young.ai

The site https://young.ai from Insilico purports to tell you what you are doing (exercise, supplements, lifestyle etc) that helps, and what is aging you. Anybody try it?

9replies Oldest first
  • Oldest first
  • Newest first
  • Active threads
  • Popular
  • JGC  albedo

     

    This morning I'm going through and adding my data. They have another predictor here, for blood results they actually support Life Extension lab PDF files, just upload and they scan and automatically pull out the data. Nifty ... and interestingly this gives a different number when uploading the last Male Elite panel I did - 35. So so far I've got 40 (Levine method), 28 (aging.ai v3) and 35 ( young.ai ).

     

    Note that on the blood tests there's three sections, the first (with the auto scanning of PDF's) is for common blood work. There's a second tab for inflammation. The third tab tells you how your values are doing relative to where they should be. 

     

    They also have a section on a new test I'm not familiar with, the 'transcriptome' which measures your DNA expression. I've not seen where you can get a test for this, any pointers?

     

    Generally very impressed with this app, it's telling me some things about myself I've not seen elsewhere. 

    Like 2
      • albedo
      • albedo
      • 1 yr ago
      • Reported - view

      DanMcL 

      Thank you for sharing. I would have played with young.ai too if as EU resident I was also able to use it but I am not allowed, likely for some regulatory issues. I think the advantage of young.ai is to provide a nice tracking tool and possibly an integration of other than blood biomarkers.

      For tracking I would also consider plotting aging.ai results along time.

      Like
      • Danmoderator
      • skipping my funeral
      • dantheman
      • 1 yr ago
      • 1
      • Reported - view

      albedo That's easy - use a VPN to alter your IP address. Private Internet Access is a excellent cheap one I use. Then connect to a server in the U.S. and voila, welcome to North America (as far as youth.ai is concerned). If you have any other issues use privacy mode on your browser, it's possible (but unlikely) they are picking through some cookies.

      Note on this service, I haven't checked the privacy settings, presumably they don't conform to HIPPA (U.S. medical data privacy regs), but for me the value of the service is above companies learning about my (excellent) health, anonymously or not. The EU is quite picky about data privacy, I deal with that in my work too, usually the privacy requirements are too onerous to be worth bothering with the region (unless there's a lot of money in it). The flip side of regulation and privacy - it does come with a cost (not trying to moralize here :) ).  

      Like 1
      • albedo
      • albedo
      • 1 yr ago
      • Reported - view

      DanMcL 

      Thanks. See you are an expert here ... :)

      Will see but for the moment I will not bother that much. I still need to understand and be a bit more confident on the aging.ai results and answer the questions I have before trusting more what they do.

      Like
      • Danmoderator
      • skipping my funeral
      • dantheman
      • 1 yr ago
      • Reported - view

      albedo one thing for sure is they are using young.ai to generate data to feed the networks to improve them. The time tracking especially will allow them to find how hyperparameters (e.g measurement, lifestyle choices, ...) effect aging.

      Like
      • albedo
      • albedo
      • 1 yr ago
      • Reported - view

      DanMcL 

      Yes I also understand it like that. Different than aging.ai where they were more interested to the same mechanism of improvement for the blood biomarkers relative importance (in guessing age) and genetics dependency (cross ethnicity). I guess those were the main results of the two papers they refer to resp. for v 1.0 and v 3.0

      Like
  • Anyone here who tried to compare aging.ai and Levine's Phenotypic Age results?

    Like
  • I wanted to compare two senolytic approaches:

    a) Fisetin: 1500mg/ day for two days. Repeat after 30 days.

    b) Dasatinib / Quercetin  one dose of 200mg D / 2000mg Q. Repeat in One week.

    I have taken a blood test before the first treatment, and a second blood test ten days after completion of fisetin treatment. A third blood test will be be taken after completion of the second treatment.

    On the basis of the first two test results, I calculated my "Biological Age"  using both Levine's method, and Aging.AI(3.0). The results are very different:

    Before Fisetin treatment:
    Levine DnamAge: 62.9      Aging.Ai: 39

    After Fisetin treatment: 
    Levine DnamAge: 61.7     Aging.Ai: 35

    My actual age is 69

    While both methods show a reduction in age, which indicates that the treatment is effective, I am not sure how  I can interpret the "Biological Age".

    Nonetheless, we could possibly use them as tools, depending on the circumstances. In my opinion, Levine's methodology can be very useful for predicting the Biological Age of people that make no major interventions. Then, the "usual" aging process can presumably predict accurately the probability of death in the next 10 years (and the age of a person that would have the same probability). The statistical parameters show that it can be trusted.

    However, if we assume it is possible to rejuvenate someone and effectively make the person identical to one with half the age by major interventions, Levine's method could not possibly "predict" the true new biological age, since the actual age component is so strong in the formula. No combination of blood parameters can calculate the age of 35 for a 70-year-old, even if the person is fully rejuvenated. In other words, Levine's methodology is appropriate for minor interventions. On the other hand, the methodology of aging.ai could conceivably predict the new Biological Age for a rejuvenated person. Therefore, for people that make major interventions in their lifestyle with the hope of rejuvenation, the methodology of aging.ai makes more sense.

    I suspect that for people that make no major interventions in their lifestyle, the two methodologies might give comparable results. 

    I would love to have feedback on my thinking.

    Like 1
  • The way I will use these tools is only for comparison with myself and not the general population.

    I do the first test and that will be my baseline number.

    Any unexpected increase would tell me I'm on the wrong track.

    Any unexpected decrease would mean that I'm doing the right things and therefore I must continue to do them.

    Like
Like Follow
  • 13 days agoLast active
  • 9Replies
  • 192Views
  • 4 Following