5 ms·
It's been known for a long while that model outputs = data for training another model to copy the original model's behavior, also known as distillation. What I
by cs702 8mo ago
It's been known for a long while that model outputs = data for training another model to copy the original model's behavior, also known as distillation.
What I didn't know is that the three groups mentioned "created over 24,000 fraudulent accounts and generated over 16 million exchanges with Claude, extracting its capabilities to train and improve their own models." There's some irony in that, given that Anthropic and all other established AI shops have been criticized for using copyrighted materials without permission to train their own models. I wouldn't be shocked if we subsequently find out tat every major AI shop has secretly engaged in distillation at some point in the past.
Still, wow, 24,000 accounts. I can't help but wonder, how many other AI shops have surreptitious accounts with other AI shops right now?
- taytus 8mo agoThis reads like AI slop.
- lumost 8mo agoAlso makes you wonder how much of the user growth could just be distillation attempts from one model vendor to another.
- cs702 8mo agoYeah, those 24K accounts are likely high-volume. 16M "exchanges" / 24K accounts = 667 "exchanges" per account. How many tokens per "exchange"? I imagine a lot, because these accounts were likely maxing out on context.
- lejalv 8mo agoSo they did pay to distill a piratic model. More than can be said from Anthropic et al.’s leeching of a substantial proportion of human culture