AI-Estimated Facial Age of Longevity Experts: An Exploratory Photo Experiment

Short answer: this experiment compares AI-estimated apparent facial age from video screenshots of well-known longevity personalities. It does not measure biological age, rate of aging, healthspan, or rejuvenation. The useful signal is narrower: how old a face appears to one age-estimation model under particular image conditions.

What this experiment can and cannot measure

I originally built this as an informal way to compare people who publicly discuss anti-aging. The photographs and calculations remain interesting as an exploratory dataset, but I now think the terminology needs to be stricter. Throughout this article, a negative age gap means only that the software estimated a face as younger than the person’s chronological age. I call this the apparent facial age gap, not “biological age” or “rejuvenation age”.

This distinction is supported by independent testing. NIST’s Face Analysis Technology Evaluation found that age-estimation accuracy varies by algorithm, image quality, age, sex, region of birth and interactions among those factors. The same person’s estimate can shift when facial expression or eyeglasses change, and estimates from video frames can vary by several years. In NIST’s 2024 evaluation, mean absolute error on a standardized visa-photo dataset was about 3.1 years for the tested algorithms, and performance differed across demographic groups.

Reference: National Institute of Standards and Technology, NIST Reports First Results From Age Estimation Software Evaluation, based on NISTIR 8525.

In this article, I used an age recognition system from the ToolPie service: https://age.toolpie.com/

I’ve been using this service for about a year.

There are other similar systems, but either they are not convenient, or it’s not possible at all to upload a third-party photo. I’ve tested this service multiple times on different acquaintances, and it shows good accuracy. I tried to choose photos taken head-on and with a straight head position. This is because, for instance, when the shot is taken from above, the system shows the age to be younger.

The age recognition system for photos is based on the use of artificial intelligence and machine learning. First, the computer is trained on a large set of photos of people of different ages, then it uses this experience to estimate the age of a new person in a photo, analyzing their facial features and other signs.

The order of this list doesn’t follow any particular logic. I found information about these people on the internet in roughly this order based on queries related to anti-aging and rejuvenation.

So, here’s my list:

1. Aubrey de Grey: A biomedical gerontologist who co-founded the SENS Research Foundation, an organization dedicated to studying and defeating the diseases of aging.
2. David Sinclair: A professor at Harvard Medical School known for his work on understanding why we age and how to slow its effects.
3. Valter Longo: A biochemist known for his research into aging and life span, particularly his work on fasting.
4. Peter Attia: An anti-aging physician who runs a medical practice focused on health, longevity, and preventive medicine.
5. Andrew Weil: Known for his work in integrative medicine and his “Anti-Inflammatory Diet”.
6. Sara Gottfried: A Harvard-educated MD and author, known for her work on hormonal balance.
7. Joseph Mercola: A well-known figure in alternative medicine.
8. Ray Kurzweil: A futurist and an AI visionary at Google, known for his claims about the potential for life extension through a combination of supplements, diet, and future advancements in biotechnology.
9. Deepak Chopra: A man who believes that one can significantly impact their aging process through a combination of meditation, yoga, and a specific diet.
10. Sadhguru (Jagadish Vasudev): A guru who promotes holistic well-being, emphasizing inner transformation and mindfulness practices.

Below I analyze the screenshots person by person. For each image, I subtract chronological age from the software’s estimated age. The result is an apparent facial age gap: for example, -10 means the model judged that particular image about 10 years younger than chronological age. Because lighting, camera angle, expression, facial hair, cosmetics, image quality and model bias can all affect the estimate, the numbers should be treated as repeated observations from one image-analysis tool, not as measurements of biological aging.

For instance, the first video from 2020, with an age determined to be 44 years, minus the actual age of 57 years, gives us -13 years – Aubrey de Grey appears to be 13 years younger than his actual age in this video.

1. Aubrey de Grey. Born: April 20, 1963

https://www.youtube.com/watch?v=oWYBdrQPQjM – video from 2020: 44 (determined) – 57 (actual) = -13 years
https://www.youtube.com/watch?v=anyWHDrmZbg – video from 2022: 45 (determined) – 59 (actual) = -14 years
https://www.youtube.com/watch?v=lg9v9ii2Kkk – video from 2023: 50 (determined) – 60 (actual) = -10 years
Average: -12.3 years.

Well, quite an impressive result, although in recent years Aubrey has been showing signs of aging. Or could this thick beard be preventing the artificial intelligence from accurately determining his age?

2. David Sinclair. Born: 26 June, 1969

https://www.youtube.com/watch?v=IEz1P4i1P7s – video from 2019: 40 (determined) – 50 (actual) = -10 years
https://www.youtube.com/watch?v=r2DxoqkGQJI – video from 2022: 39 (determined) – 53 (actual) = -14 years
https://www.youtube.com/watch?v=fsUUgTPtwnc – video from 2023: 39 (determined) – 54 (actual) = -15 years

Average: -13 years (more detailed analysis)

David Sinclair’s screenshots produce a consistently negative apparent-age gap, and the model estimates the later images as younger relative to chronological age. That makes him an interesting case to keep tracking. It does not show that he has halted biological aging: the three images were not captured under standardized conditions, and differences of several years are within the kind of variability seen in independent age-estimation testing. See also my more detailed analysis.

3. Valter Longo. Born: October 9, 1967

https://www.youtube.com/watch?v=jNI8IlMhg3A – video from 2020: 47 (determined) – 53 (actual) = -6 years
https://www.youtube.com/watch?v=z4GD5cKG1sE – video from 2022: 47 (determined) – 55 (actual) = -8 years
https://www.youtube.com/watch?v=7PdKZPBen0w – video from 2023: 52 (determined) – 56 (actual) = -4 years

Average: -6 years.

Certainly, he’s doing better than most regular people, but it’s clear that Valter is not the frontrunner.

4. Peter Attia. Born: 19 March, 1973

I had to put in some extra effort to get good screenshots of him: he likes to wear a baseball cap when giving speeches. I’m not sure how such headgear might affect the age estimation result.

https://www.youtube.com/watch?v=vDFxdkck354 – video from 2017: 35 (determined) – 44 (actual) = -9 years
https://www.youtube.com/watch?v=ExZ9JF69px0 – video from 2021: 44 (determined) – 48 (actual) = -4 years
https://www.youtube.com/watch?v=v8VFTQ74bqo – video from 2023: 51 (determined) – 50 (actual) = +1 year

Average: -4 years.

The model’s estimate moved upward across these three screenshots. That is an interesting within-person pattern, but it does not show that Peter Attia is biologically aging faster or that his lifestyle is causing harm. Camera conditions, cap use, expression, lighting and ordinary model error are plausible alternative explanations. The result is useful for generating questions, not answering them.

5. Andrew Weil. Born: June 8, 1942

https://www.youtube.com/watch?v=SYmb0DYRQKQ – video from 2011: 62 (determined) – 69 (actual) = -7 years
https://www.youtube.com/watch?v=bcSr7hueQOo – video from 2021: 56 (determined) – 79 (actual) = -23 years
https://www.youtube.com/shorts/TlS3OQ7jwaI – video from 2023: 65 (determined) – 81 (actual) = -16 years

Average: -15.3 years.

This is the most negative average apparent-age gap among the people examined up to this point. The large year-to-year spread is a reason for caution, not a reason to infer rejuvenation. Age-estimation models can have age-dependent bias, and the photographs were not standardized. Andrew Weil remains interesting to track, but this experiment cannot connect the facial estimate to his anti-inflammatory approach.

6. Sara Gottfried. Born: February 23, 1967

Here is the first woman on my list. We need to be especially attentive here because cosmetic procedures may strongly influence the results. Of course, men could also take advantage of this, but it’s likely more common for women. In any case, cosmetics are important too, and her experience is very interesting. So, let’s take a look:

https://www.youtube.com/watch?v=BImf6R97UiE – video from 2019: 41 (determined) – 52 (actual) = -11 years
https://www.youtube.com/watch?v=O_DgxuZtgDs – video from 2021: 37 (determined) – 54 (actual) = -17 years
https://www.youtube.com/watch?v=BrFT6N7KaL0 – video from 2023: 34 (determined) – 56 (actual) = -22 years

The model estimates these particular screenshots as substantially younger than chronological age. That is visually interesting, but cosmetics, lighting, image processing, expression and sex-related algorithm error are all plausible contributors. NIST testing has found age-estimation error to be higher for female faces in many tested systems, so I would not interpret this pattern as evidence of biological rejuvenation.

On the other hand, here is a video from the same year, and the result is not as astonishing:
https://www.youtube.com/watch?v=OxzpB-ZZomQ – video from 2023: 42 (determined) – 56 (actual) = -14 years (photo)

Average: -16 years.

Including the second 2023 image, the average apparent facial age gap is about -16 years. The within-year difference between the two 2023 screenshots is itself useful: the same person’s estimate can move substantially with the image. Cosmetics may contribute, but this experiment cannot quantify how much.

7. Joseph Mercola. Born: July 8, 1954

https://www.youtube.com/watch?v=qsRBYKsdshQ – video from 2019: 54 (determined) – 65 (actual) = -11 years
https://www.youtube.com/watch?v=t1e45Rhz0wE – video from 2020: 53 (determined) – 66 (actual) = -13 years
https://www.youtube.com/watch?v=M4QlWWt_lW4 – video from 2023: 63 (determined) – 69 (actual) = -6 years

Average: -10 years.

Joseph started well, but something seems to have knocked him off track in the last year. Nonetheless, I respect him for his public stance.

8. Ray Kurzweil. Born: February 12, 1948

https://www.youtube.com/watch?v=ybb052CNE3I – 2018: 54 (determined) – 70 (actual) = -16 years
https://www.youtube.com/watch?v=JbKNbMgRbJA – 2021: 58 (determined) – 73 (actual) = -15 years
https://www.youtube.com/watch?v=8SLnP8zikXI – 2023: 59 (determined) – 75 (actual) = -16 years

Average: -15.7 years.

Ray consistently maintains a good result, but at this rate, he risks losing this game of time and might not live to see the immortality he predicts.

9. Deepak Chopra. Born: October 22, 1946

https://www.youtube.com/watch?v=CHmnPVApfFE – 2018: 52 (determined) – 72 (actual) = -20 years
https://www.youtube.com/watch?v=mJp9MwjkTrI – 2020: 51 (determined) – 74 (actual) = -23 years
https://www.youtube.com/watch?v=znV8nSd2odg – 2023: 56 (determined) – 77 (actual) = -21 years
Average: -21.3 years.

Deepak Chopra has a very large negative apparent-age gap across these screenshots. The consistency is interesting, but it still cannot distinguish appearance from model bias. NIST evaluations show that region of birth and demographic characteristics can affect age-estimation accuracy, so cross-person comparisons here should be treated cautiously.

10. Sadhguru (Jagadish Vasudev). Born: September 3, 1957

Even though Sadhguru is not directly involved in anti-aging, I decided to include him in my list because he teaches about life, and I’m curious to see if his teachings help to stay young.

https://www.youtube.com/watch?v=0vOwBb-w0Qc – 2018: 36 (determined) – 61 (actual) = -25 years
https://www.youtube.com/watch?v=wvIbG6zMbME – 2021: 30 (determined) – 64 (actual) = -34 years
https://www.youtube.com/watch?v=4VkIJKuDyMQ – 2023: 28 (determined) – 66 (actual) = -38 years
Average: -32.2 years.

This is the largest negative apparent-age gap in this small dataset, but it is also a warning about interpretation. A 66-year-old being estimated as 28 is more plausibly evidence that the model is being strongly affected by image or facial characteristics than evidence of literal rejuvenation. Facial hair, skin texture, lighting, ethnicity or demographic bias in the model may all contribute. I would want repeated standardized photographs and preferably several independent algorithms before treating this as more than an outlier.

So, what do we have? By my voluntary decision, I select the following individuals for further study:

• David Sinclair – good results over time and in general, he appeals to me (I’m straight, just in case).
• Andrew Weil – also very good results.
• Peter Attia – I choose him for the worst results among those considered. To understand what he might have done wrong.
• Sara Gottfried – she will be interesting to a female audience.
• I could have chosen Deepak Chopra, but he overlaps with Sadhguru, and Sadhguru has higher metrics, so we keep Sadhguru.

To make the ranking easier to scan, I summarized the article’s screenshot-based results in the infographic below.

This chart should be read as an exploratory comparison of how old these personalities appeared to age-estimation software, not as proof of true biological rejuvenation.

Now, let’s try to compare their approaches with each other.

David Sinclair’s approach is purely scientific. Unlike Peter Attia, Sadhguru and others, who focuses more on lifestyle factors like diet and exercise, Sinclair emphasizes the importance of genetic and cellular mechanisms in aging. He’s particularly invested in the development of medicines that can mimic the effects of diet and exercise at the cellular level. Sinclair is more interested in how the manipulation of certain genes and cellular processes can potentially reverse aging. Perhaps he has already invented some magic pill, but we can’t rely on that yet, so we need to find another example to emulate. Read more about his approach here: David Sinclair’s Anti-Aging Essentials and Lifestyle

Andrew Weil, Sara Gottfried, and Sadhguru advocate for a holistic approach to health and wellness, emphasizing the importance of both physical and mental wellbeing in anti-aging and rejuvenation. Each of them underscore the importance of good sleep hygiene, stress management techniques, and avoiding harmful habits like smoking or excessive alcohol consumption. They all promote a balanced diet and regular physical activity as crucial components of any anti-aging regimen.

Dr. Sara Gottfried’s approach to anti-aging and rejuvenation is unique as it emphasizes hormonal balance and its crucial role in aging. She specifically focuses on women’s health and the unique biological and hormonal differences that affect their aging process. Beyond medical treatments, she also advocates for lifestyle interventions such as a healthy diet, regular exercise, good sleep, and stress management. She has a strong belief in the mind-body connection and promotes practices like yoga and mindfulness meditation. Lastly, she uses genetic testing to create more personalized anti-aging strategies, integrating traditional medical knowledge with a more holistic perspective of health and wellbeing.

Dr. Peter Attia, our outlier, focuses on “healthspan” – a life free of chronic diseases and full of vitality. He is known for his rigorous personal experiments with diet, physical exercises, and various biohacking methods. He has a strong focus on physical fitness and has participated in extreme endurance events. His training regimen is intensive and meticulously tracked. Attia practices intermittent fasting and follows a low-carbohydrate, high-fat diet, known as a ketogenic diet. He has also experimented with prolonged fasts. Could it be that his workouts and fasting are excessive?

Unlike Peter Attia’s more rigorous, science-backed methodologies, Weil places greater emphasis on lifestyle changes, natural remedies, meditation, and proper nutrition. His strategies focus more on overall well-being and mental health in addition to physical health.

On the other hand, comparing Weil to Sadhguru, an Indian yogi and mystic, shows some similarities in their focus on holistic wellness and mindfulness practices. But Sadhguru’s approach is deeply rooted in spiritual traditions and yoga practices of India, which can be considered more esoteric compared to Weil’s more Western-oriented, integrative medicine approach.

The key factors identified from the different approaches of David Sinclair, Andrew Weil, Sara Gottfried, Peter Attia, and Sadhguru include:

• Cellular Mechanisms and Genetics: As emphasized by Sinclair, understanding and manipulating our genetic and cellular mechanisms might hold the key to reversing aging.
• Holistic Health: Advocated by Weil, Gottfried, and Sadhguru, this approach emphasizes the importance of physical and mental wellbeing, incorporating lifestyle changes, stress management, good sleep hygiene, and avoiding harmful habits.
• Hormonal Balance: Gottfried highlights this as a crucial factor, especially in women’s health. Maintaining hormonal balance can significantly affect the aging process.
• Diet and Exercise: A balanced diet and regular physical activity are universally acknowledged as crucial components of any anti-aging regimen. The type of diet and intensity of exercise can vary, as seen in Attia’s approach of a ketogenic diet and rigorous physical regimen.
• Mind-Body Connection: Practices such as yoga and mindfulness meditation, as emphasized by Sadhguru and Gottfried, can play a significant role in maintaining overall health and wellbeing, thus contributing to anti-aging.
• Biohacking and Personal Experimentation: Attia’s approach showcases the potential of personalized regimens based on self-experimentation. However, it also raises the question of whether extreme regimens might be counterproductive.

What this experiment gives us is not a ranking of whose anti-aging approach “works.” It gives us a set of reproducible image-level observations and several hypotheses worth testing. To connect an apparent facial-age gap with slower biological aging would require standardized photography, repeated measurements, multiple independent algorithms and, ideally, independent biological-aging or clinical outcomes. Until then, the photographs are evidence about appearance as judged by one model, not evidence that one person’s regimen extends life.

Method reference

Hanaoka K, Grother PJ, Ngan ML, Yang J, Quinn GW, Hom A. Face Analysis Technology Evaluation: Age Estimation and Verification. NISTIR 8525, National Institute of Standards and Technology, 2024. NIST continues to update its age-estimation evaluation as new algorithms are submitted.

Whom else would you like to analyze in a similar manner? Write in the comments, and I will conduct such a review. If you have run a similar experiment with standardized images or several age-estimation models, please share the method as well as the result.