> ## Content Index > Fetch the complete content index at: https://www.liamdarmody.ai/llms.txt > Use this file to discover other available public pages before exploring further. # AI Experiment 007: Can a Public Fascination Feed Make AI Remember Me? - URL: https://www.liamdarmody.ai/ai-experiment-007/ - Published: 2026-09-25T16:33:35.000Z - Updated: 2026-09-25T16:33:35.000Z - Description: I started publishing a weekly feed of everything that fascinates me, in my own words. The hypothesis: consistent first-person takes, published on a schedule, will strengthen how AI engines describe me. Here's the setup and how I'll measure it. - Author: Liam Darmody - Tags: AI Lab **The short answer:** I built a public weekly feed of the things that fascinate me, each with my own take in my own words, and I'm testing whether publishing it on a schedule makes AI engines describe me more accurately over time. If the "Liam Darmody" entity gets sharper in AI answers, the feed did its job. If not, it didn't. ## What I'm doing There's a line I say all the time: "follow your fascination & feed your curiosity" because that is the approach that has changed my life several times in the last 6 years. So I used my Muse, Shamus, to create a Fascination Feed. It is a weekly public page of what fascinated me that week, with a sentence or two on why, in my words. For years the interesting things I found each week went nowhere. A link dropped in a chat, a screenshot saved to a folder I never open again, a thought I meant to write down and didn't. So I just built the thing I wished existed. The first edition covers September 16 through 22 and has seven entries. It lives [here](https://muse.ai/s/liam-s-links-cxe6akxyxhxxxmrc?ref=liamdarmody.ai), and it's listed on the [Microsites page of my website](https://www.liamdarmody.ai/microsites/), amongst other Muse microsites I've been creating. ![Screenshot 2026-09-25 at 12.15.50 PM.png](https://storage.ghost.io/c/4c/2e/4c2ec722-2bd5-4a69-ad93-121c6c1c3e73/content/images/2026/09/Screenshot-2026-09-25-at-12.15.50---PM.png) ## How the feed works Every week I publish one edition with a handful of entries. Each entry has a title, the source, a link, and a short "why it fascinated me" take that I often dictate to Muse via Wispr Flow while walking my dog (Lucky DarmoDoodle). Entries arrive in two ways: 1. During the week I drop links in a chat with a one-liner about what grabbed me. 2. I react to content surfaced to me through my Muse Feed, and then once a week Shamus (my Muse) provides the list of things I reacted to and asks me what resonated with me about each one. ![Screenshot 2026-09-25 at 12.28.46 PM.png](https://storage.ghost.io/c/4c/2e/4c2ec722-2bd5-4a69-ad93-121c6c1c3e73/content/images/2026/09/Screenshot-2026-09-25-at-12.28.46---PM.png) My answers become the takes. ![Screenshot 2026-09-25 at 12.32.42 PM.png](https://storage.ghost.io/c/4c/2e/4c2ec722-2bd5-4a69-ad93-121c6c1c3e73/content/images/2026/09/Screenshot-2026-09-25-at-12.32.42---PM.png) And very page carries the same identity footer and Person schema that sits on everything I publish: who I am, what I do, where to verify it so that LLMs can clearly attribute this to me. The feed is a microsite, not a social post, so it can be crawled, quoted, and cited. ![Screenshot 2026-09-25 at 12.21.13 PM.png](https://storage.ghost.io/c/4c/2e/4c2ec722-2bd5-4a69-ad93-121c6c1c3e73/content/images/2026/09/Screenshot-2026-09-25-at-12.21.13---PM.png) ## The hypothesis Publishing a weekly public feed of consistent first-person takes will strengthen the "Liam Darmody" entity in AI answers over time: more accurate descriptions of who I am, more mentions of the feed itself, more citations tied to my name. The mechanism I'm betting on is boring, and that's the point. Entity clarity comes from repetition of consistent, attributable claims. Every edition adds another week of "Liam Darmody said this about that," all in one crawlable place, all pointing at the same identity. ## The caveats Four caveats, because this one has some real structural wrinkles. 1. The feed lives on a public muse.ai link, not my domain. Citation and domain credit accrue to muse.ai, while the entity credit, my name showing up in answers, stays with me. I'm testing whether the entity half works on its own. If engines start describing me better but never cite the feed's domain, that's still a result, just a split one. 2. This is a sample size of one person. It tells me what happens for Liam Darmody. It does not tell me what happens for you. The worth-borrowing section is about what you can use from it, not a promise it will work the same for you. 3. AI engines re-crawl on their own schedules. A weekly publishing cadence means nothing if crawlers visit monthly. The cadence is part of what's being tested: does showing up every week move the needle faster than a static page would? 4. The cadence is also the risk. This experiment only works if I keep publishing. Historically, I've struggled to publish longer form content (like this newsletter) consistently. I'm so busy and ADHD that sitting down to write & synthesize all of the things that I found interesting that week is a herculean effort for me. Using AI to make it easier for me has been really helpful. 5. It started as the Curiosity Feed and I renamed it the Fascination Feed before this experiment published, because the tagline was already the name. If the name changes again mid-experiment, I'll note exactly when, because a rename is a variable. ## Worth borrowing You don't need Muse or my feed for this. Try this: 1. Publish your takes somewhere crawlable, in your own words. A social post that sinks in an algorithmic feed is not an entity asset. A page is. 2. Attach the same identity block to everything. Name, role, location, what you're known for, worded the same way every time. Repetition is the mechanism. 3. Pick a cadence you can actually keep, and write it down. Weekly, monthly, whatever. The experiment is the schedule as much as the content. 4. Check your work the way I'm checking mine: ask four engines who you are, write down what they say, keep publishing, and ask again in a month. If you try it, I'd love to hear what moves. If you try it out, let me know what you think! Hi5 & have a great weekend! LD 🌶️ *Sources: the Fascination Feed itself ([https://muse.ai/s/liam-s-links-cxe6akxyxhxxxmrc](https://muse.ai/s/liam-s-links-cxe6akxyxhxxxmrc?ref=liamdarmody.ai)); measurement log to be published alongside the first monthly results. Baseline check runs before propagation; first comparison one month later.*