Field Notes
Can ChatGPT Make an Editable Diagram? 5 Ways I Tested
Mermaid import, draw.io's own Generate tool, Eraser, BioRender, and one workspace that hands back the .drawio file directly.
I use AI for real work. Then I write down what actually happens.
Field notes on multi-model workflows, long-term memory, and the odd ways models behave when the task is not a toy prompt.
Start here — a short path into how I think
Field Notes
I Tested Gemini 4 Argon’s “Less Hedging” Claim on Real Work
Against Claude Opus 5.5 and GPT-6.1 Sol. One got the math wrong, one forgot the handlebar, one added text nobody asked for.
Field Notes
Custom GPT and Gemini Gem Alternatives After the 2026 Retirement
Both labs ended the walk-into-a-room assistant. A job map for where those rooms actually go, and why persona is not a project.
Field Notes
How to Get 4K Images From ChatGPT (I Tested 5 Methods After Images 2.5)
ChatGPT Images 2.5 can go up to 4K in the API. The ChatGPT app still won't hand it to you. Here's what actually worked for me.
Experiment
How I Validate Business Ideas With AI Before Quitting My Job
I'm a working PM who hasn't picked an idea yet. This is the workspace, the three-round process, and the kill list I use to figure out what I could run alone in 2026.
Field Notes
How to Check if Text Has an AI Watermark (I Tested 5 Methods)
No public text checker exists yet. I ran five workarounds people share for Claude and Gemini, and only one works, for files only.
Workflow
How I Read a 154-Page Threat Report Without Reading All 154 Pages
Search first. Open the page. Summarize last, or never. Long PDFs on a cheap model: index, page jumps, one artifact.
Workflow
ChatGPT Study Mode Forgets Your Syllabus. I Fixed It With One Markdown File.
Study mode is back, but it still can't remember what you learned last week. Here's the PLAN.md method I use to actually finish learning something with AI.
Workflow
Why My AI Prompt Library Died (And What I Keep Instead)
I built a careful personal prompt library. It still died. What travels now is role, residue, and a short ticket.
Field Notes
How to Remove Gemini’s SynthID Text Watermark (I Tested 5 Methods)
Gemini has quietly marked your text since 2024, and it has quirks Claude doesn’t. Five methods, one that actually improved the writing.
Field Notes
Looking for Manus Alternatives After Restore Failed? Wrong Question
Manus data deletion, one-time restore, websites still down, and how to choose a cloud or local agent so your next eight months don't live only in a vendor sandbox.
Field Notes
How to Remove Claude’s AI Text Watermark (I Tested 5 Methods)
Five methods people try. Only one improved the writing instead of just hiding the fingerprint.
Field Notes
Is Hermes Agent Safer Than OpenClaw? Wrong Question.
OpenClaw, Hermes, DeepSeek Harness. Different names, same "local = private" illusion.
Workflow
AI Workspace for Product Managers: Keep Your Work Connected
The problem is not finding another model. It is keeping research, decisions, drafts, and context connected as the work changes.
Workflow
Leaving Poe for a long-term AI workspace
What I actually packed when the knowledge base went away. Not a ranking of bot stores.
Workflow
I Stopped Exporting AI Output Into Docs. Here's What Changed
Export is the handoff now, not where thinking starts. Personal workspace first, company tools second.
Experiment
I Asked 4 AIs to Review Each Other's Work. The Most Brutal Critic Wasn't the One I Expected.
Same messy dilemma, four models, anonymized peer review. Self-preference bias and four structural blind spots.
Work & culture
Vibe Working: When Hiding AI Use at Work Becomes the Job
People hide AI use from their boss, then show their setup to strangers online. Why that split exists, and what it is turning into.
Work & culture
From Vibe Coding to Vibe Working: Why Fast AI Output Isn't Always Safe to Ship
If it is a draft, I can vibe. If it carries my name, I need to audit it.
Privacy
I Didn't Plan To Use My AI Workspace For Health. Now It Holds Our Family Medical Records.
Patterns and appointment prep for family health — not diagnoses. What stays offline.
Latest writing All writing →
Oct 18
Oct 16
Sep 27
Sep 24
Sep 18
Sep 13
Sep 6
Sep 1
Aug 27
Aug 21
Aug 20
Aug 16
Jul 22
Jul 20
Jul 16
Jul 2
From Vibe Coding to Vibe Working: Safe to Ship
Work & culture
Jun 25
I Asked 4 AIs to Review Each Other's Work
Experiment
Jun 6
The AI Homogenization Problem
Experiment
Jun 16
Vibe Working: When Hiding AI Use at Work Becomes the Job
Work & culture
Jun 6
Browse by topic
Occasional notes, no schedule
On AI, work, and the things models get wrong. A few emails a month — only when there is something worth saying.
Newsletter signup coming later. For now, follow via RSS or Medium.