As The Geek Learns AI Engineering Digestfrom As The Geek Learns

Archive

Archive

One dated file per issue. Newer days will stack here.

  1. Issue 45 MiniMax opens M3.1 Flash, and a Studio can just hold it. Fri 23 Oct 2026
  2. Issue 44 Flint lands on the Mac first. Thu 22 Oct 2026
  3. Issue 43 llama.cpp learns to listen and talk. Wed 21 Oct 2026
  4. Issue 42 PyTorch 2.15 brings FP8 and much faster MoE decoding to the Mac. Tue 20 Oct 2026
  5. Issue 41 Reflection opens Beam, a 501B Apache-licensed coding model. Mon 19 Oct 2026
  6. Issue 40 Meta opens Muse Spark Open, and Apple runs it on Mac Studios. Fri 16 Oct 2026
  7. Issue 39 Anthropic opens OpenTelemetry tracing for Claude agents. Thu 15 Oct 2026
  8. Issue 38 Google opens Gemini 4 Argon for developers and enterprises. Wed 14 Oct 2026
  9. Issue 37 Small AI models may be months from self-replicating on IoT. Tue 13 Oct 2026
  10. Issue 36 Claude Managed Agents can write and run multi-agent workflows. Mon 12 Oct 2026
  11. Issue 35 vLLM discloses 11 more security bugs overnight. Fri 09 Oct 2026
  12. Issue 34 Claude Haiku 5.5 lands at a tenth of Haiku 4.5’s price. Thu 08 Oct 2026
  13. Issue 33 Ollama 0.40 makes MLX the Mac default. Wed 07 Oct 2026
  14. Issue 32 llama.cpp v0.6.0 turns a month of tips into GA. Tue 06 Oct 2026
  15. Issue 31 vLLM 0.31 locks down per-request media knobs. Mon 05 Oct 2026
  16. Issue 30 OpenAI disrupts a model-distillation campaign. Fri 02 Oct 2026
  17. Issue 29 AWS opens GPT-6 Astra UltraFast on Bedrock. Thu 01 Oct 2026
  18. Issue 28 AWS opens Bedrock Managed Agents for OpenAI. Wed 30 Sep 2026
  19. Issue 27 Anthropic ships Claude Sonnet 5.5. Tue 29 Sep 2026
  20. Issue 26 NVIDIA opens OpenShell for agent runtime safety. Mon 28 Sep 2026
  21. Issue 25 Ollama 0.40 RC puts MLX first on Apple Silicon. Fri 25 Sep 2026
  22. Issue 24 llama.cpp ships v0.5.0. Thu 24 Sep 2026
  23. Issue 23 Anthropic ships Claude Opus 5.5. Wed 23 Sep 2026
  24. Issue 22 xAI ships Grok 4.7 on the API. Tue 22 Sep 2026
  25. Issue 21 Anthropic opens the lab’s R&D pace metrics. Mon 21 Sep 2026
  26. Issue 20 OpenAI opens a misalignment disclosure framework. Fri 18 Sep 2026
  27. Issue 19 llama.cpp patches RPC RCE in b11000. Thu 17 Sep 2026
  28. Issue 18 Ollama 0.34.1 ships MLX create GA. Wed 16 Sep 2026
  29. Issue 17 Hybrid Apigee speaks MCP. Tue 15 Sep 2026
  30. Issue 16 API Gateway speaks MCP. Mon 14 Sep 2026
  31. Issue 15 Metal fusion gets one table. Fri 11 Sep 2026
  32. Issue 14 DeepSeek-V4.1-Flash ships. Thu 10 Sep 2026
  33. Issue 13 vLLM 0.29.0 goes GA. Wed 9 Sep 2026
  34. Issue 12 mlx-vlm 0.7.0 ships for Apple Silicon VLMs. Tue 8 Sep 2026
  35. Issue 11 Gemini Enterprise Projects go GA. SGLang 0.5.19 ships. Mon 7 Sep 2026
  36. Issue 10 GPT-6 Astra ships. Gemini Enterprise Workflow Builder goes GA. Fri 4 Sep 2026
  37. Issue 9 NVIDIA agrees to acquire Hugging Face for $12.93B. Claude Fable 5.1 ships. Thu 3 Sep 2026
  38. Issue 8 Flash Next MTP no longer dumps state every round. Metal 4.0 tensors light up on M5. Tue 1 Sep 2026
  39. Issue 7 M3 Ultra gets its own Metal fa-vec tables. Flash Next long-context gen got cheaper. Mon 31 Aug 2026
  40. Issue 6 llama.cpp can run Qwen3.8-Flash-Next. Newer Macs get faster Metal. Fri 28 Aug 2026
  41. Issue 5 Reports split on NVIDIA buying Hugging Face, the open-source Hub Thu 27 Aug 2026
  42. Issue 4 vLLM 0.28.0 lands as Gemini Enterprise adds Legal and Financial Services previews Wed 26 Aug 2026
  43. Issue 3 llama.cpp 0.3.0 ships Metal FA as MLX 0.32.2 and AWS ARD drop Tue 25 Aug 2026
  44. Issue 2 Claude computer use leaves beta as llama.cpp wires GLM MTP and SenseNova lands on MLX Mon 24 Aug 2026
  45. Issue 1 Glimmer ships on-device as frontier APIs drop price and EU Article 50 goes live Sat 22 Aug 2026