As we continue to realize everything that we can do with artificial intelligence and large language models, scientists are taking aim at longevity.
Generally, humans have always wanted to live longer. But most efforts in this direction have been the stuff of science fiction. That is, until recently.
Researchers are making impressive advances into things like epigenetic programming, where scientists can, to some extent, actually turn on and off genes related to the aging process through something called “methylation.”
There’s also the wealth of information that AI can come through to show us patterns on who lives longer, and how to support centenarianism or other forms of longevity.
Aging and Disease
In fundamental ways, aging is different than a disease. It’s largely a system of measurement where we look at average lifespan and individual lifespan comparatively.
I liked this introduction to a talk that Nobel laureate Venki Ramakrishnan did at Columbia, where the writer attributes this idea to the speaker:
“Whether drinking from mythical springs of the Fountain of Youth or feeding on human blood as a vampire, cheating aging and death has long been the subject of fairytales. While immortality is still an impossibility, recent years have seen dramatic progress in the science of aging and adding healthy years to our lives.”
Ramakrishnan cites certain therapeutics like supplements and epigenetic science, as well as common sense interventions like good sleep, good diet and exercise. This is where AI can come in: by analyzing vast troves of information with wearables, LLMs will likely be able to tell us exactly what impact we are having on our bodies each day.
Being Real about Longevity Advancement
I also saw a recent TED talk by Matt Kaeberlein where he talked about the real benchmarks in solving longevity.
“I’ve seen a lot,” Kaeberlein said, detailing his experience over the years, including at the University of Washington, where he is a professor of pathology.
The longevity business, he said, faces a credibility challenge, partly because people aren’t clear on what can and can’t be done with modern science.
Are we close to solving longevity? Kaeberlein suggests that we are not. He points out how despite inventing medications like rapamycin, originally an immunosuppressant for organ transplants, and pioneering epigenetic programming, nothing has been as effective in longevity trials as caloric restriction in mice.
“One way to appreciate this is just to ask the question,” he says, “is this: if we look at our ability in the basic research world to slow aging, increase lifespan, improve health span, are we continually year over year over year, getting bigger and better? And the answer, unfortunately, is no.”
Epigenetic tests, he argues, are “not ready for prime time” as they can give wildly differing biological ages.
“What do they actually measure?” he asks, of the tests. “They measure average methylation status at specific regions in the genome, usually a few dozen. That’s what they measure. They don’t measure biological age. And then these specific regions where the average methylation status is quantified will correlate to something that doesn’t even correlate directly to biological age. Usually, depending on the clock or the algorithm, it’ll correlate to chronological age. It’ll correlate to mortality risk.”
Three Facts
Kaeberlein gives us three facts about longevity research, pointers toward the gist of the picture he is painting of our current efforts. One is that aging is, in his words, “a modifiable biological process.”
The second, proceeding from the first, is that anti-aging work with AI will be ultimately quite effective.
“As we understand the biology of aging, and we can, that’s going to be way better for human health than what we’ve done in the past,” he says, “which is waiting until people are sick, and trying to target their individual disease in isolation. The impact from targeting aging is orders of magnitude greater than targeting individual diseases.”
The third of his enumerated facts is that the scientific community now has the knowledge to make certain things happen.
“We don’t know everything, but we know enough today to have an impact on the aging process.”
Canine Research and More
For ballast, Kaeberlein talks about a study of 50,000 dogs, where scientists evaluated anti-aging interventions. He also mentions a term, “healthspan,” which includes this neat four-word mnemonic for drivers of lifestyle change: eat, move, sleep and connect.
However, he urges people to look out for “the noise around the signal,” suggesting that there are “charlatans” trying to take advantage of research news to make false promises. He cites the value of rigorous predictive biomarkers, and the role of supplements.
At the end of the day, though, Kaeberlein gives us some expectation that this kind of work is going to continue to take off.
As for right now:
“I think we are at the point where we have a set of medications,” he says. “They’re all FDA approved. I don’t know if they slow aging in people, there’s some reason to think that they could, but they absolutely can help prevent people from getting sick.”
Into the Future
Of course, there’s also Ray Kurzweil’s “singularity” theory, which holds that regardless of what we do with these human bodies, there may come a time when a kind of mind-meld connects us to digital consciousness. Absent this eventuality, though, we’re going to keep trying to keep our biological people alive. Stay tuned.
