AI Coding Agents: Cursor, Devin, and the Future of Software Engineering
This week we dive deep into the AI coding revolution. Cursor ships multi-file editing, Cognition's Devin hits production milestones, and Claude Code gets autonomous mode. Are software engineers b
Scott WuAman Sanger
codingCursorDevin
Google DeepMind's Gemini Ultra 2.0: Multimodal AI Gets Serious
Google releases Gemini Ultra 2.0 with native audio, video, and code generation in a single model. We test it against GPT-5 and Claude 4 across real-world multimodal tasks. Plus: Google's new AI S
Logan KilpatrickJeff Dean
GeminiGoogleDeepMind
Claude 4 vs GPT-5: The Benchmark Battle Nobody Expected
Anthropic drops Claude 4 just one week after GPT-5 launch. We do a head-to-head comparison on coding, reasoning, creative writing, and real-world tasks. The results might surprise you.
Dario AmodeiSimon Willison
Claude 4AnthropicGPT-5
ThursdAI - Jan 15 - Agent Skills Deep Dive, GPT 5.2 Codex Builds a Browser, Claude Cowork for the Masses
Deep dive into agent skill architectures. GPT 5.2 Codex demonstrates building a full browser from scratch. Claude Cowork opens to general availability. The era of personalized AI is here.
Nisten TahirajLDJRyan CarsonSwyx
agentsOpenAIAnthropic