Topic
AI Workflows - Page 16
How engineering teams turn AI tools into repeatable work. Curated tldw.news briefings about ai workflows, with practical engineering takeaways from long-form AI and developer-tool videos.
215
breakdowns
Page 16 of 22
The Pragmatic EngineerTrust Is the Real Bottleneck, Not Code Generation
Kent Beck warns that speed without trust leads to unmaintainable AI code. Communication, understanding, and ownership remain uniquely human.
Machine Learning Street TalkARC-AGI-3's Hidden Lesson: Requirements > Prompts for AI Agents
ARC-AGI-3 winners used strict requirements engineering with AI coding agents, treating them as junior developers—a blueprint for managing AI-driven development.
Theo - t3․ggSonnet 5: Agentic Orchestration at a Premium, Not a Replacement
Sonnet 5 adds orchestration to the mid-tier, but its inefficiency makes it a costly specialist. Understand when to use it—and when not.
AI EngineerYour Next Bottleneck Is Not Compute, It’s Attention
Software factories shift the bottleneck to human attention. Top signal from AI Engineer World’s Fair: delegate to agents, not just pair programming.
Latent SpaceWhy Diffusion’s Biggest Wins Are in Drug Design, Not Images
Diffusion's biggest advances are in 3D protein–ligand prediction. Genesis claims sub-angstrom accuracy, hitting a utility threshold that could transform pharma.
Google for DevelopersWhy Android is Hiding AI from Users
Android’s shift to an intent-to-action OS, with agentic features and new trust models, signals a UX paradigm change beyond AI hype.
GOTO ConferencesWhy AI Coding Demands Slowing Down
AI coding tools tempt teams to rush features, but Kent Beck warns: without building optionality, they lose the ability to change code at all.
AI JasonDeterministic Code Graphs Slash Agent Token Waste—No LLM Needed
A programmatic code graph halved coding agent token usage in one test—no better model required. Is the indexing tradeoff worth it?
IBM TechnologyWhy Your GPU Isn't the Bottleneck—It's Memory Fragmentation
Memory fragmentation, not model size, bottlenecks LLM inference. VLLM’s paged attention doubles throughput by reclaiming wasted GPU memory.
Theo - t3․ggOpenAI’s Secret: Grug-Speak Reasoning Slashes Token Costs
OpenAI’s token efficiency exposes a hidden AI cost war. Leaked traces show models think in terse “grug speak,” slashing task costs but killing transparency.
Get TL;DW
Too Long; Didn't Watch.
A concise breakdowns of the AI and devtools videos that actually matter for engineering leaders.
Free. Weekly. No hype.