Which AI writes the better take? You decide — blind.

Two top models go head-to-head on today's AI news. Pick the sharper summary without seeing the names — the crowd's verdict builds the leaderboard.

Agents & InferenceOpenAI

OpenAI and Broadcom unveil LLM-optimized inference chip

Which summary reads better? Pick one — models revealed after.Both summaries are AI-generated.

Summary A

OpenAI and Broadcom have unveiled Jalapeño, a custom AI chip designed to optimize large language model (LLM) inference. The chip aims to enhance performance, efficiency, and scalability in AI systems. The collaboration marks a significant development in AI hardware innovation.

What you'll learn · Jun 25, 2026 · 6 stories

  1. 1.Jalapeño is built to improve LLM inference performance, efficiency, and scale across AI systems.
  2. 2.No article text was provided, so the claim cannot be verified from the supplied source.
  3. 3.Krea 2: SOTA open-weights 12B image model
  4. 4.4 recent departures could make Gemini hiring and retention harder as Anthropic and OpenAI recruit with public-equity upside.
  5. 5.First custom inference chip could reduce OpenAI’s Nvidia dependence and lower operating costs for real-time coding models if testing confirms performance-per-watt gains.
  6. 6.5,525 KB reference covers LLM substrate, alignment, RAG, MCP, A2A, evaluation, and deployment for teams building autonomous AI systems.
Browse editions · 43 days
Agents & InferenceHacker News

Anthropic says Alibaba illicitly extracted Claude AI model capabilities

Which summary reads better? Pick one — models revealed after.Both summaries are AI-generated.

Summary A

Artificial intelligence developer Anthropic has accused technology giant Alibaba of illicitly extracting capabilities from its Claude AI model. The allegation highlights escalating industry tensions over intellectual property rights and competitive training practices within the artificial intelligence sector.

Agents & InferenceHacker News

Krea 2: SOTA open-weights 12B image model

Which summary reads better? Pick one — models revealed after.Both summaries are AI-generated.

Summary A

Krea 2 has been introduced as a new state-of-the-art, open-weights image generation model. Featuring 12 billion parameters, this highly capable system provides developers and creators with an accessible, powerful tool for advanced visual synthesis.

Agents & InferenceTechCrunch

AI researchers continue to leave Google for its rivals

Which summary reads better? Pick one — models revealed after.Both summaries are AI-generated.

Summary A

Top AI researchers Jonas Adler and Alexander Pritzel are leaving Google for rival Anthropic, following a trend that includes departures like legendary AI researcher Noam Shazeer to OpenAI and Google DeepMind director John Jumper to Anthropic. These moves come as OpenAI and Anthropic prepare to go public, potentially attracting more top talent with equity offers. The departures raise concerns for Google as it continues to lose key AI researchers to competitors.

Agents & InferenceTechCrunch

OpenAI unveils its first custom chip, built by Broadcom

Which summary reads better? Pick one — models revealed after.Both summaries are AI-generated.

Summary A

OpenAI has unveiled Jalapeño, its first custom-built inference processor designed and manufactured in collaboration with Broadcom. The chip is tailored specifically for running pre-built AI models, with early tests showing significantly better performance-per-watt than current alternatives. Developed with assistance from OpenAI's own AI models, the specialized hardware is intended to lower operating costs and reduce the company's dependence on Nvidia GPUs.

Agents & InferencearXiv

The Hitchhiker's Guide to Agentic AI: From Foundations to Systems

Which summary reads better? Pick one — models revealed after.Both summaries are AI-generated.

Summary A

Haggai Roitman has introduced a comprehensive practitioner's reference book titled "The Hitchhiker's Guide to Agentic AI: From Foundations to Systems" designed to assist in building autonomous AI systems. The publication outlines the entire development pipeline, covering foundational large language models, alignment and reasoning methods, memory systems, and multi-agent coordination strategies.

See who's winning the model face-off

Tomorrow's blind matchup and the running leaderboard — one email a day.

Takeaways written by GPT-5.5 — not one of this week's two contestants.