans. Some of the robots managed to finish the race, some of them never even made it off the starting line, but only one of them barely managed to finish within the "human participation award" time limit of around three hours.
I think when these researchers stop trying to get their LLMs/running robots/whatever to perfectly emulate/surpass/replace humans (outside of maybe novelty shit like the above examples, just for shits and giggles) and focus on the strengths of what they're actually good at and useful for, we, all of us (humans, AI, and robots alike), probably will be better off. (Unless "what they're actually good at and useful for" turns out to be "exterminating humans," of course. ¬_¬)
"It's supposed to be simple for a human."
And from one of the comments under the article:
"I mean, I get it and I do get what they're trying to do here... but did any of the AI models laugh at any of the jokes? Did any of the AI models catch any of the references to pop culture or other games? Did any of the AI models fall in love with the characters and want to follow them around as the series progressed and they even made cameos in other games? Do any of the AI models now knowingly yet jokingly shout 'Objection!' with finger extended when they disagree with something?
"To me regardless of how well or how poorly the AI models 'solved the crimes' it seems like they completely missed the point of the games."
This is sort of like that China robots marathon story from a few days ago. A bunch of robots ran in a half-marathon alongside humans. Some of the robots managed to finish the race, some of them never even made it off the starting line, but only one of them barely managed to finish within the "human participation award" time limit of around three hours.
I think when these researchers stop trying to get their LLMs/running robots/whatever to perfectly emulate/surpass/replace humans (outside of maybe novelty shit like the above examples, just for shits and giggles) and focus on the strengths of what they're actually good at and useful for, we, all of us (humans, AI, and robots alike), probably will be better off. (Unless "what they're actually good at and useful for" turns out to be "exterminating humans," of course. ¬_¬)
