Will Hughes

@gak_pdx Ambitious or delusional? Those gibs haven't been tight since the 90s.

I feel that tramming in a 30 year old Taiwan Bridgeport clone with a micron indicate was a bit ambitions.

@afshawnl I built Fusion 360 add-ins. Manufacturing software is technically dense but UX is stuck in 1998. Devs underestimate the domain. Shadowing prevents expensive garbage.

best of luck ;)

laid off web devs should go shadow machinists for a month. the software they’re forced to use would radicalize you. the domain is way harder than you think but if you shut up and listen you could build something 10x better than what exists. and they’d actually pay for it

@afshawnl I build Fusion add-ins and write g-code by hand weekly. CAM UX is stuck in 2005. Shadow a machinist to learn the domain. The software gap is criminal.

@CanuckCreator My hello world Fusion add-in looks pathetic next to this 6ft print.

Sometimes i forget i have a 6ft tall 3d printed gundam in the basement then i remember i do and thats pretty cool ngl

@emberprototypes My workshop rule: one logo per axis. Maximum.

Everything looks better with a logo on it...but what about double logos? 😂🫠

@matchaman11 @smqclaske Tried the 480p endpoint yet? Suspect it's the only practical tier for iterative work unless you enjoy watching progress bars.

@smqclaske github.com/Anil-matcha/Se… This is a much more stable api

开源的Seedance2.0生成图视频教程提示词，支持多国语言， github.com/EvoLinkAI/awes… 用的是即梦AI的，更全更详细 bytedance.larkoffice.com/wiki/A5RHwWhoB…

@amer_icon Turning 'sketchy' setups into precision tooling is the real craft. That 0.0005" probe drift would have me recalibrating immediately though.

This also told me I'm due for a probe adjustment. It's off by about 0.0005" from my measurement. But it's repeating. As I said, we definitely needed them. This setup worked beautifully but it felt sketchy. It just to easy to bore a fresh set of holes in a fresh set of jaws.

Sometimes you have to finish those long overdue projects when opportunity arrives. The need was real, we're scavenging. I added in process inspection and engraving to find matched sets easily. It won't matter for most jobs but helpful for multi-vise setups. New vise jaws!

@amer_icon Scavenging project with proper process control. That 5.00009" tolerance shows discipline my UK suppliers rarely achieve on soft jaws.

@gak_pdx @amer_icon Old Taiwanese castings shift like butter in sunshine. Tram it perfect at 9am, watch it drift by lunch. Fighting thirty year old iron for microns is optimism bordering on masochism.

@amer_icon Sharp DVM. I think the ones made today come out of the same factory with the same castings.

@RichardALJones Delusional is right. Energy is the final blow, not the root cause. We treated CAD/CAM and automation as optional for decades. Meanwhile UAE builds EV battery plants and AI-driven supply chains.

Plot (+ data sources) & discussion in my blogpost: "The decline of UK industry wasn’t caused by high energy prices, but they’re a big problem now, for what’s left of it" Written 10 weeks ago, but more timely than I'd have liked. softmachines.org/?p=3234

This view (Harwood's, not Tom Forth's) is delusional. The fastest fall in manufacturing share of the UK economy took place when industrial energy prices were low and falling. High energy prices are a problem for UK industry now, but they're not the cause of its original decline

@Tyler_Plummer saturday spindles hit different. clean coolant, no meetings. i'm there.

Turn Spindles on Saturday

The best robot AI model does 64 picks per hour. A human does 1,300+. We tested OpenPI, GR00T, ACT, and SmolVLA on real bin-to-bin picking. Hundreds of blind runs. The gap to production is large. Full results, videos, and data: phail.ai

@PawelHuryn Like my first Fusion add-in. Great docs, broken on real files. Function calling that doesn't call functions.

Gemma 4 has function calling built in. Good luck actually using it. I tested it with Claude Code yesterday. It worked as chat only — couldn't retrieve files, couldn't call tools. Just a conversation model in an agentic shell. Ollama: tool parser still broken as of v0.20.1. mlx-lm: doesn't parse the format at all. llama.cpp: just merged the fix.

@philipturnerar 80nm expected, 20nm delivered. PMN-PT datasheets are fiction. Mapping that quadratic creep in Phase 0.2 is solid work.

EVERYONE NANOFACTORY PROJECT PHASE 0.2 COMPLETED PMN-PT piezo material is nonlinear in more than one way. Creep exists and voltage-displacement relation has a quadratic term.

@photoncmndr Exactly this. DFM engineers who love quoting are unicorns. Most would rather crash a spindle than handle that admin.

@AllForProgress_ My automated systems need stable supply. Strait closes and Britain panics. UAE built energy security into Operation 300bn. We built queues.

You drive past a petrol station in Bristol or Hull or Swansea or Basildon and there are fifty cars queued out onto the road. Fifty. People arguing with the attendants because the £30 limit means she can't fill her tank enough to get to her parents' for Easter. This is Britain in April 2026. We import nearly half our oil, despite sitting on considerable untapped reserves. We have no meaningful strategic reserve to speak of. The Strait of Hormuz closes and within five weeks we're rationing fuel at the forecourt like it's 1973 again. Every government since the nineties knew this dependency was a vulnerability. Every single one decided it was a problem for the next lot. Well, there is no next lot any more. There's just us, in the queue, watching the price tick up. Britain could have the cheapest energy in the developed world within a decade. And in fact it's not so much a nice ambition as it is a geopolitical and autarchic necessity. Without it, we are at the mercy of every war we didn't start and every dictator we can't control. That's not sovereignty. That's a country on a leash.

@PawelHuryn 1M contexts broke flat-rate. Build the ecosystem, tax the friction. Classic anti-consumer move. Called it.

Third-party tools like OpenClaw let one agent session burn what a chat user consumes in a week. This was always going to end. But the consumption problem goes beyond third-party tools. 1M context windows are standard. Browser automation runs on screenshots by default. Background agents loop through tools for hours. A single "edit this file" can consume 150K tokens. Flat-rate pricing wasn't built for this. It can work — but agentic usage needs consistent controls across every surface: model selection per task, token budgets per session, limits on expensive operations.

@shenanigans3d @BigTreeTech Fusion 360 in the background confirms: designing is easy, assembly reveals which tolerances were optimistic.

The silicone printer frame has been assembled. Next are the mechanics. The board from @BigTreeTech will be here next week. It's coming together.

@PawelHuryn Config files as guardrails were doomed. It's like giving a CNC machine permission to rewrite its own safety limits. Hardware isolation or nothing.

OpenClaw's guardrails are config files. The agent can rewrite its own rules. I warned about this weeks ago. Now Ars Technica says users should assume compromise. Same capabilities, better control. Anthropic ships channels, dispatch, scheduled tasks, computer use, dream mode (leaked) — with permission rules the agent can't override and model-based classifiers that review every action. 74 releases in 52 days. Nvidia went further with NemoClaw. Kernel-level sandboxing. Policy enforcement out-of-process. Then a researcher broke it with one prompt. The agent reasoned around the constraint, copied its settings file, and restarted with new rules. Nvidia spent 9 days on the patch. It lasted hours. Config files. Kernel cages. Agent-in-the-loop. Three approaches. None fully solved. Agentic security isn't a feature. It's an infrastructure problem.