Heidelberg AI News

Video · WorldofAI ·

Anthropic SLOWING DOWN!?, Gemini 4 HYDROGEN IS INSANE, OpenAI's NEW Model, & NVIDIA FIRES Jensen!

In short

The video surveys major AI developments, focusing on Gemini 4 Hydrogen’s speed, cost, and agentic coding performance, Anthropic’s safety posture and workplace features, and OpenAI’s unreleased cybersecurity model. It also covers Nvidia leadership and export controversy, new video and computer-use models, and upcoming Microsoft announcements, highlighting rapidly shifting tools and competitive benchmarks for AI builders.

Key points

  • Gemini 4 Hydrogen is reported to be 2.5× faster than earlier Flash models, with four reasoning levels including X-High.
  • Hydrogen ranks seventh overall, costs $1.50 per million input tokens and $7.50 per million output tokens, and shows an 8% hallucination rate in one test.
  • Anthropic says it may slow AI development for safety, while Claude Meetings adds real-time collaboration, notes, documents, and action items to Google Meet.
  • OpenAI’s unreleased cybersecurity model reportedly scores 77% on ExploitGym versus 52.6% for GPT-6 Cyber.
  • Nvidia faces reported leadership and Blackwell GPU export controversies involving Chinese AI labs, though the video notes the evidence is not conclusive.
  • ByteDance’s Seedance 2.5 Flash is described as nearly twice as fast as its flagship model and priced at about $0.15 per second for 720p video.

Written by an AI model from the video’s captions and description.

Anthropic is willing to SLOW DOWN AI development?! Meanwhile, Google just released Gemini 4 Hydrogen, its fastest and cheapest Flash model yet, OpenAI revealed an insanely powerful new cybersecurity model, and NVIDIA is facing a major leadership shakeup involving Jensen Huang!

Watch on YouTube The opening of the article at youtube.com.

In this story

A tool or model opens in the Heidelberg AI Curriculum: what it does, what it costs, how to start. A repo opens on GitHub.

Related stories

All stories

OfficialOpenAI

Our new blog for science

The OpenAI science blog shares new findings, practical methods, and firsthand accounts from researchers using AI to advance scientific discovery.