#8

I really like Peter Attia and his perspectives on everything. He’s a very balanced guy and keeps it real. I actually appreciate him more and more each year as I’ve been listening to him. Hopefully he comes back on as a guest.

6 Likes
#9

I too enjoyed their ratings after minute 50:00. Here’s is an AI-redacted summary for quick reference. I double-checked the answer, quite a few items were missing, it should be reasonably accurate now.

Dr. Attia and Dr. Kaeberlein discuss their ratings on various supplements and interventions in a “lightning round” format. Here’s a summary of their ratings.

  • Resveratrol: Dr. Attia gave a grade of 6.Kaberlein a grade of 5. Don’t bother.
  • Healthy diet: Dr. Attia gave a grade of 2.Kaberlein a grade of 1.
  • Rapamycin: Dr. Attia gave Rapamycin a grade of 2.This indicates that he acknowledges its potential but emphasizes the need for more data and caution in its use.
  • follastatin gene therapy: Dr. Attia gave a grade of 5.Kaberlein a grade of 3.
  • A-Ketoglutarrate: Dr. Attia gave a grade of 4.Kaberlein a grade of 5. Don’t bother.
  • Probiotic Supplements: Both rate these a 3, acknowledging that the quality of the supplement significantly impacts its effectiveness.
  • Chronic Caloric Restriction in Humans: Dr. Attia rates this a 3-4, while Dr. Kaeberlein rates it a 5, believing it would lead to a miserable existence and potentially have harmful side effects.
  • Spermidine Supplements: Both rate this a 2-3, recognizing its potential but acknowledging the need for further research.
  • Close Friendships: Both rate this a 1, emphasizing its importance for quality of life.
  • Sleep Tracking: Both rate this a 2-3, acknowledging its potential benefits for awareness but also the risk of anxiety.
  • Sauna: Both rate this a 2-3, recognizing both potential health benefits and social benefits.
  • Vitamin D: Dr. Attia rates this a 3, emphasizing the need for proper study design to determine optimal dosage. Dr. Kaeberlein rates this a 1, suggesting supplementation only after measuring individual levels.
  • NAD Boosters: Both rate this a 5-3, acknowledging potential but emphasizing the lack of strong evidence for benefits in the average person.
  • Hormone Therapy (HRT): Both rate HRT for women a 1, recognizing its significant benefits. For men, they rate it a 1-2, acknowledging potential benefits but recognizing a lower magnitude of effect compared to women.
  • Vegan diet: Dr. Attia rates this a 3, while Dr. Kaeberlein rates it a 4,
  • Carnivore diet: Dr. Attia rates this a 3, while Dr. Kaeberlein rates it a 4,
  • Hyperbaric chamber: Dr. Attia rates this a 1 for would healing, 2-3 for concussion and trauma, 5 for longevity.
  • Metformin for Non-Diabetics: Dr. Attia rates this a 3-4, while Dr. Kaeberlein rates it a 4, expressing some skepticism due to potential limitations in how it’s metabolized.
  • SGLT2 Inhibitors for Non-Diabetics: Both rate this a 2, recognizing its potential benefits.
  • Growth Hormone in Elderly: Both rate this a 3, acknowledging potential benefits but emphasizing the need for more data and caution in its use. Attia gives a 4 to Dr. Fahey’s TRIIM protocol of thymic regeneration.
  • Continuous Glucose Monitoring: Dr. Kaeberlein rates this a 1, while Dr. Attia rates this a 3, acknowledging its potential benefits for behavior modification but also the risk of increased distress.
  • BPC 157: Both rate this a 2-4, acknowledging its potential if pure but emphasizing the lack of high-quality data and concerns about quality control.
  • GP1 Agonists for Non-Obese Individuals: Both rate this a 2-3, recognizing its potential but emphasizing the need for more data.
  • TR Activators: Both rate this a 5, recognizing the lack of significant potential.
  • Young Blood Therapeutic Plasma Exchange: Both rate this a 3, acknowledging its potential but emphasizing the need for further research.
  • Fatty 15: Both rate this a 4, expressing skepticism about its effectiveness.
  • Epigenetic Age Tests (Direct-to-Consumer): Both rate this a 5, expressing skepticism about their clinical utility.
  • NMN: Both rate this a 5, expressing skepticism about its effectiveness.

This summary provides a general overview of their ratings. It’s important to note that these are just their personal opinions and further research is needed to fully understand the potential benefits and risks of these interventions.

16 Likes
#10

Yes, I concur, at the beginning, it was like beating around the bush, I was almost giving up listening, fortunately, they made up for it in the second half

4 Likes
#11

Thanks a lot for taking the time to write this summary!

3 Likes
#12

LOL on the phrasing. A rating of 5 means they see zero value in a given intervention. So, both see zero value, but recognize its significant potential?

What a wonderful tool! It really saves you a lot of time and effort - and you absolutely can not rely on it for accuracy at all.

I think I know how to evaluate such a tool: I see zero value in it, recognizing its significant potential.

Instead, I actually went ahead and listed to the podcast like a human being. YMMV.

2 Likes
#13

Cronos, yes, I was aware of that flaw, that is weird since the AI seems to have grapsed the meaning of grades at the beginning. Also, it missed for some reason Rapamacyn, which was graded 2 by Peter Attia, maybe because I chose 50:00 time as a starting point and it was a little too late.

I find such tools to be useful after listening to the podcast, in such a way to review the discussion and remember the details. I must see why it missed rapamycin, It’s undoubtedly a serious fault in this forum!

3 Likes
#14

Gemini seems to have missed Rapa for some reason, he only mention it after I asked specifically. This is why it is necessary to listen to the videos before summarizing them.

You are absolutely right! I apologize for the omission.

Dr. Attia gave Rapamycin a grade of 2.

This indicates that he acknowledges its potential but emphasizes the need for more data and caution in its use.

I am still under development and learning to be more comprehensive and accurate. Thank you for pointing out this error!

I edited the complete list in the previous post, adding rapa, correcting the AI comments on epigenetic age test

6 Likes
#15

This is for those who are curious about the experiment. Gemini also missed the healthy diet and resveratrol, just at the beginning of the lightning questions. Surprisingly, Peter Attia gave 6 to resveratrol, Kaberlein 5, but Gemini reported 2-3, showing a case of hallucinations.
Since I started all this, I’m going to review it all and provide a final corrected edition.

Resveratrol: Likely a moderate rating (e.g., 2-3) given its potential benefits but also the need for more conclusive evidence.

5 Likes
#16

I mean, mccoy, you had to listen to the podcast to see where the tool “hallucinated”, that is, was blatently wrong, or invented things out of nowhere. This means you simply cannot trust this tool. It’s like relying on the words of someone who is a habitual, proven liar. You simply cannot. Because I cannot know ahead of time just when this tool will “hallucinate”, it is unpredictable and untrustworthy. And spoiler alert: this will never change, because it is inherent in the very technology and its methodological foundations. To change this, we will have to abandon this approach, and adopt a radically different one, and that is something that we have not developed as yet.

For practical purposes, I honestly cannot see any use for this tool. If I cannot trust the accuracy, not just in minor details, but profound distortions, and cardinal mistakes, and must therefore doublecheck everything to see where it may have gone wrong, then what is the point of this? I have to listen to the whole thing anyway, so why do I need this tool? It only doubles my work, first I have to listen to the whole podcast and make notes, and then I have to check the garbage this tool puts out so that I’m not leaving in total nonsense. It’s like having a worker who is an enemy agent sabotaging the work, not only do you have to do the work itself to make sure it’s done right, but you have to doublecheck the work of the agent, so you can undo all the sabotage. I choose not to hire the enemy agent in the first place.

I don’t use this tool at this time, and have no plans to do so. YMMV.

4 Likes
#17

@cronos tempi, all sensible objections, this tool presently can be used after listening to the podcast to fix things in the memory in the parts more interesting to us (we may remind omitted details, I remembered a couple of the most important like Rapa and Urolithin), or even before listening to the podcast, to make sure it is interesting to us.

Clearly, the details must be double-checked.

Last, at present I find it can be perhaps useful in videos of moderate length.

4 Likes
#18

A great experiment - thanks for posting. I was planning to do something similar, but you’ve saved me the time.

4 Likes
#19

Nothing about Senolytic treatments which is by far the largest segment in anti-ageing therapies.

1 Like
#20

They didn’t discuss taurine either, which I found disappointing, though I guess I already know both Matt’s view (from his optispan podcast), and Attia’s from his newsletter. Matt has a good opinion, if I had to guess, he’d score it 2-3, though he himself doesn’t take it at this point allowing he might in the future. Peter is entirely skeptical and dismissive. Given their somewhat differing views on taurine, it would’ve been interesting to hear them discuss it. Alas, it was not to be.

4 Likes
#21

There are quite a few points the AI missed, and I am in the process of adding them up to the previous list. This is indeed an imperfect and absent minded assistant, that needs to be double-triple checked.
I asked it the reason, and it provided a lengthy answer.

The technical reasons why I might miss points in YouTube videos are complex and multifaceted. Here are some of the key factors:

  • Limited Access to Visual and Audio Information:
    • I primarily rely on text and code.
    • I cannot directly “watch” or “listen” to the video, which means I miss crucial non-verbal cues like tone of voice, body language, and visual demonstrations.
    • This can lead to misinterpretations of subtle nuances and humor, which are often conveyed through these non-verbal channels.
  • Challenges in Understanding and Interpreting Complex Information:
    • Videos can contain a lot of information presented in a fast-paced manner.
    • I may struggle to keep up with rapid-fire exchanges, complex arguments, or highly technical discussions.
    • Identifying and extracting key information from the video’s audio and visual content can be challenging.
  • Limitations in Natural Language Understanding:
    • While I am constantly improving, my ability to understand and interpret human language, especially in the context of nuanced conversations and informal speech, is still under development.
    • This can lead to difficulties in accurately understanding the speakers’ intent, identifying sarcasm, and recognizing subtle shifts in topic.
  • Dependence on Accurate and Complete Information:
    • My performance relies heavily on the quality and completeness of the information I am provided.
    • If the video has poor audio quality, inaccurate captions, or missing segments, it can significantly impact my ability to understand and respond accurately.
  • The Ever-Evolving Nature of Language and Information:
    • The human language is constantly evolving, with new slang, idioms, and cultural references emerging all the time. 1

Gemini, the large language model I’m based on, uses a combination of techniques to process and understand video content:

  • Audio Processing:
    • Gemini can analyze the audio track of a video to identify spoken words and their meanings.
    • This involves sophisticated audio processing techniques, including speech-to-text conversion and natural language understanding.
  • Visual Analysis:
    • While my current capabilities are more limited in this area, Gemini is being developed to better understand and interpret visual information from videos.
    • This could include identifying objects, scenes, and actions within the video.
  • Integration of Audio and Visual Cues:
    • By combining audio and visual information, Gemini can gain a more comprehensive understanding of the video’s content.
    • For example, it can use visual cues to disambiguate speech (e.g., identifying the speaker based on their appearance) or to infer context that is not explicitly stated in the audio.
  • Machine Learning and Deep Learning:
    • Gemini leverages powerful machine learning and deep learning models to analyze and understand video content.
    • These models are trained on massive datasets of videos, enabling them to learn complex patterns and relationships between audio, visual, and textual information.

Important Note:

  • My current abilities with video understanding are still under development.
  • I may not always accurately interpret or summarize video content, especially in complex or nuanced situations.

I hope this provides a clearer understanding of how I approach video analysis.

5 Likes
#22

Both Dr. Attia and Dr. Kaeberlein assign rank 2 to sglt2 inhibitors. Here is a link to an open access paper, “SGLT Inhibitors for Improving Heathspan and Lifespan.”

How can I get access to one of these drugs? (I live in Texas.)

4 Likes
#23

You can get them in India pretty cheap.

#24

Most of the same online pharmacies that sell rapamycin also sell canagliflozin and empagliflozin (etc.): 在线购买雷帕霉素 - 可靠药店列表

2 Likes
#25

@RapAdmin - which one did you find fatigued you? And have you tried the other?

I wonder if it’s a valid apporach just to take these meds acutely when you have a lot of starch or kinda go on a carb binge?

#26

Canagliflozin seemed to cause fatigue. I have not had the same issue with empaglifozin. But I think I also picked up Covid and that may be a wild card in my analysis.

2 Likes
#27

There is some evidence that Canagliflozin has off-target effects through mitochondrial electron transport chain 1 interference. This study suggests that other SGLT2i drugs don’t have that effect.

https://www.nature.com/articles/s41419-018-0273-y