Activity
Mon
Wed
Fri
Sun
Dec
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
What is this?
Less
More
Clief Notes
55.8k
5.0
Free
390 contributions to Clief Notes
I built a careful loop to make AI ideas better... One plain sentence beat it.
I spent two days building a loop to make Claude's ideas better. Check every idea against 5 rules, fix what fails, check again. Then I tested it against one lazy sentence: "Review these ideas and improve them until they are good." The lazy sentence won 92 times out of 96. This is part four of Second Eyes, my test of the advice "get a second opinion from a different AI," and it didn't go the way I planned. Report: https://claude.ai/artifact/Mp1xCfPYov52HccRQZYSaB WHERE WE ARE The kickoff: https://www.skool.com/cliefnotes/second-eyes-does-a-second-ai-actually-catch-more-im-testing-it-properly Part one: I planted 288 mistakes in 12 renovation cost estimates and had Claude (Anthropic's AI, the one I use most), ChatGPT (OpenAI's) and Gemini (Google's) review them. A second company caught no more than a second Claude. https://www.skool.com/cliefnotes/i-expected-a-second-ai-from-another-company-to-catch-more-it-caught-less Part two: my everyday pair, Claude Code (the version of Claude that runs code) with Perplexity (an app that picks a model for you and searches the web) as the second look. Half a point better than asking Claude Code twice. https://www.skool.com/cliefnotes/i-was-sure-switching-ais-was-the-smart-move-300-reviews-later-im-not Part three, which I never posted: brainstorming. Claude plus another app gave about 2 more different ideas in 20 than two fresh Claude chats. But asking the same Claude chat for "10 more, different from the ones above" gave 19.3 out of 20. Not a fair fight, since only the same chat could see the first list. WHAT THIS ONE IS Part four makes it fair and asks two questions.
I built a careful loop to make AI ideas better... One plain sentence beat it.
1 like • 1d
@Clare Martinez internalising it - unconscious competence. That’s the key way “a saved note” (and yes, I know I’m drawing a long bow) is useful enough that you stop looking the original up again.
0 likes • 1d
@Clare Martinez putting the reps in and paying attention. That’s the only practice in any field that I’ve ever found that helps anyone make that jump.
I actually slept during this one - meet Temper
Build video number three is up — and this one exists because of the two before it (and Alex's comments 🤣). Temper is the crew member whose entire job is quality. A topic near and dear to my heart. Some receipts from this build: 106 work orders. Seven staged increments. And — this is the part I'm still slightly smug about — I went to bed at 10pm mid-run and let it keep going, because the evidence said I could. Eleven files were waiting for me in the morning. The system worked *just fine* without me watching and held everything back for my quality checks, which is either the whole thesis or deeply unsettling, and honestly it's both... The voice for this one has been recorded for over a week. I *could* have shipped it sooner, but instead I realised that making videos is apparently a thing I do now for fun, so I took the time to actually build a visual style that's more me for it. And I took notes so I can build the factory and the next one won't take a week. That one's for a future video. Still long. Still detailed records. You might just know how this goes by now. 🤷🏾‍♀️ https://youtu.be/eEtfXYoTzI8 Tell me what you think of the style. Tell me what you think of the build. I can't wait to talk about this one.
0 likes • Aug 9
@Jessy Dominguez good luck with that.
1 like • 1d
@Clare Martinez no, I haven’t, because by the time it hits the review, it’s already been built. Massive waste of time and tokens to leave it that far down the chain to do (and essentially sets you up for a rebuild on purpose).
Should I Reverse my own rule on dashboards for my ICM?
Found a piece of my own client work this week that had been sitting stalled for thirteen days. Fully drafted. Ready to go. It was stuck because one handoff step never happened, and nothing caught it until I stumbled on it by accident. Around the same time, I'd been asking clients a blunt question: what are you actually paying me for? AI can build the website, write the ads, draft the campaign now. Three separate conversations, no coaching between them, same answer: assurance. They're not marketers. They don't know what's happening day to day. They want to know someone trustworthy is watching it. So the plan was to automate proof. AI reads the data, drops five plain bullets into the agenda instead of a generic monthly PDF. Here's what we did. Here's why. Then I looked at why nobody caught the stalled one. I never had a name for "stuck because someone outside the loop hasn't gotten back to me" as its own thing, separate from normal waiting. Both just read as "waiting." If you can't tell the two apart, nothing will flag the difference until you trip over it by accident. If the product is "I'll make sure it's handled right," the proof of that can't be an AI summary nobody double-checks. An unchecked summary just hides the same invisible-work problem one layer deeper. When part of your process depends on a real person outside your system getting back to you, how do you tell "still waiting, that's normal" from "quietly stuck," before you stumble into it by accident? Who's actually watching for that, if you're honest? I had been disciplined about not building a UI for my ICM because I thought it was a waste of time. But now I feel like I need a dashboard for KPIs and statuses so I know what's actually going on under the hood. Anyone else running into this? And by ICM I mean my Operating System or Second Brain.
Should I Reverse my own rule on dashboards for my ICM?
1 like • 25d
You can either do it by thinking through your states and getting a really good query (scripted, tested, don’t let it be created fresh each time) that is surfacing that OR a good view over it. I, personally, have both. Use the query most of the time - but check the view when housekeeping is required. Depends how your brain works… and on the volume and complexity…
A rival could copy the system you paid for. Has it met your numbers?
Nothing I'm about to say here, Jake hasn't already said to you in his own words. You just agreed to pay for a tool, an agency build, or an AI strategy deck described as proprietary, custom, or built specifically for you. Or maybe you're spending time and money in tokens to build it yourself. Before that invoice clears, ask what you actually bought/built. In 1999, the record industry was selling a file. Napster made that file free to copy. The industry's first move was legal, not commercial: the RIAA sued that December, an injunction landed in March 2001, Napster shut itself down that July, and by June 2002 it had filed for bankruptcy. Two and a half years, mostly lawyers, and the file had already gotten away. Software is having the same year right now. In October 2026, Kotaku's Lewis Parker reported on old games, including Halo: Combat Evolved and Grand Theft Auto: Vice City, running in browsers after being pulled apart and rebuilt from the files that ship to players. Parker called it "IP protection whack-a-mole on a scale I don't think we've seen since the advent of pirated DVDs." That's 1999 again: a technology outran what rights holders could control, fast. The limit this time: it works on what ships to a customer's machine. It doesn't reach a company's own servers. https://kotaku.com/we-might-be-cooked-as-these-vibe-coded-web-browser-ports-of-halo-the-simpsons-hit-and-run-and-gta-vice-city-seem-to-work-perfectly-2000743300 Napster duplicated the recording. This makes it cheaper to reconstruct software from what has already been distributed. The techniques and legal questions differ; the economic threat is familiar. Either way, what you paid for stops being rare once a working copy exists outside your control. In McKinsey's 2026 State of AI survey of 1,719 respondents, 32% said their organizations had decided against buying at least one software product or feature because they could build it with AI coding tools instead. That's a purchase avoided, not a subscription canceled. It isn't 32% of software spend disappearing, and it isn't a SaaS collapse. More than three in ten organizations are telling a researcher, in writing, that their build-or-buy answer has shifted.
A rival could copy the system you paid for. Has it met your numbers?
1 like • 26d
@Curtis Hays and… just in case anyone doesn’t get my humour because I haven’t been around here for a while… I have a saying I like to use “if you’re always the smartest person in the room, you should find some better rooms”. I know enough to know there are plenty of smart people here that I have learnt a lot from. About things that matter more than ICM. And it’s a massive compliment from someone whose opinion means a lot. Ngā mihi, Curtis. ❤️
1 like • 26d
@Aurel Babiš I’d recommend a re-read. Sure you guard it - but take a look at what Curtis suggests happens when someone pirates it rather than developing it… It’s a matter of time until it becomes obsolete.
Congratulations Jake
Just wanted to drop a congratulations note to you and your family @Jake Van Clief on the birth of your baby. Wishing you and your family a happy, rested first few weeks. From myself and I know everyone here.
Congratulations Jake
1 like • 28d
Congratulations!!! ❤️
1-10 of 390
Mira Bradshaw
7
5,339 points to level up
@mira-bradshaw-7707
Data tech and product leader. Experience leading data science and data governance.

Active 1d ago
Joined Jul 27, 2026
Powered by