I think the best generative AI can do is mediocre. If you train on what humans have done you will only get as good as 50% of it’s best ellements unless you literally copy the best ellement. But generating something “new” should only yield 50% since you are generating from the low quality ellements as well right?
This whole thing about LLMs being able to output something of higher quality than what it ingested is just flawed.
AI such as Deep Blue was able to outperform humans at a specific tasks because humans wrote the algorithms. And they spent a shitload of time and energy to achieve that. It wasn’t “generated” by a machine. It was not vibe coded or easy to do. It was human excellence.
Generative AI is just humans being happy with good enough, contempted with mediocrity.
AI such as Deep Blue was able to outperform humans at a specific tasks because humans wrote the algorithms.
This isn’t true of modern game-playing AI like Alpha-Go or recent Stockfish versions. These learn by playing against themselves over billions of games, and the strategies they develop aren’t guided by human input. These kinds of results aren’t achieved by LLM bots because there’s no equivalent to “winning the game” in a chatbot conversation that can easily be rewarded automatically.
I think the best generative AI can do is mediocre. If you train on what humans have done you will only get as good as 50% of it’s best ellements unless you literally copy the best ellement. But generating something “new” should only yield 50% since you are generating from the low quality ellements as well right?
This whole thing about LLMs being able to output something of higher quality than what it ingested is just flawed.
AI such as Deep Blue was able to outperform humans at a specific tasks because humans wrote the algorithms. And they spent a shitload of time and energy to achieve that. It wasn’t “generated” by a machine. It was not vibe coded or easy to do. It was human excellence.
Generative AI is just humans being happy with good enough, contempted with mediocrity.
This isn’t true of modern game-playing AI like Alpha-Go or recent Stockfish versions. These learn by playing against themselves over billions of games, and the strategies they develop aren’t guided by human input. These kinds of results aren’t achieved by LLM bots because there’s no equivalent to “winning the game” in a chatbot conversation that can easily be rewarded automatically.