#25

Sure - as with anything, but I have much more faith that we can move AI quickly up the ladder of judgement and analysis (i.e. choosing what information to believe / use and what to throw away) than the vast bulk of humanity.

1 Like
#26

Maybe if AI is integrated with this forum this will eventually solve the main information overload issues :smile:

1 Like
#27

What would be really interesting (I think, but what do others think?) is this effort to sort of open-source our protocols and results. Sort of like an open source version of what Bryan Johnson and Oliver Zolman have done with their site the Rejuvenation Olympics. Basically an open source rejuvenation olympics for the rest of us (who don’t want to pay $200 a shot for the True Diagnostics test every X months).

In an ideal world it would take our inputs, and (here is when things get more difficult) our blood test results - and use whatever biomarker is available to us, or that we have used, and add that to our “results” information to help rate and rank the participants.

In an ideal world I’d love some sort of AI parsing program that would allow us to simply upload our blood test results from LabCorp or Quest, etc. - and have our blood test results automatically parsed from the PDF and entered into our database of test results. That way we can all share our “results” and we can all compare… and learn from each other.

6 Likes
#28

Actually, we will be integrating AI into this forum soon - I just upgrade very slowly to minimize my software management overhead. See: Discourse AI Features | Discourse - Civilized Discussion

5 Likes
#30

That looks great. Certainly a better way to extract the essential info from the super-long threads and discarding all the crap, like @SNK’s posts and my jokes, would be great.

Yes, we’ve talked about this and of course it’s a great idea. I’m still working on a full battery of tests - functional, blood, Dexa, VO2 max, etc. to measure longevity progress (or for a longevity leaderboard). It would be great if we had some kind of form here to just fill in the scores. I think we’ve talked about it enough to know (more or less) what to measure. There are a bunch of apps out there that services use as a personal health concierge - Inside Tracker is one. I haven’t had a chance to look at all of them - but some of them are total crap.

Certainly Rapamycin is becoming much more visible as it makes the rounds of all the big media outlets. Soon we’ll be overrun with reporters. The Kardashians are sniffing around…David Sinclair is putting out feelers with @RapAdmin 's people to see if they could set up a meeting. No wonder there are so many new members…soon there’ll be a waiting list.

Hope I didn’t say too much…there are so many spies around, looking for a scoop.
What’s next? @Agetron signing autographed posters?

4 Likes
#32

The current biological age tests are inferior in some ways, possibly relying too much on association and fake U-shape curves (For those who don’t know, U-shape curves show lots of dubious assosciations like being overweight, and having high LDL, glucose, blood pressure, and so on all are good for “longevity” than normal or low, when there is reverse causation and confounding factors in many cases.)

I think the first step either way is someone figures out a really good longevity score based on blood tests that considers causal factors like apoB, non-HDL-c and heart disease, and weighs accurately the differents parts of the longevity or healthspan score based on level and quality of evidence.

Then people if some service allows API access upload this data before initiating treatment, or enter manually. This calculates longevity or healthspan score. Then a person initiates treatment and uploads again after some period. Change is marked. Person can share or group this together with other people if they wish, it can probably be done in some way for other rapamycin users and we have mean change in this score if there is any.

If people are really enthusiastic and there’s plenty of users (so there are some enthusiastic users) the treatment can be done in a blinded fashion with QR codes like the microdosing LSD experiment. NYT readers are going to fall off their chairs if/when they see a blinded improvement in longevity score for any treatment, it might be a future longevity drug or combo from the wormbot experiments.

But either way… there needs to be a good longevity score or healthspan score even maybe someone like Peter Attia would consider who is probably more evidence-based or traditional.

This longevity score + open source + platform could easily allow people to test future compounds with more potential than rapamycin or combinations (like those maybe found in the wormbot experiments: Super longevity drugs coming soon 2027-2028)

The bottleneck is otherwise going to be human testing, new outcome (longevity score), and so on.

3 Likes
ARPA-H launches new program aimed at extending the healthspan of Americans
#33

Yea. I’m a bit more skeptical on the scores and Blueprint type metrics used to rank/gauge progress, but I definitely think the direction they’re trying to go in is worth pursuing. I do wonder if there is some standardized way to get accurate, consistent, provable data on conventional lab results to ensure quality data. That’s the tricky part. If only there were many open-source lab testing services that can digitally verify people’s reported lab results, this would be a great way to create a more confident dashboard of everyone’s regimens and progress.

But even then, it’s tricky because you cannot prove someone’s reported regimen of X mg * Y days/week. That will always remain self reported, and maybe it’s too wishful thinking, but the protocol itself is where I hope there’s some way to get reliable data on. Something like Bitcoin is cryptographically verifiable so no one can game the system. But if we want this kind of data to be valuable and grow in the long term, there needs for there to verify its accuracy.

2 Likes
#34

Yea I was thinking of that earlier. Screenshot your results → let AI parse into the form/field for each metric tested. Again, this is valuable in so far as honest players are involved, which I imagine currently is true. But in the long run, how you prevent spam/bots/fake results from polluting the integrity of the data is a challenge I don’t see a solution for currently. Unlike as mentioned, we have lab testing services that digitally verify your results. Maybe that’s what we should start haha. Open source lab testing service with authentication for users wanting to report their data.

2 Likes
#35

You can prove it by measuring Sirolimus levels as part of the blood testing schedule, if that is necessary, but won’t be able to do a self-blinded experiment then unless that data is withwhold from users. Just thinking out loud I imagine this would require even larger amount of users so there is enough people who want to participate (but the site and longevity drug field/rapamycin is growing, at least previously it was exponentially). Shouldn’t be much of a problem for rapamycin or similar drugs with blood tests for it. New drugs from wormbot or whatever will be more difficult but it might better than nothing but will not persuade the most important influencers or decision makers from a strict utilitarian perspective, but I don’t know. Maybe wormbot with possibly fake data from some participants could be useful (If the longevity score or whatever deviates way too much with a wormbot superlongevity drug, then it could be thought of impossible to fake or very unlikely).

Another user was working on blood testing rapamycin via finger pricks, maybe custom blood tests could be done for specific wormbot possible future superlongevity drugs (招募荷兰志愿者：使雷帕霉素血液检测在整个欧洲成为现实), if deviations are not that large.

The provable blood tests would be via API access then. I don’t know about cryptographically verifiable blood tests but I like the idea.

2 Likes
#36

I read all of them plus one.

1 Like
#37

Discourse looks like a step in the right direction. Good job!

1 Like
#38

We could synchronize/share the user authentication systems used in this forum with the dashboard/supplemental data so that you’re only drawing from users here in the forum (thus eliminating the bots/spam, etc.), and you have a rating/ranking system of this forum on the users (posts/likes, etc.) that allow you to tell at a reasonable accuracy the real people who are participating here vs. bad actors.

On the blood test results side of things, there are only a few major companies (in the US) really doing this at scale (Labcorp and Quest) and if we can get an AI system to parse the data from the typical PDF results that many of us get, you getting a lot of data in addition to the raw blood test results (e.g. name/ date of testing, etc. which you wouldn’t reveal to public, but which an administrator could check on to provide another level of scrutiny if needed).

And, ultimately there isn’t much of an incentive to “hacking” the system to get a higher rating / ranking. There are no prizes as such, the real benefit is tracking your own metrics and trying to optimize them, and learning from others and tracking how they are doing so you can learn from what is working or not working for other people.

The real incentive here is optimizing your own health…

2 Likes
#39

You honestly keep reading my mind lol. I was also picturing some (doesn’t have to be AI) image recognition or screenshot recognition algo that simply takes in the raw data and somehow ‘verifies’ it through a backchannel or voluntary third party. Yes, many ways to skin the cat to improve reliability, but does come to down community driving demand for it.

Is the auth in this forum exposable to third parties through some API? I agree, proof-of-community is prob one of the best ways in the future to combat AI when it comes to fake accounts, and having long-term user metrics from this forum as a proxy for reliability of their input data is a good way to boost confidence of other users.

I was working in pathology and if I was an entrepreneur, it would be cool to start a laboratory testing service that simultaneously provides verification services.

Yes, I’m hoping the real incentive for others continues to be internal motivation.

2 Likes
#40

This would be great. I can enter all my blood test data by hand but just using the PDF is so much easier. Then a chart to show multiple blood test results over time.

I think that’s true now, but once you have a leaderboard…
That’s why the Rejuvenation Olympics gets the test results directly from TruDiagnostics.
But, of course, just using an epigenetics test is not enough. As @John_Hemming says, functional tests are important. The Rejuvenation Olympics don’t seem to show any clear benefit of taking rapamycin even though lab results indicate that they should.

There are better, more comprehensive, BioAge Tests that rely on more blood biomarkers than the 9 used for the Levine PhenoAge Test. See this thread:

https://spotify.localizer.co/t/biomarker-optimization-crowdsourcing-biomarker-knowledge-by-michael-lustgarten-and-sergey-vlasov/11868?u=ng0rge

And I think you’d have to include things like blood pressure and something like AnthropoAge or some measure of body composition (DEXA?) Also sleep score…good sleep, I would argue, affects longevity.

1 Like
#41

+1 on that idea.

Yeah - imagine some sort of open source, puiblic sharing of before/ after info. I’ve got a heap of real time data on the impact rapa has on HRV via whoop. + 3 blood tests, 1 of which is on off cycle.

Bucketing it up into correct categories and relying on people to do the right thing in terms of data integrity may be tough - but it’s a great idea/ intiative.

Encourage newbies to get a before blood test and upload it…get AI to read it, then subit afters.

The sort of “study” that will maybe never happen because there is no funding for a cheap, easily accessible drug.

healthmatters.io will take a blood test from anyone and scan/ upload the data for $15 - so it’s available

2 Likes
#42

The people here are great and there is a plethora of information here, but the site has years-long threads that are nuts to peruse. I left a note about it in Feedback to let the admin know I was really frustrated. I started here visiting most every day until I realized it is too difficult to make replies in these long threads (due to interruptions in comments by other users with tangential aspects that also go on for pages, which are then interrupted by more tangential comments, ad infinitum, until I lose track of replies to the comment I am interested in because not everyone will quote or reply directly to a comment). The search is helpful but not consistent. I hesitate to use words like “anarchy” or “chaos” lol, but that is kind of what it feels like to me (I get that others like the freewheeling nature).

1 Like
#43

I get the feeling some people dislike 1,000 post threads that would take a solid week to read though…

Yes, I am not the most active of moderators… But will try to break threads up to less than a 24 hour reading time.

And the coming ai integration should make things easier on new readers…

1 Like
#44

Sermo, AI Summary Agent, Rapamycin Blood Work database and Case Studies:

Has anyone (doctor or medical researcher) tried Sermo lately? I used to have access to Sermo and I quite liked the website’s design but that was many years’ ago. I would assume they have kept up with searchability and indexing part of their site. Sermo is mainly a case study website participated by doctors.

Sometimes I wish to have the ability to apply an AI agent on top of a thread to summarize the key points or input, without having to read everything posted there.

I do think posting blood results is a great idea and to mine the database as this is probably the largest rapamycin human data pool there is.

Lastly, I think it needs to make it “personal” to have more engagement or attract more users as then we would have unique content and increased practical value. What I meant by that is if we all share personally particulars like conditions and medications, why and how then we can all follow it up like case studies, which is what Sermo excel. Someone presents a case, doctors around the world give their diagnosis and input. There are usually follow up by the case originating doctor so it’s a feedback loop that benefits all.

1 Like
#45

Good to know. Even if these may not be immediately/quickly addressable, I think having the fingerprint or accumulating opinions on this will help keep it closer to top of mind for everyone. Hence why I even posted on this topic. Although I’m no moderator or forefront user, I’ll be looking more on this topic.

1 Like
#46

I never did address the dashboard aspect you brought up. I think it is an interesting idea and could absolutely be useful if you want to look something up or aggregate findings systematically. My comment was definitely tangential (sorry about that) and focused on user-to-user engagement, conversation, consideration of ideas, in a “forum” setting which is something I am particularly interested in (and have moderated elsewhere). I like your idea and am glad you posted about it. With the way this site is attracting more and more people something of that nature would be very helpful.