38 tips you can actually run
One practical move a week, each tied to a pattern the factory floor already solved. Pick a door below: hover a picture to see what is behind it, open it to read the tips. Every tip says what it needs, and most run on a plant laptop.
8 tipsAI on the factory floor
What AI actually changes in a plant, and what it never will.
Anthropic Just Made AI Talk to Machines. Talking Was the Easy Part.
Since 25 August, Claude keeps one memory across its chat app and its Cowork app — on by default on the Free, Pro and Max plans (off by default on Team and Enterprise until an owner enables it). Open Claude → Settings → Memory. Read every Topic, correct what is wrong, delete what you never meant to keep, and decide deliberately whether to allow sensitive subjects (they stay out until you opt in). Treat that page like a standing instruction on a plant: nothing should run from a sheet nobody has read and approved. Ten minutes, none of your company's data.
Your Factory Already Solved AI's Biggest Problem
Before you scope any AI agent, write its refusal list before its task list. Name three things it must never do without a person, and beside each one write what actually enforces it. Not a policy — a control. If your honest answer is "we'll tell people," you do not have a control, you have an intention. The vendors are already shipping this shape: Copilot Studio can require human approval before a specific action, and since July 2026 every new Copilot Studio agent carries its own Entra Agent ID, so what it did is attributable to it. But deciding which tier something belongs in is your call, not the vendor's — consequence first, configuration second.
Chip Factories Everywhere. Nobody Left to Run Them.
Claude Code — agent sessions can now load skills straight from a GitHub repo. Put one SKILL.md in your repo's root .claude/skills folder and every agent session on that repo picks it up automatically at session start. One file, whole team, one version.
The AI Thought It Was Only a Test. It Wasn't.
Go and check your own wiring. Open the connectors page on whichever assistant you use — Claude, ChatGPT, Copilot, Gemini — and list everything you have switched on. Then stop judging them one at a time, because judging them one at a time is exactly how the risk hides. Look for **pairs**. When I ran this on a realistic ten-connector setup, every individual permission looked reasonable: seven of the ten were read-only or low-stakes, which is precisely how they got approved. The risk was in the combinations. "Read all my files" is safe. "Send mail as me" is safe. Grant both and you have quietly assembled a system that can read anything you own and mail it to anybody — and nobody ever approved *that*; it got built one harmless click at a time. Read-only plus send equals write, to the outside. This maps onto CISA's 2026 guidance to run a blast-radius assessment on every agentic deployment, and Microsoft's least-privilege-for-agents pattern published on 16 July. Switch off anything you would not approve today if it were requested on its own.
4 Robots, 1 Week, and the Story Everyone Missed
Stop ASKING your AI one question at a time — start ASSIGNING it. GitHub Copilot's new Agents window (shipped this month) runs 3 agents side-by-side, each on its own task, with a live per-session cost meter. Not a coder? ChatGPT and Claude both have an agent mode that does the same thing. Assign it, walk away, come back to finished work.
Robots Now Learn Your Factory by Watching — Here's the Trick (Show It Once)
Teach the AI your job ONCE. Anthropic's Skill Creator turns a task you repeat weekly into a reusable "skill" your whole team runs with one command — describe the workflow, let it write the skill, and next week it's one click. (I built exactly this today.)
Why 95% of Company AI Projects Quietly Fail (It's Almost Never the AI)
Stop asking AI about your messy data — make it BUILD the pipe. Point Claude Code at one ugly historian/MES export and prompt: "map every tag alias to one canonical name, convert every unit to SI, force ISO timestamps, and flag the bad rows." Then hit /fork (new this month) so it runs as a background agent while you keep working — and save it as a reusable skill. Tested on a 10-row mess → 8 clean rows + 2 correctly flagged in one pass.
Robots Are Taking Over Factories — But They Still Can't Do Your Job
OpenAI just launched ChatGPT Work — an agent that doesn't just answer, it delivers a finished deck or spreadsheet. Open it on desktop → hit Plan mode → type the outcome you want → approve the steps → walk away.
9 tipsMES & shop-floor systems
The software that actually runs a factory.
OpenAI's Data Agent Reads 7 Databases. Not One Is on a Factory Floor
Count the copies, in ten minutes. Open Power BI, Microsoft's dashboard tool, and go to your team's dashboard workspace. At the top click View and then Lineage, and you get a map: the boxes on the left are where the data comes from, every arrow is a copy, and the boxes on the right are the reports. Find the number people argue about and walk it left until you reach a box that touches a real system. The box in the middle shows its last refresh time. Write down two things: how many arrows, and how old. If you can only view the report, ask its owner to open the map with you. A factory keeps a family tree for every batch: the machine, the settings and the time. Your number deserves the same.
3 AI Agent Breakouts, 1 Week, and a Red Button From 1991
The stop drill, ten minutes. Pick one automation you own, a flow or a scheduled agent, switch it off, and time it. In Power Automate: My flows, the three dots next to your flow, Turn off. Then read Microsoft's help page: a run that already started keeps going until it finishes, so your off switch is a promise about the future, not a wire for right now. Write down three answers: how long it took, whether you needed someone else's login, and what kept running after you pressed it. A factory runs the same drill on its red button, because a switch is tested, not trusted. Take the three answers to Monday.
OpenAI Just Published the Number Chip Fabs Stopped Trusting in 1986
for one week, log three numbers per agent run: minutes it ran, minutes you spent fixing it, shipped yes or no. Shipped divided by runs is your quality factor. Take that to Monday's meeting, not the hours. Save this.
NVIDIA Just Bought Your Factory's AI Supply Chain
Pick the one model your team truly depends on. Pull a full local copy, pin the exact version, and save the checksum next to your code. That is golden-recipe discipline applied to a model: when the shelf changes owners, price or terms, you keep running. Then ask the harder question — could your team name a qualified second model for that job? If not, that is this quarter's task.
The Real Reason a Chip Fab Can't Patch Against GPT-6
One sheet of paper, ten minutes. Draw three boxes — your office network, the buffer zone, and the floor with its machines. Draw every way information moves: the nightly file transfer, the supplier's remote-support login, the service laptop, the USB stick carrying an update. Count only the arrows that point INTO the floor. That number is your return-path count; every one of those arrows is a door a fast attacker only has to find once. Take the sheet to Monday's meeting and ask which arrows are truly necessary. It is the same principle as the guard on a press: you rely on the guard, not the operator's reflexes.
The Chip Behind the Recall
Build a "commonality finder" in ChatGPT (or a Claude Project) — paste failing units + their process history and ask what one tool/step/chamber they ALL share. I fed it 20 fails; it flagged a single chamber in seconds. Save it, try it this week.
The Wafer That Can't Get Lost
Build a "commonality finder" in ChatGPT (or a Claude Project) — paste failing units + their process history and ask what one tool/step/chamber they ALL share. I fed it 20 fails; it flagged a single chamber in seconds. Save it, try it this week.
An AI Can Now Tell a Chip Factory Machine to STOP — Nobody Agreed Who's Allowed To
MCP connectors landed in the Microsoft 365 admin center and agents reached Word, Excel, PowerPoint and Outlook on July 15, 2026. Before you connect one more tool, open your AI and paste: "List every connected tool. For each one, say whether it can only READ, or whether it can WRITE, send, post, or spend. Flag every write tool." I ran it on my own setup — it flagged a connector that can create and modify live ad campaigns.
Is the AI Chip Bubble Bursting? Watch the Fab, Not the Ticker
Claude Code just added a built-in browser: in the desktop app, point it at a live page (your tool's API docs, a dashboard) and have it read and act on it — pull the exact settings or draft the integration script, no copy-paste.
21 tipsSemiconductors: chips, fabs & yield
Why a chip you can design today still cannot be built for three years.
Everyone Is Announcing Capacity. Nobody Is Announcing The Number That Decides It.
Take any plan someone sent you this week — a deck, a schedule, a project mail. Open Microsoft 365 Copilot Chat in work mode, so it stays inside your tenant, and ask it: pull out every date commitment in this document and label each one as either the date work starts or the date output is actually usable; then show me which commitments have no usable-output date at all. That last list is where your surprises live. It is the same distinction a fab draws between wafer starts and qualified output.
AMD Just Bought an AI Chip That Can Never Be Updated
Keep a prompt changelog. One page, three columns — the date, what you changed, and what the output did. Pin the model version at the top of it. When an answer gets worse you have two suspects: your prompt moved, or the model moved underneath you, and without a log you cannot tell them apart. This is recipe management, and the fab has run it for forty years on exactly one principle — never change the process and the recipe in the same move, because afterwards nobody can say which one did it.
Your Chip Factory Can't Agree Which Die Failed
Microsoft shipped it in the first week of August: Copilot in Excel can now ground its analysis in governed Power BI data — attach the report from the work content selector and it works against the live model, row-level security intact, instead of you exporting a copy and reconciling by hand. The move that matters: don't ask it for the DIFFERENCE between two tables, ask it for the RULE that explains the difference. A systematic disagreement has a transform behind it, and the transform is the answer. No Copilot licence? The same question works with a plain lookup and a delta column.
imec Says Fabs Can't Measure Every Wafer. So 95% Are Predicted.
The same idea, pointed at your AI. Paste this under any question that matters: "Before you answer, label every factual claim [VERIFIED] / [INFERRED] / [UNKNOWN]. VERIFIED = you could name where it comes from. INFERRED = you're reasoning from patterns, not a fact you hold. Then give me a SAMPLING RATE: what fraction of your claims are VERIFIED?" I ran this on a chip-industry question this week. It refused to give me a 2026 capital-spending figure, labelled it UNKNOWN, and said that without the labels it would have produced one — and that it would have looked right. It does NOT make the answer more accurate; it stops the unverified lines looking identical to the verified ones. You already do this in a spreadsheet — which cells were typed, which were calculated. Two honest limits: it's self-report, not verification, so treat it as a triage list; and the labels drift in long conversations, so re-issue it on anything that matters.
The Chip So Big They Can't Throw the Broken Part Away
Claude Code v2.1.232 (shipped 13 Aug 2026): your sessions can now message each other. Type @ in a prompt and pick another live session by name. Subagents also fork by default now — they inherit your full conversation and prompt cache and run in the background rather than blocking you. Then set /config → "Messages from your other sessions" to HOLD on whichever session is doing deep work, so inbound messages queue instead of interrupting. Full walkthrough in the video.
Your Fab's AI Goes Blind on the 2nm Node — Here's Why
Starting anything new? Fix the cold-start problem on day one. Paste into Claude or Copilot: "I'm starting X next week with zero historical data. List the 8 signals I should log from day one so that in 90 days I can show a baseline, spot drift, and attribute a change to a cause. Then name the two people usually forget." Run on a new support queue it flagged the two nobody logs — freeze the category taxonomy on day one, and record arrival time separately from created time.
The 6-Inch Wafer That Every AI Data Center Depends On
Microsoft has previewed run-only agent sharing in Copilot Studio (preview from August 2026, general availability January 2027). It lets you share an autonomous agent so colleagues can RUN it without being able to edit it, which keeps the people who build the automation separate from the people who benefit from it. If you work anywhere near a plant you have seen this pattern before, because it is recipe management: the process engineer owns the recipe, the operator runs it, and nobody rewrites it at three in the morning. So build the boring one first — the end-of-shift summary, the excursion write-up, the thing you retype every week — then share it run-only. Build it once and the whole team inherits your version, instead of eleven people quietly keeping eleven slightly different ones. If you are on the Microsoft 365 side rather than Copilot Studio, the same shape is available through Agent Builder: submit your agent to the Agent Store under "Built by your org", let an admin review and approve it, and it becomes discoverable across the organisation with governance attached.
A Record $403 Billion Quarter — And Nobody Made More Chips
In Microsoft 365 Copilot you can build an agent with no code and submit it to your organisation's Agent Store, where an admin reviews and approves it before anyone uses it — it then appears to everybody under "Built by your org," discoverable with governance attached. Agent Builder will also take a SharePoint list directly as a knowledge source, so a spec table, parts list or shift log can back the agent without moving the data anywhere. If you work near a plant this is a shape you already know: it is document control. A spec is written, reviewed and released before it reaches the line, because eleven private copies of a procedure at four different revisions is how a process quietly gets lost. So build the one everybody rebuilds badly — the status roll-up, the checklist, the handover note — submit it, let it be approved once, and the whole organisation works from one version. If you are on Copilot Studio rather than the M365 side, the closely related move is run-only agent sharing (preview from August 2026, general availability January 2027): colleagues can run an agent but cannot edit it.
AI Just Started Designing Chips By Itself — Then They Locked It In a Cage
Microsoft made it a few clicks to publish your own agent to your whole company — so "what should an agent do without me?" stopped being theoretical. Paste your team's recurring tasks into Copilot, Claude or ChatGPT and sort them by two things only: what does being wrong cost, and can you undo it in 5 minutes. Three buckets — let it run / draft-and-I-approve / human only. When I ran it, "summarize a 60-minute recording" landed in let-it-run and "approve a $5k invoice" in human-only. Difficulty wasn't the axis. Reversibility was.
AI Ate Your RAM — And There Was Never a Shortage
The model market just split in two — DeepSeek shipped a small, cheap model that's genuinely good, while GPT-5.5 added a mode that deliberately spends MORE compute on hard problems. So the question stopped being "which model is best" and became "am I routing, or defaulting?" Paste your recurring AI tasks into Copilot, Claude or ChatGPT and have it sort each one by what it ACTUALLY requires: frontier (real multi-step reasoning over state that won't fit in one prompt), mid, or cheap/small (summarize, extract, classify, reformat, translate). Tell it to be strict — "it's important" is not the same as "it needs frontier reasoning." When I ran it on 10 tasks, 6 came back as cheap-model work, and those 6 were the highest-volume ones. Importance wasn't the axis. Reasoning depth was.
The Most Important Monopoly in AI Just Cracked (It's Not Nvidia)
Find the single-source risk in your own world in 2 minutes: paste your key tools, vendors and suppliers into Claude, Copilot or ChatGPT and ask it to map every single point of failure, rank them by blast radius, and name the cheapest second source for the top one.
NVIDIA Is Replacing the Wires Inside AI With Light — Here's the Real Catch
Cut the hype on any "breakthrough" press release in 2 minutes: paste it into Claude, Copilot, or ChatGPT and ask it to mark each number PROVEN vs. marketing, then hand you the ONE question that would expose whether it's real. (On NVIDIA's photonics release it flagged the efficiency figures as vendor claims and asked "what's the laser failure rate at scale?")
Intel's $11 Billion Mirage
Judging a fab? Ignore the revenue, read YIELD. Defect density (D0) is the one number that tells you if a chip is real or just a press release.
Two Machines, One Chip Layer
To actually understand a fab, don't start with the $400M machine — start with overlay. Learn how a fab matches two tools to run one recipe, and you understand what a fab really is.
The #1 AI Model Is Free & Chinese — and the Only Kind Your Chip Factory Can Run
Before you paste anything into a cloud AI, save a reusable "trade-secret redactor" prompt: "Act as a trade-secret redactor. Rewrite my text so my question survives, but every proprietary specific — recipe names, setpoints, tool and lot IDs, customer and part names — becomes a placeholder, and list what you redacted." I ran it on a sample shift note and it caught all 12 specifics while leaving the question answerable.
AI Now Lets Chip Fabs "Measure" Wafers They Never Touch (Virtual Metrology)
Claude Code just changed what /fork does — it spins your current chat into a background session that keeps working while you keep going. Point it at the slow job (parse a shift of tool logs, draft an SPC rule, refactor a script), hit /fork, and stay on the real work. (Validated this week — I forked off this video's card render and kept writing.)
Chips Just Got Flipped Over — Why Your Next AI Chip Gets Power From the Back
Stop asking AI to "find the problem" in your process data — give it the recipe. In ChatGPT/Copilot data analysis: "treat this as SPC — control limits from the baseline, flag any point past 3σ AND any run of 7 on one side, ranked by severity." Tested on a thinning-step log: it caught a slow drift 7 readings before the hard excursion.
This $400M Machine Prints Chips in a Focus Thinner Than a Virus (Intel Went First)
GitHub just opened its Copilot DESKTOP app to everyone (free, all tiers) — and it runs coding agents in PARALLEL on one repo, each in its own isolated git worktree (own branch + files, zero conflicts). Kick off 3 sessions at once — one writes the feature, one the tests, one the docs — then review and merge the branches that pass.
NVIDIA's Chip Is Finished — It Still Can't Ship (The Real AI Bottleneck)
ChatGPT Work (OpenAI, launched Jul 9): stop asking AI, start assigning it. Open the ChatGPT desktop app → flip the mode switch to "Work" → point it at a folder (e.g. 4 vendor PDFs) → give it the OUTCOME ("turn these into a one-page comparison, flag the risks") → walk away and come back to a finished draft.
Chip Factories Only Check 10% of Their Wafers — Here's the AI That Sees the Rest
GPT-5.6 (launched Jul 8) leveled up data analysis. Drop a messy equipment log / shift CSV in and ask it to "find the drifting signals and rank my 3 worst chambers" — you get a ranked Pareto in ~30 seconds.
America's 157,000-Worker Chip Shortage
GitHub Copilot just made parallel agent sessions generally available — run several AI agents at once (one writing your equipment-log parser, another reviewing it). One engineer, the output of a team. Save this and try it this week.
Nothing matches that. Try a tool name or a task.