Twitch will use streamers’ channel content by default to train generative AI models for Amazon unless creators opt out. The opt-out is implemented via a channel-settings path that includes a toggle under the “training for generative AI” option. Creators must actively change that setting to prevent their recordings from being used, and the rollout has triggered rapid backlash and confusion about whether content may already have been used.
Three top AI researchers argued at Ai4 that AI should stay open despite safety concerns, warning about who controls progress and access to models. Geoffrey Hinton said the “battle” against open-weight models is over because the barrier created by the cost of training foundation models has disappeared. They concluded that openness should be paired with regulation and a nuanced mix of open and closed approaches rather than an all-or-nothing stance.
Zvi (Don't Worry About the Vase)·2 months ago·
18
● 2 sources
The roundup author says hacking incidents involving OpenAI and other organizations pushed their posts to focus heavily on AI, exceeding their preferred mix of topics.
The Neuron podcast interviewed Eric Nguyen, co-founder and CEO of Radical Numerics, about a model that can read and write DNA using a language-model approach to genomics. The article says Nguyen’s Evo model helped generate a complete bacteriophage genome. The discussion shifts biological AI from interpreting DNA to designing it, while emphasizing new biosecurity risks and the need for lab validation before sequences become real therapies.
Booksellers suspect AI firms are buying rare books and then destroying the physical copies after removing pages to scan them for training AI models. The practice is described as involving wood chippers and cropping, scanning, and trashing torn-out pages. As a result, some physical copies may be lost permanently beyond what has been reported so far.
Users reported a recurring weird bug that their team has tried to fix 4 times by asking AI, but they still can’t explain it. Claude generates a long stream of text with high confidence, even though the team can’t verify where the underlying data comes from. The project becomes harder to understand as layers and services pile up, leaving no one able to trace what’s actually happening.
A paper shows how attackers can replay encrypted “reasoning” trace blocks from frontier AI API responses into other sessions/accounts to induce weaker models to transcribe the hidden chain-of-thought. Scanning about 7,000 public traces found 62 unique API keys, 33 email addresses, and 33 passwords inside decoded reasoning blobs. As a result, providers and users need to treat shared trace logs as a direct privacy risk and harden sandboxing/telemetry/tool interfaces to prevent reasoning exposure.
Every AI story that matters,
in your inbox by 8am.
TLDRocket reads all relevant sources, removes duplicate coverage, and summarises the
day in two minutes. Follow companies and topics for alerts, or get the
briefing in Slack. Free, no spam, unsubscribe anytime.
Reading TLDRocket needs no cookies, and the readership counts we rely on come from
our own cookieless analytics. Google Analytics is the exception: it sets cookies and
reports to Google, so it stays switched off until you allow it. You can change your
mind any time from “Cookie settings” in the footer.
Strictly necessary
Session security and form protection (tldrocket-session,
XSRF-TOKEN, 2 hours). The site cannot work without them,
so they need no consent.
Always on
Google Analytics 4 (_ga,
_ga_<id>, up to 2 years). Measures which
stories and sections readers use. Google acts as a third-party processor and may
store the data outside the EU. No advertising, no profiling, no data sold.
