obvious interpretations + sheet champion

· Zi Wang · 1 min read

Z / Runner 🏃🏻‍♂️

🏃🏻‍♂️ deeper down the design/proto phase of ai(journal); have an objectivity engine, need to eliminate bias: sample size of 1; false consensus effect. Teasing out the tension between objective [features + tech] vs. subjective [opinions, preferences, beliefs, habits] hard & will get harder. Longevity space new, mired in opinions… diff from previous startup experiences. N of 1, availability heuristic maybe hurting the exploration process, over-index on personal recall vs. generalizable research data. (need to balance the 2). Reduce reliance on "would I use this …"

  • Hate market research, but needed if serious about newco. Otherwise, we are just eating "opinion onions". Have to be more disciplined (knowing the pain vs. knowing the remedy).

  • 🏃🏻‍♂️ need touch grass (sample outside of peer groups), sparing w/ medical professionals; reach 100 "No" w/ skeptics.

  • Post prototype: need to formalize and have a type 1 vs. type 2 decision.

  • 🏃🏻‍♂️ type 1 = irreversible, high consequence, costly to fix (1-way door).

  • Does 90-9-1 network dynamic [lurkers > contributors > creators] apply in the longevity space? (if yes, then how to leverage; if not, how to create).

  • For ai(sheet), should ask "have you ever paid to get automated + continuous + evidence based interpretation"?

  • 🏃🏻‍♂️ use this mental model (notebooklm vs. gemini deep research), analysis of targeted domain vs. exploration + brainstorming.

  • 🏃🏻‍♂️ i have gone thru this process w/ supp.co but my need is to source the evidence directly from pubmed & ground the research for my own biomarkers. Hence I think ai(sheet) maybe a pathway to realize this concept. As for dosage & brand, all Sports Research for Q4… avoiding cross-branding matrix for now.

  • For ai(journal), what are people currently using to record their longevity (fitness, sleep, diet, mental) data?".

🏃🏻‍♂️ ai(sheet) & ai(journal), stating the obvious/obvious (obv^2), value props and usage are distinctly different: JTBD for ai(sheet), data, interpretation (monthly/quaraterly); ai(journal), first-party data (daily/weekly)?

🏃🏻‍♂️ shamir had a good suggestion (per yc) of using a pre-mortem to tease out product truth.

  • 🏃🏻‍♂️ i agree excel is an amazing tool, global finance would die w/o it… but how do you make it accessible for longevity.

Stephen / Basketball 🏀

  • 🏀 what's more than gemini deep research? bottom up instead – convince n++ close to us?

  • 🏀 not grass, but rubber meets the road.

  • 🏀 slow vs fast thinking?

  • 🏀 bezos! but all type 2 until vc raise or token launch.

  • 🏀 nah, per "play mode" below.

  • 🏀 automated / continuous / evidence are the easy parts, while interpretation is the hard one. should people just follow simple rules / high authority on fitness instead? how to show incremental value of sheets over plain gemini? but, let's not be stuck and can start with consumerlabs – z, search and follow up on "fish oil" below.

  • 🏀 z, for your 14 pills: post the brands + packages, your exact doses + timing? let's get to the bottom – and expose the scam of expensive pees.

  • 🏀 i did go through these as you linked them before; there are no ways for any human experts or ai models to give definitive answers – unless the prompt is hyper specific.

  • 🏀 focus on ONE and go as deep as possible. how recording mental (and from its ai coaching) may improve my life?

  • 🏀 journaling is obviously valuable to you (and billions but not to me or aaron. what can i try the next few days to see good value?

  • 🏀 ah, "create (immediate) value for users (today)" – per "8.8 years" below. solve a 'hair on fire' problem; launch now. find the 90/10 solution. it's better to have 1(00) people who love you than a million who kind of like you. make something people want.

🏀 "debate if, what, and how much fish oil supplements to take - among peter attia, rhonda patrick, david sinclair, andrew huberman, gabrieele lyon, thomas delauer, bryan johnson, brad stanfiled, robter horst, michael doyle, consumerlab.com."

  • (excel) world champion: "There's a super active community, we've got this giant Whatsapp group for the different things people want to talk about.. Excel is used in solving a maze, scoring poker hands, or even sorting Kings and Queens into the battles in which they fought.".
  • 🏀 the ui (accessible) and the topic (longevity) are easy, but the ai?