Post #4041552

2026-07-23 11:08 UTC

@willowwren@hcommons.social I'm doubting the OpenAI framing. Their story is murky: 'Our benchmarks run in a highly isolated environment, with network access constrained to the ability to install packages through an internally hosted third-party software that acts as a proxy and cache for package registries.' More likely, imho, is OpenAI likely targeted Hugging Face, got caught, and decided 'went rogue' (becoming normalized as cover fire, as you've noted) sounded better than 'caught stealing'. #ai

Replies (1)

@willowwren@hcommons.social 2026-07-23 12:18

@f800gecko@mastodon.online I think that it's quite likely that a company built on theft would continue a pattern of theft. Their version of events is highly suspect, because they turn an incident of bad consequences into an advertisement that "our tech is so good it's scary," a strategy OpenAI is known for. I'm more concerned here with the fact that the NYT wrote them a scifi cover story where they imagine the AI as little robot hackers from Tron sneaking out of confinement to raid the library because they "figured out" this novel way to solve their problem, independent of their human overlords. The story drips with credulous assumptions of intentionality and all but affirms that the code is a sentient agent. Despite suing NYT (for theft), it's clear that NYT is shilling for the company here. #ai #nytimes

Open ##4041551