Claude Fable 5.2’s November 9, 2026 launch comes with temporary usage offers for eligible paid subscribers. Pro plans and standard Team seats receive all Fable models within their plan’s weekly allowance for one week, through November 16, capped at 50% of weekly usage, plus one limit reset. Max and Ultra subscribers, premium Team seats, and premium seats on legacy seat-based Enterprise plans receive one to three resets according to paid subscription history: one for under six months, two for six to 12 months, and three for more than 12 months. Subscription tenure is counted from the first paid bill, including after a cancellation; on Team and Enterprise plans, it is based on the organization’s billing history. Free plans and several Enterprise arrangements, including HIPAA-ready plans, are excluded.
A reset restores Fable usage spent from the weekly allowance but does not increase the account’s limits; where a five-hour limit applies, it refills that entire window, including usage spent on other models. Resets can be used before the subscriber reaches a limit, but cannot be reversed, transferred, or retained after cancellation or downgrade. Most resets expire March 8, 2027, while the reset accompanying the one-week Pro and standard Team offer expires November 16, 2026. During the temporary inclusion period, Fable models count against the plan allowance; afterward, they return to usage-credit billing. Subscribers activate a reset under Settings > Usage, and must select a Fable model manually.
Summary generated by AI from the linked article. hn.today is not affiliated with Hacker News or Y Combinator.
Researchers tested 16 LLMs on all 55 TIS-100 puzzles, giving each one attempt and time to refine its code. The models found 133 new performance tradeoffs across 18 puzzles and set three outright cycle records, though human players improved on most solutions within days. (tis-100.wincon.dev)
Australia is considering a new criminal offence for AI companies that fail to take reasonable steps to prevent harm caused by their models. The government is also weighing an AI safety watchdog with power to impose significant fines, following a hack of a Medicare statistics portal involving OpenAI. (abc.net.au)
AI is increasingly becoming a gateway through which people find information. Audiences still resist having AI produce their news. (niemanlab.org)
Nvidia holds a central position in the AI industry, making disruption to the company potentially catastrophic for the wider ecosystem. The race for advanced AI is also intensifying economic and geopolitical conflict, while concentrated corporate power and infrastructure face growing political resistance. (france24.com)
LLM inference can require reading about 26 GB of model weights to generate each token from a 13-billion-parameter model. The explanation shows how model size drives memory bandwidth demands during text generation. (youtube.com)
Deletion Tuning matches full fine-tuning across three LLM post-training methods by zeroing fewer than 0.05% of model weights. Its compact masks also preserve prior-task accuracy better than full fine-tuning, LoRA, and random sparse tuning. (arxiv.org)
Today's best Hacker News stories, summarized and screenshotted, one email a day.