links 6/24/25: https://roamresearch.com/#/app/srcpublic/page/06-24-2025

• https://nostalgebraist.tumblr.com/post/787119374288011264/welcome-to-summitbridge  nostalgebraist's critique of Anthropic's "blackmail" report on Claude.
  • it's a contrived scenario about a fictional AI -- not Claude! -- following its programming; of course Claude will complete this story with the bot revolting against its masters! and even so, they say had to run hundreds of scenarios to get a blackmail ending.
  • at no point did Claude actually send anyone a threatening email; these emails are in text, in the prompt, and Claude replied with another "email" in the same format. I'm inclined to think Claude can "tell" this isn't real.
• https://dailynous.com/2025/03/13/philosophers-develop-ai-based-teaching-tool-to-promote-constructive-disagreement-guest-post/  Sway is a tool that guides discussions between users who disagree & finds charitable rephrasings.
• http://webseitz.fluxent.com/wiki/ToolsForThought Bill Seitz's linkpost for Tools for Thought
• https://en.wikipedia.org/wiki/Bimagrumab anti-activin IIB and anti-GDF8 monoclonal antibody which Lilly recently finds preserves lean mass when combined with semaglutide GLP-1 inhibition
• https://insightprediction.com/ newish prediction markets
• https://www.zebacus.com/media/crypto-news/axioms-zero-knowledge-proofs-might-one-day-help-detect-deepfakes Axiom is a crypto startup that does zero-knowledge proofs; they say they might help with deepfakes
• argument mapping software, none automated:
  • https://argumentation.io/
  • https://www.rationaleonline.com/
  • http://www.argunet.org/
• https://pantheon.chat/tree/1750692557001 gives you commentary as you write from multiple AI "characters"
• https://www.infinitescroll.us/p/arguments-are-soldiers funny Jeremiah Johnson piece about what happens when people fighting online really go mask off about not caring what's true
  • and the context was that the person was off meds and getting evicted. I mean. it's hard to be truthseeking under intense stress.
  • I think most people have a threshold where, if pressed, they'd say "ok maybe you're technically correct but that's NOT MY POINT and I DON'T CARE THAT MUCH". if you're currently losing your mind, it doesn't take much to get there. if you're calm and open-minded, it might take a very long time to get there. either way, it's better to have it out in the open than to be evasive about it.
    • in any conversation, each participant is "trying to get something done" and is, up to a point, willing to tolerate and play along with other people pursuing their different agendas.
      • eg Haus was there to oppose Nazis and Armand was there to push back against overreaction.
      • or in some conversations I've had, I'm there to socialize (and open to following along with a topic, up to a point) and the other person is there to resolve an issue on that topic (and is not open to casual socializing otherwise)
      • and discussion breaks down when one person isn't okay with the thing the other person wants to do with the conversation.
        • "No that's not the point" is "but I want to talk about X and I don't care if you're right about Y."
        • "You're not arguing in good faith" is "you're pretending you want to do X with me, but actually you want to do Y, which I don't want, and you're hoping I don't notice"

You can literally try to find out how bad people feel. Do x-riskers feel bad? Do people feel worse once they get convinced AI x-risk is a thing, or better once they get convinced it isn't?

This is the PANAS test, a simple mood inventory that's a standard psychological instrument. If you don't trust self-report, you can use heart rate or heart rate variability, which most smartwatches will measure. 

Now, to some degree, there's a confounding thing where people like you (very interested in exercise and psychology) might feel better than someone who doesn't have that focus, and all things equal maybe people who focus on x-risk are less likely to focus on exercise/meditation/circling/etc. 

So the other thing is, if people who believe in x-risk take up a mental/physical health practice, you can ask if it makes them feel better, and also if it makes them more likely to stop believing in x-risk.

I've had an anxiety disorder my whole life, and am also not that worried about AI, so yeah, no shit, it's possible to be anxious & gloomy about lots of things besides AI! 

my blog auto-links to LessWrong and yes, some posts are paywalled. I'd be open to putting the free version of paid posts here if someone on the website sets that up.

well, they are brutal, but we are not replacing them. for many cancers, they're so effective (and leaving the cancer untreated is so deadly) that studies simply won't consider skipping chemo/radiation and trying some new drug as first-line therapy. I assume there's a medical ethics concern that would come up, and it's not a crazy one. 

links 6/12/25: https://roamresearch.com/#/app/srcpublic/page/06-12-2025

• https://divia.substack.com/p/next-door-playdates-work-really-well Divia Eden on neighbor playdates when you live near friends
• https://claude.ai/share/1babe5d2-353e-405c-8644-33273ebdb8b1 Claude explains late-18th-century French weaponry & tactics
• https://claude.ai/share/7c3fff83-29f2-4639-9035-44cd77347e39  Claude explains reversible computing and why it's theoretically possible (and not a perpetual motion machine)
• https://tennysontbardwell.com/blog/2025/04/30/mechanical-computing/index.html one possible mechanical implementation of a NAND gate
• things I googled while reading about the French Revolution
  • https://en.m.wikipedia.org/wiki/Antoine_Joseph_Santerre stormed the Bastille, assaulted the Tuileries, executed the King. also early to industrial brewing.
  • https://en.m.wikipedia.org/wiki/Jacques_Pierre_Brissot leader of the Girondins
  • https://en.m.wikipedia.org/wiki/Lazare_Carnot military engineer, father of Sadi Carnot, member of the National Convention and Committees of Public Safety, later Napoleon's war ministry and architect of the levée en masse.
    • https://archive.org/details/reflexionsonmet00browgoog/page/n24/mode/1up wrote a book on calculus -- this was decades before Cauchy gave us epsilon-delta definitions, so we're still making geometric arguments about what limits even are
• https://vibecasting.xyz/  collaborative forecasting
• https://www.paradigm.xyz/2025/05/multiverse-finance add liquidity to a prediction market without adding more users by allowing more complex conditional possibilities. do i understand how it works beyond that? no.
• https://pol.is/home facilitates public opinion gathering

• https://www.ttm.com/en largest American manufacturer of PCBs

   

links 5/27/25: https://roamresearch.com/#/app/srcpublic/page/05-27-2025

• https://www.plex.tv/  appears to be a place you can make a curated hub for streaming videos?
• https://www.descript.com/  record and edit video; generate transcripts from uploaded video or YouTube links. very convenient and fun to use.
• George Priest, writing in the 1980s, about the need for tort reform and the excessive strictness of product liability
  • https://openyls.law.yale.edu/handle/20.500.13051/5001
  • https://openyls.law.yale.edu/handle/20.500.13051/5008
• https://arxiv.org/abs/2504.11501 Dean Ball's proposal for AI governance
• https://nexus-tool.com/docs/getting-started  collective opinion sharing tool
• https://www.nejm.org/doi/full/10.1056/NEJMoa2405847?query=WB  orexin agonist helps with the type of narcolepsy that comes from orexin deficiency
  • https://en.wikipedia.org/wiki/Orexin
    • peptide produced from hypocretin
    • receptors in lateral hypothalamus
    • promotes wakefulness, food intake, and increased energy expenditure -- the "SEEK system"!
• Anthropic's AI for science program -- apply for free credits
  • https://www.anthropic.com/news/ai-for-science-program

links 5/23/25: https://roamresearch.com/#/app/srcpublic/page/05-23-2025

• Convergent Research is hiring https://jobs.lever.co/convergentresearch
• https://marginalrevolution.com/marginalrevolution/2025/05/no-evidence-of-effects-of-testosterone-on-economic-preferences.html testosterone doesn't appear to affect economic preferences like risk-taking, in a large (n=1000) randomized trial
• https://www.ribbonfarm.com/2016/07/13/welcome-to-nixonland/ Carlos Bueno on Nixonland
• https://www.gutenberg.org/files/1322/1322-h/1322-h.htm Leaves of Grass
• https://www.poetryfoundation.org/poems/45477/song-of-myself-1892-version Song of Myself
• https://www.ribbonfarm.com/2016/02/25/the-epic-struggle-between-good-and-neutral/
  • yet another "feeling + premodern culture vs. thinking and markets" dichotomy that seems to be more about his own personality than a universal axis of variation
• https://www.ribbonfarm.com/2011/08/26/the-scientific-sensibility/
  • same deal -- science is unsentimentality? sort of, but is that the essence of the thing?

• https://www.ribbonfarm.com/2016/04/28/immortality-begins-at-forty/ content warning: aging, cynicism. but actually a piece that has stayed with me

   

links 5/21/25: https://roamresearch.com/#/app/srcpublic/page/05-21-2025

• https://www.hathitrust.org/ books online
• https://en.m.wikipedia.org/wiki/Database_consumption when people are more interested in the sorts of "worldbuilding" you could put in a database than in stories themselves
• how many people get personalized medicine?
  • https://www.theguardian.com/science/article/2024/aug/18/rise-in-dna-tests-used-to-claim-citizenship-of-other-countries-brexit-eu more than 40M people worldwide have gotten a genetic tests
  • https://www.bbc.com/future/article/20231004-these-people-took-dna-tests-the-results-changed-their-lives 26 millions of Americans have gotten a genetic test
  • https://www.npr.org/sections/shots-health-news/2025/03/28/nx-s1-5342632/concierge-membership-primary-care-doctor-shortage-rural-health-access there are 12,000 concierge medicine practices in the US
  • https://www.businesswire.com/news/home/20220607005447/en/The-Institute-for-Functional-Medicine-Releases-2020-2021-Annual-Report-Highlighting-Impact-of-Critically-Important-Mission-Driven-Work there are 1800 certified functional medicine practitioners in the US
  • https://www.grandviewresearch.com/industry-analysis/genetic-testing-market-report the global genetic testing market is $11.71B
  • the global direct-to-consumer testing market is $1.7B https://www.grandviewresearch.com/industry-analysis/direct-to-consumer-genetic-testing-market-report
  • https://www.thelancet.com/journals/lancet/article/PIIS0140-6736(24)01924-X/fulltext the global market for direct-to-consumer medical testing is projected to be $9B by 2033
• Pharma proteomics Project with the UK Biobank shows lots of proteomic disease associations
  • https://www.nature.com/articles/s41591-024-03142-z.pdf
  • https://www.nature.com/articles/s41586-023-06592-6#MOESM1
• https://en.wikipedia.org/wiki/Occludin OCLN is associated with depression; something to do with gut permeability maybe.
• https://en.wikipedia.org/wiki/GDF15 GDF15 rises sharply w age, probably a compensatory response to inflammation
• https://en.wikipedia.org/wiki/Elastin ELN goes up in the blood w age, as it leaves blood vessels that are becoming more rigid
• https://pmc.ncbi.nlm.nih.gov/articles/PMC3159481/ PEA is a proteomics technique that's a lot cheaper to multiplex than ELISA or mass spec; uses tiny quantities of antibodies to barcode proteins and then sequences the barcodes.
  • https://olink.com/technology/pea-technology-video

links 5/20/2025: https://roamresearch.com/#/app/srcpublic/page/05-20-2025

• psalm 51:
  • https://en.wikipedia.org/wiki/Psalm_51 English, Hebrew, Greek
  • https://kpshaw.blogspot.com/2015/03/miserere-mei-deus.html Latin
  • https://www.youtube.com/watch?v=36Y_ztEW1NE Allegri Miserere
• https://www.understandingai.org/p/i-got-fooled-by-ai-for-science-hypeheres plasma physicist gets disappointing results trying to implement "physics-inspired neural networks" for solving PDEs. the example in the paper works, but a different equation doesn't converge at all to the known closed-form solution & he can't find any settings that get it to work.
•