Topic
AI Infrastructure - Page 2
Model platforms, cost, latency, and operational trade-offs. Curated tldw.news briefings about ai infrastructure, with practical engineering takeaways from long-form AI and developer-tool videos.
384
breakdowns
Page 2 of 39
AI EngineerThe Cheaper Model That Costs More: Why Custom Evals Matter
Cheaper AI models can be more expensive. Learn why custom evaluations are the only way to right-size model cost without sacrificing performance.
Hugging FaceMistral's 1T Model Challenges Closed AI—But Verify the Claims
Mistral releases its largest open-weight model at 1T parameters, emphasizing enterprise control and cyber defense. The real signal is the compute…
AI EngineerRevolut cut AI coding costs 90%. The catch is GPU infrastructure
Revolut cut AI coding costs 90% by self-hosting open models on 56 GPUs. The real constraint is no longer model quality but time-to-first-token and GPU tuning.
AssemblyAIVoice Agents' Real Bottleneck Is Intelligence, Not Voice Quality
AssemblyAI launches TTS-1 at $15/M chars, betting voice agents need low-cost, scalable TTS. But the real battle is agent intelligence, not voice quality.
Latent SpaceSupabase bets Postgres on agents, not migrations
Supabase is rebuilding Postgres around AI agents: declarative schemas, compute next to data, and SQLite scale-out. Migrations remain off the table.
IBM TechnologyAI audits need independence before access; legacy PHP inventory can't wait
AI auditors with redaction rights are a PR exercise. Plus, why unpatched PHP 8.1 is still a blind spot for most teams.
AI EngineerAgent Auth Is a Gateway Problem, Not a Token Problem
When agents act on behalf of users, who decides which actions are allowed? A talk from Keycard makes the case for gateways over credentials.
Latent SpaceAnthropic's Efficiency Is a Product of Loops and Rust, Not Just a
Anthropic's 5x revenue-per-employee surge reveals that efficiency comes from loops and architecture, not just better models.
AI Engineer300 agents in 60 people: The ungoverned AI agent sprawl
A 60-person company discovered 300 AI agents running daily. The security and cost surprises that followed apply to every team deploying agents today.
Theo - t3․ggInference costs are crashing — don't let bad math scare you off
Theo exposes the math behind the $132M token-spend scare: cache hit pricing and rapid cost declines make agentic coding far cheaper than headlines suggest.
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.