Will AI help us live longer?
The latest AI copilot from Sam Altman & Arianna Huffington focuses on your health
Big bets are being made on AI copilots and the latest one focuses on your health.
This week, Arianna Huffington (founder of the Huffington Post and Thrive Global) and Sam Altman (OpenAI founder) co-wrote an article in TIME, explaining why they’re making a collective investment in a new “hyper-personalized AI health coach” called Thrive AI:
A staggering 129 million Americans have at least one major chronic disease—and 90% of our $4.1 trillion in annual health care spending goes toward treating these physical and mental-health conditions. That financial and personal toll is only projected to grow.
What does it need to achieve this?
It will be trained on the personal biometric, lab, and other medical data you’ve chosen to share with it.
It will learn your preferences and patterns across the five behaviors: what conditions allow you to get quality sleep; which foods you love and don’t love; how and when you’re most likely to walk, move, and stretch; and the most effective ways you can reduce stress.
Combine that with a superhuman long-term memory, and you have a fully integrated personal AI coach that offers real-time nudges and recommendations unique to you that allows you to take action on your daily behaviors to improve your health.”
That all sounds pretty good to me.
Based on my experience, though, we’re a way off achieving truly personalized recommendations.
Why?
It takes more effort than you think to generate useful data.
When you do, it doesn’t always create the positive results you’re hoping for.
Two personal examples…
Fitness:
Initially, I loved my WHOOP band. I quickly saw a correlation between increasing my exercise time and my fitness metrics. Being a data nerd, I also enjoyed understanding the relationship between sleep, exertion, stress, and various other lifestyle decisions, which caused my Strain and Recovery scores to vary.
There was one big problem.
The more fixated I became on what affected my (at the time severely child-impacted sleep), the worse it got. The first thing I’d do each morning is analyse my sleep:
How long had I spent in REM vs deep sleep?
How did it rate against target levels?
Was I ‘fully recovered’?
Honestly, it became such an obsession my wife had to step in and take my WHOOP away.
Nutrition:
After emerging from the years of interrupted sleep I mentioned above, I realized last year that it was time to investigate why I felt so tired every afternoon. I was certainly very fit. And my diet was healthy (or at least that’s what I thought). But, without fail, after lunch, I’d either need a nap or struggle to concentrate.
So, I signed up for a ZOE, continuous glucose monitor (CGM).
Cue a new unhealthy obsession with experimentation and data - this time about food.
Now, don’t get me wrong, the test did reveal a couple of interesting insights:
THE OBVIOUS ONE: a sugary morning bun doesn’t boost long-term energy.
THE SURPRISING ONE: bananas spike my blood sugar almost as much as a cinnamon bun.
But the experience was not positive. Not least, because despite the promises, the technology is seemingly not advanced enough to provide truly personalised results and nutrition recommendations. Instead, I was presented with generic advice, which I’d already discovered from reading this book.
The future of AI copilots
Now, this isn’t to say I’m against the idea of a hyper-personalised health coach or AI copilots, in general. Assuming we have, As Altman and Huffington write:
”…assurances that these technologies are reliable and that their personal health data will be handled responsibly.”
But in the short term at least, there are challenges:
Creating a useful data set relies on significant human input. In the CGM’s case, being incredibly diligent about entering a food and drink diary. In the case of WHOOP, tracking different activities and rest periods.
There can be unexpected negative consequences - while I undoubtedly developed positive fitness habits while using my WHOOP, it was to the detriment of my sleep.
While the promise of a longer life might sound like perfect motivation, unless there’s some short-term payoff, most humans struggle to build a new habit. Take my CGM example - the reality was that constantly thinking about what food I should or shouldn’t eat got in the way of enjoying it. The juice didn’t feel worth the squeeze.
There’s the challenge for all of us excited to take advantage of this fast-evolving technology. Whether you’re trying to improve health outcomes, improve productivity, or like me, create a copilot to help you grow your career, the success of these new products lies in whether they can quickly create value for the humans using them.
In that way, nothing ever really changes.