DeepL

Headlines · 3

1. 5d
   Mistral CEO warns Europe against quitting frontier AI

   Mistral CEO Arthur Mensch warned that Europe would receive the “leftovers” if it stopped building frontier AI models, while reliance on American or Chinese models could create a “big trade balance problem.” Mensch and DeepL CEO Jarek Kutylowski argue that Europe needs its own models and AI value chain. By contrast, European Commission president Ursula von der Leyen and Legora CEO Max Junestrand favor capturing value through applications. The Commission has begun work on seven AI compute hubs as part of a €30bn project.

   The Next Web · 🔥 0
2. Oct 21
   Nvidia's Huang says telecom base stations could become distributed AI networks

   At Nvidia's GTC event in Berlin on October 21, CEO Jensen Huang said all telecom operators will eventually provide AI compute and that base stations could evolve into distributed AI networks. He said roughly 100,000 sites could form a distributed data-center network of about 100 gigawatts, and that after nearly eight years of development Nvidia and Nokia had completed a first commercial phone call on T-Mobile's network. Huang also repeated his five-layer "cake" model of AI—energy, chips, infrastructure, models and applications—and said building a 1-gigawatt data center currently costs about $50 billion, as reported by 36Kr and IT之家.

   36氪快讯 · 🔥 0
3. Oct 13
   AI translation is fluent now, but human checkers remain essential

   BBC reports that Google says people now translate one trillion words a month with Google Translate, while direct sales of machine translation outside general AI subscriptions were estimated at $1.4bn last year. OKA (Open Knowledge Association), a Swiss non-profit, ran a Wikimedia-funded study covering 119 articles, 10 language pairs and more than 1,000 hours of translator work, finding that 27% of AI-generated text was still corrected before publication, three-quarters of those corrections were errors the AI introduced, and 5 to 6% of errors changed the meaning of the original. Legal and medical translators cite AI rendering "Le demandeur" (the plaintiff) as "the requester" and translating the Greek letter μ as milli rather than micro, a thousandfold dosage error; CNRS researcher François Yvon says the way these systems function makes 100% reliability hard to imagine.

   BBC Technology · 🔥 0

Experience and discussion from the community

Share my DeepL experienceAsk about DeepL