Testing Gemini 4 Hydrogen on Six Coding Projects
AICodeKing tests Gemini 4 Hydrogen on six coding projects and finds a fast worker that needs independent verification.
Your #1 AI video source
AI videos, nothing else.
One-sentence takeaways and concise summaries of important AI videos.
Our Latest AI Videos
The AI Grid Examines Restricted Cybersecurity Models
The AI Grid discusses restricted cybersecurity models and the tradeoffs between defensive access and wider availability.
Ken Wong on Enterprise Agent Context and Accountability
Ken Wong discusses why enterprise agents need organizational context, governance and clear accountability.
Alex Finn Reviews AI Models and Collaborative Agents
Alex Finn shares impressions of a frontier model and discusses collaboration features in AI agent tools.
Videos in our catalogue by publication date
Igor Pogany Tests Claude Whiteboards, Sheets and Data
Igor Pogany tests Claude whiteboards, sheets and data views, finding whiteboards useful for visual planning.
Theo Browne Revisits Gemini 4 Carbon Latency
Theo Browne reexamines Gemini 4 Carbon and argues that latency makes it difficult to use in his coding workflows.
Dwarkesh Patel Reconsiders Continual Learning
Dwarkesh Patel explains two reasons he now thinks AI research could be automated despite current limits on continual learning.
Pat Simmons Uses Fable 5.2 to Build Three Blender Projects
Pat Simmons tests an AI-driven Blender workflow on two architectural recreations and a browser horror game.
Dwarkesh Patel on Recursive AI Improvement and Research Automation
Dwarkesh Patel lays out a speculative case that automated AI research could accelerate future AI progress.
Nick Saraev and Jack Roberts on AI Policy and Office Tools
Nick Saraev and Jack Roberts discuss AI service dependencies, European policy and the expansion of AI productivity tools.
Stefano Corazza on AI Design and Creative Work
Stefano Corazza discusses how AI design tools can support editable creative work and how teams evaluate subjective outputs.
AI Copium on Nvidia’s Transition and AI Research Acceleration
AI Copium reviews Jensen Huang’s departure from Nvidia and industry claims about AI-driven research acceleration.
Nate B Jones on Nvidia Leadership and AI Computing Costs
Nate B Jones considers how Nvidia’s leadership transition could affect AI computing costs and the companies that depend on its hardware.
Testing Gemini 4 Swarms, Video Editing and Benchmarks
A hands-on review examines Gemini 4 model variants, video editing, a Factorio agent swarm and an early model benchmark.
Melvin Johnson on Frontier and Open-Weight AI Models
Melvin Johnson discusses Gemini releases and argues that frontier labs currently lead open-weight models.
AICodeKing Explores Meta's Astryx Design System
AICodeKing demonstrates Meta's Astryx design system and explains how its CLI supplies templates and component documentation to coding agents.
Theo Browne on Choosing AI Coding Models
Theo Browne compares two coding models and explains how he chooses between thorough checking and creative alternatives.
Bijan Bowen Tests Gemini 4 Hydrogen on Creative Tasks
Bijan Bowen tests Gemini 4 Hydrogen on creative coding and 3D tasks, reporting successes, limitations and differences in test setup.
Pat Simmons Demonstrates Agents for a Mock Store Launch
Pat Simmons demonstrates a team of AI agents coordinating tasks for a simulated e-commerce launch and discusses limits of a phone-call experiment.
Igor Pogany Explores Claude Fable 5.2 Creative Examples
Igor Pogany surveys creative examples of Claude Fable 5.2, from interactive games and presentations to video production.