Login
Toward a Human Future for AI
I was invited by Janna Anderson and Lee Rainie, of the Imagining the Digital Future Center at Elon University, to contribute my thoughts to their latest study, titled Building a Human Resilience Infrastructure for the Age of AI: Experts Call for Radical Change Across Institutions, Social Structures, which just came out. Here is the full report, which runs 376 pages. I am generously sourced on pages 11, 16, 142, and 358. There is a lot of great stuff in the report, which I highly recommend. For what it’s worth, here is the full text of what I sent them.
We are digital beings in a digital world. That’s the main thing. And this world is still very new.
We’ve operated in the natural world for as long as we’ve been a species, and we are experts at it. But the digital world is not only new, but sure to be with us for many years, decades, centuries, and millennia to come. And we still lack countless graces we take for granted in the natural world, such as privacy and independence from algorithmic manipulation.
Making full sense of this new world is very hard because we understand everything metaphorically, and natural world metaphors mask what’s really going on in the digital world. So, while we speak of “domains” with “locations” that we “build” and “own” (though we only rent them), and speak of “loading” and “transferring” “packets” of data in “up” and “down,” data are actually collections of ones and zeroes that are by design immaterial non-things that are instantaneously both here and elsewhere, even though “where” only makes full sense in the natural world. How will all this change and make whole new kinds of sense after a few more decades of digital existence?
Progress is the process by which the miraculous becomes mundane. In the digital world, that transition is now happening almost instantly, and in many domains, because AI is endlessly useful.
Big AI does its best to ingest the totality of human expression in all digital forms, and then to make any and all of it available in the most useful ways it can. At the moment (for me, Noon in The Bahamas on February 2nd, 2026), it does this by bringing hunks of that expression back to us, on demand, in constructive conversational forms. Big AI is the world’s largest Magic 8 Ball, within which floats a polyhedron of answers with trillions of facets, each ready to help.
As with all tech, Big AI has its downsides. (Just ask Gregory Hinton or Gary Marcus.) But its usefulness verges on absolute, so we can’t stop using it, no matter how abyssal some credible prophesies may be.
But there is one saving upside. It’s the same one that saved us from HAL 9000 in the movie 2001: A Space Odyssey. It’s our humanity and independence. Specifically, in the form of personal AI.
We need personal AI for the same reason we need personal homes, shoes, and computers. We need it to know our natural and digital selves as fully as possible, and to participate with full agency in society, its economies, and its governance.
Think about all the data in our personal lives that is not in our full control, and could use some AI help: our schedules, our past and future work, our property, our finances, our obligations, our writing and correspondence, our photographs, our sound recordings, our videos, our travels, our countless engagements with other persons online and off, our many machines, you name it.
Truly personal AI—the kind you own and operate, rather than the kind that is just another suction cup on a corporate tentacle—is as hard to imagine in 2026 as personal computing was in 1976. But it is no less necessary and inevitable. When we have it, many of the questions that challenge us will have new and better answers. And new challenges.
Every form of life, from the microbial to the human, is fraught with challenges. Personal AI is necessary for us to meet and surmount our challenges in the digital world, and to answer all the questions posed to us in this very research exercise.
Amara’s Law says we overestimate in the short term and underestimate in the long. I’ve been doing both all my life, and in all my answers to good questions asked by Pew over the years.
Perhaps the most glaring example of short-term overestimation was my response to a request by The Wall Street Journal in 2012 to compress my new book, The Intention Economy, to a single cover piece for the paper’s Marketplace section. My editor at the Journal suggested writing about how the intention economy would look ten years in the future, which is three years ago as I write this. The piece I wrote was titled (by the WSJ) “The Customer as a God.” In retrospect, I was wrong. The economy I described still hasn’t happened. We are not gods in the marketplace. But there are encouraging signs, and I’m still sure my prophecy will prove out. Meanwhile, the first half of Amara’s Law applies.
I’ve been young for so long that I now have the life expectancy of a puppy. So I don’t expect to see personal AI or the intention economy prove out in my lifetime. But I am sure both are worth working toward, so that’s what I do. And I advise anyone wishing to make the world better to look for their best work to manifest somewhere beyond their own life’s horizons.
