A customer loves the feature you've launched. Then they ask "could it also do X?"
New PMs can hear that as "you haven't done enough".
Steve Jobs said people don't know what they want until you show them.
Sometimes, you've shown them something they love. Now they can see what else is possible.
3
An AI recommendation can change someone's career.
For anyone building this kind of tech, the law is catching up. Connecticut already says using the software is no defense to a discrimination claim. Australia follows on 10 Dec, California and Colorado in Jan 2027.
Sources below.
2
1
37
California: cppa.ca.gov/regulations/ccpa…
Colorado: leg.colorado.gov/bills/sb26-…
Connecticut: cga.ct.gov/2026/ACT/PA/PDF/2…
Australia: oaic.gov.au/privacy/australi…
25
All that power and it still got stuck for an hour. Its browser pane was hidden and a hidden tab only runs its timers once a second
Gave Fable 5.2 my health tracker app to speed up. It was off like a greyhound, almost like it enjoys having something to chase.
In tests my walk now loads in about half a second, down from over 5.
One snag, Claude’s hidden browser spoiling a timing run, wasn’t Fable’s fault.
18
Gave Fable 5.2 my health tracker app to speed up. It was off like a greyhound, almost like it enjoys having something to chase.
In tests my walk now loads in about half a second, down from over 5.
One snag, Claude’s hidden browser spoiling a timing run, wasn’t Fable’s fault.
After reading this I asked Claude to build a "dream" process. Definitely recommend this. The first run caught two of my own rules clashing.
One thing I added was a check for stale memories. Basically anything untouched for 8 weeks. The result was 5 more flagged items.
Learn from this. Just don't hand your agent's memory to a lab.
I've been doing a lot of this with plain .md files for months. One small index the agent always loads. Domain files it opens only when a task needs them. It's in my repo, so I can read it, diff it, fix it, and the same files work in Codex.
Even Claude Code's own memory loads just the first 200 lines, as an index.
Their tests back the rest:
- Better models keep less.
- Telling the model how to write memory hurt. Say what memory is for, then let it judge.
- Check notes regularly. They call it dreaming.
I added a /dream to Mini Harness this week: merge duplicates, delete what the code contradicts, keep the index accurate. Without it, every knowledge system degrades over time.
1
80
Humans and AI seem to share this quirk.
It’s called the adjacent possible. It’s where a step opens options you couldn’t see before.
If you build and constantly improve, partly with your own ideas and partly with AI’s suggestions, you’ll likely have experienced the AI suggesting something that makes a big improvement.
You wonder why it wasn’t just built that way in the first place.
For humans, a door often needs to be opened to see something else that was possible. For AI, it might simply be that the additional context helps. Either way, it looks like the same quirk.
1
2
91
“The adjacent possible” is the complexity scientist Stuart Kauffman’s term. Steven Johnson, who wrote about it in Where Good Ideas Come From, credits him on his newsletter’s About page.
adjacentpossible.substack.co…
41
Using @todoist with Claude? Give it a rule to put a link back to the chat at the top of each task it adds.
Later, one click and I’m back in that chat.
In my setup the desktop app gives Claude the link. In a browser, Claude can’t read it, so paste it in yourself.
How to connect Todoist to Claude or another AI assistant, from Todoist’s help page: todoist.com/help/todoist/tod…
1
99
Old HR tech was about easier and faster.
New HR tech is also about making people processes better.
Of course, it's AI enabling this. Which is somewhat ironic.
1
1
48
And that's not to say, every AI feature being added makes the people side more effective. But it's good to see.
SHRM, the US HR association, says 75% of HR people using AI reported better quality work, at least a little.
Source: SHRM, 5 August 2026. shrm.org/topics-tools/news/a…
That's not a survey about HR tech, but HR people say AI is improving their work.
1
17


