Business
Machine Translation in Action: Duolingo turns to AI translation to expand its most popular courses to all 28 user languages
AI is bringing a massive boost in productivity to Duolingo, maker of the most popular app for learning languages.
Latest
Machine Learning Research
Next-Level DeepSeek-R1: DeepSeek-R1’s update leads all open models and brings it up to date with the latest from Google and OpenAI
DeepSeek updated its groundbreaking DeepSeek-R1 large language model to strike another blow for open-weights performance.
Letters
How AI Fund Is Building AI Builders: AI-enabled coding empowers anyone build software that boosts productivity and makes life easier. Here’s now non-engineers at AI Fund are building apps for fun and profit.
Everyone can benefit by learning to code with AI! At AI Fund, the venture studio I lead, everyone — not just the engineers — can vibe code or use more sophisticated AI-assisted coding techniques.
The Batch Newsletter
DeepSeek-R1 Refreshed, AI’s Energy Conundrum, Agents Get Phished, Machine Translation in Action
The Batch AI News and Insights: Everyone can benefit by learning to code with AI! At AI Fund, the venture studio I lead, everyone — not just the engineers — can vibe code or use more sophisticated AI-assisted coding techniques.
Data Points
How AI models can encourage bad behavior: Edge Gallery deploys mobile models without internet connection
BAGEL, an open ByteDance model that can read and write images and text. Perplexity’s Labs, a new tool to generate research artifacts. A database security failure in Lovable’s coding platform. MIT Technology Review’s new report on AI’s energy footprint.
Data Points
DeepSeek-R1 regains open-weights crown: Researchers find critical vulnerability in GitHub MCP server
NLWeb, an open-source framework to bring AI chat to any website. FLUX.1 Kontext challenges GPT-Image with image generation and editing. LMEval, a new open-source suite for iteratively benchmarking models. Amazon’s new content deal with The New York Times.
The Batch Newsletter
Claude 4 Advances Code Gen, How DeepSeek Built V3 For $5.6m, Google I/O Roundup, O’Reilly Versus OpenAI
The Batch AI News and Insights: I am alarmed by the proposed cuts to U.S. funding for basic research, analyzed here, and the impact this would have for U.S. competitiveness in AI and other areas.
Letters
Cut Research Funding, Weaken the Nation: Open scientific research makes the U.S. stronger. Cutting funding risks national competitiveness and security.
I am alarmed by the proposed cuts to U.S. funding for basic research, analyzed here, and the impact this would have for U.S. competitiveness in AI and other areas.
Tech & Society
Did GPT-4o Train on O’Reilly Books?: Study shows OpenAI’s model can identify verbatim excerpts from paywalled books
A study co-authored by tech-manual publisher Tim O’Reilly shows that OpenAI trained GPT-4o on parts of his company’s books that were not made freely available.
Machine Learning Research
How DeepSeek Did It: Researchers describe training methods and hardware choices for DeepSeek’s V3 and R1 models
DeepSeek made headlines late last year, when it built a state-of-the-art, open-weights large language model at a cost far lower than usual. The upstart developer shared new details about its method.
Business
Google I/O Overdrive: Google’s new AI offerings include Veo 3 video generator, lightweight Gemma 3n, updates to Gemini Pro and Ultra, and more
Google revamped its roster of models, closed and open, and added more AI-powered features to its existing products.
Machine Learning Research
Claude 4 Advances Code Generation: Anthropic debuts new Claude 4 Sonnet and Claude 4 Opus models, featuring top benchmarks in coding
Anthropic continued its tradition of building AI models that raise the bar in coding tasks.