AI Briefing
Recent AI news includes Itaú building generative AI for its customer base, Stack Overflow for Agents tripling signups with a ChatGPT plugin, Benefit Street Partners using Codex for faster question answering, the introduction of the Rosalind Academic Plan for sponsored access to GPT-Rosalind, and the release of Beam, a 501B open model.
Research
Scaling Indic Pretraining: 2.8 Trillion Tokens and an Open Data Generation Pipeline
A new approach to pretraining has achieved 2.8 trillion tokens and an open data generation pipeline for Indic languages.
Why it matters
This matters for operators and executives because it shows the importance of supporting many languages in AI models to improve performance in different applications
Scaling Indic Pretraining: 2.8 Trillion Tokens and an Open Data Generation Pipeline
Summaries of publicly reported developments, with links to originals. Not original reporting. Minor product updates and unverified rumors are excluded.