I’ve spent over 15 years on the front lines of healthcare. I’ve personally felt where and how the system breaks, and how it impacts patients, their families, & our communities.
Now, I build AI infrastructure to fix those pain points. 🧵
#DigitalHealth#AI#Innovation#HealthIT
My claude-mem PR had a plot twist: another contributor already fixed the same OpenCode bug 3 weeks earlier.
Instead of dueling PRs, we coordinated — their fix, my bundle-level regression test.
github.com/thedotmack/claude…#OpenSource#TypeScript#OpenCode
Wait Xiaomi has also released the best 9B model in the world?!
This model is based on Qwen-9B...
And you can easily run it even on 8GB of memory!
- MIT license
- Multimodal (image + text)
- Solid for coding/agentic tasks
Perfect to use locally and offline... even on a laptop with unified memory!
claude-mem's opencode plugin failed to load — the bundle leaked non-function exports the loader called as factories. Now only the plugin function ships, tested against the real built bundle. github.com/thedotmack/claude…#OpenSource#TypeScript#OpenCode
it's very real, especially the ~guilt/pressure I had buying into a high ticket subscription had me sucked in to feeling like I had to be all-in on AI or have a bad ROI
Taking a step back for a month or 2 has helped me be more rational and moderate usage
Addictive behavior
Fixed: headroom's .xls loader fabricated digits above 2^53 — your 123456789012345678 showed as ...680. Now renders the honest float, matching the .xlsx loader. PR: github.com/headroomlabs-ai/h…#OpenSource#Python#AI
Those precisely timed, targeted posts around whatever the latest thing is.
Calculated enough that it feels almost violent.
Less conversation, more extraction, optimized for engagement. Wonder how much is genuine vs polished performance 🥱
AI Twitter has pure trainspotting energy:
Refreshing leaderboards for tiny bumps, posting every new SOTA like the 14:37 from Crewe, getting giddy over 3D renderings.
Slightly obsessive but y'all are safe, harmless observers.
Wholesome, really.
#AI#LLM#Benchmarks