3 ms·
Do we know the "fraudulent " requests really came from moonshot engineers and was not QA team running a ton of benchmarks against other models? I feel distill
by throwa356262 5mo ago
Do we know the "fraudulent " requests really came from moonshot engineers and was not QA team running a ton of benchmarks against other models?
I feel distilling something as big as Opus would require many many more samples, but I dont really know much about this subject
- maxdo 5mo agosure, sounds like QA lol Scale: Over 3.4 million exchanges The operation targeted: Agentic reasoning and tool use Coding and data analysis Computer-use agent development Computer vision Moonshot (Kimi models) employed hundreds of fraudulent accounts spanning multiple access pathways. Varied account types made the campaign harder to detect as a coordinated operation. We attributed the campaign through request metadata, which matched the public profiles of senior Moonshot staff. In a later phase, Moonshot used a more targeted approach, attempting to extract and reconstruct Claude’s reasoning traces.
- ta20240528 5mo agoAnd when you here unsubstantiated rumours* that say Anthropic has been sending exchanges to say Alibaba's Qwen, will you als oconclude the same about the entire US AI industry? I doubt it. * publish the logs.
- ifwinterco 5mo agoEven if it's true, it's not like US AI companies can complain, given their entire business is based on ripping off text without attribution
- maxdo 5mo agochinese ai is not doing the same? or they don't parse? they do except they also send thousands of sex-spies to do espionage of this kind on the scale.
- ifwinterco 5mo agoOf course they’re also doing this, my point is this is a grubby business where ethics went out of the window a long time ago. If you’re playing this game in 2026 you know the rules - anything goes
- ta20240528 5mo ago"they also send thousands of sex-spies" Could they send one (or two) my way?