229: What are we not seeing yet?
Adopting AI en masse is about much more than technology
I had every intention of writing last week but choose bureaucracy instead, and wound up with a cold. Serves me right. It’s worth noting I also have certified professional grade cold-harboring skills – once a cold virus and I agree to tango, we do so with lengthy commitment. My body’s near endless capacity to produce mucus is perhaps unrivaled in the modern world.
Which also propelled me back to reading. I got 30% of the way into Andy Weir’s Project Hail Mary. I wonder which came first: The idea of a solar-eating algae and its dire consequences, or a singing alien creature without eyes, powered by mercury? After you’ve written about the mighty potato and its ability to fuel Martians, I suppose the ideas just keep coming.
But what make Weir’s storytelling fascinating, even popular, I think, are at least two things: 1) The lone engineer figures it out – this is the bedrock of Silicon Valley and probably even America. One guy (always a guy, sigh) and his wits versus immeasurable odds. He’s the little engine that could. And 2) He makes a lot of mistakes and gets tired – novel heroes, they’re just like us! Who among us wouldn’t meet an alien life form (quite literally, you are the first human to meet a space alien) then figure out enough common language so you can ask for a two hour pause and take a nap? As much as I’d love for the character of Grace to engage in a Zen and the Art of Motorcycle Maintenance shakedown, I bet there’s a good reason Weir doesn’t.
What science fiction does really well, however, is get us to ask important questions.
How long is it going to take the majority of us to adopt AI for real?
The speed and depth of our collective ability to adopt a new technology is getting faster, for sure. As much as I’m impressed with singular leaps made by lone AI-rangers, I tend to measure impact and success via mass adoption. At what point do 60%+ of us engage repeatedly, daily, reliably in the new thing? And yes, this time is different – this time is predicated upon all technology which came before; as well as all technology training and fluency. It compounds. How much faster might we unlock civilization-changing advantage?
Imagine we’re living in what’s now southern Germany circa 1440. None of us know how to read or write. We have no reason to. We don’t know anyone who knows how to read or write – well, maybe one of the monks does? Gutenberg’s invention arrives without a ready customer base, or even a ready social fabric. No surprise it takes hundreds of years for the technology of the printing press to have impact, but then, holy smokes – because more of us can read, we popularize new things like Mass Education, Libraries, Law, Contracts, Maps.
The Boston News-Letter started publishing continuously in 1704. While the residents were not universally literate, enough were. Over the next century U.S. literacy rates rise as urbanization expands, production costs decrease and then most American adults are engaging in the technology of newspapers. Because of newspapers we build storytelling empires, and my favorite industry: Advertising.
But wait, what if we removed the requirement to be able to read?
Radio commercializes around 1920. And it’s in ~60% of U.S. homes inside 15 years. Network TV moves even faster. Between, call it 1948-1955, roughly 65% of U.S. homes acquired TV sets and were watching daily. The unparalleled speed of TV’s adoption owes itself to one important characteristic: The human was inert. You didn’t even have to be in the room for this technology to have impact.
I like to say that after Al Gore invented the Internet it took about 20 years for +60% of U.S. adults to engage repeatedly, daily, reliably in interactive technologies – running somewhat parallel with the timeline for personal computers. But a week ago I saw an astronaut asking space command to remote-in to get one of his Microsoft Outlook accounts functioning; and I imagine we have miles yet to travel.
And if we peg social media starting around 2006, Pew data puts 60%+ of U.S. adults on social platforms daily by ~2016. And we have to incorporate broadband and mobile/smartphones in this broad tech milestone. All that to say this wave took a decade, more or less, to achieve impact.
Each of these milestones changed how society functions, how stories are told, how ideas are created and distributed (and by whom). Each technology reset the status quo – reset how idea people do their work, how their industries function.
What did each technology reveal we’d been blind to?
The printing press – revealed knowledge had been a protection racket. Gutenberg didn’t just spread books. He made the question, “Who gets to decide what’s true?” urgent and dangerous.
The newspaper – created a shared civic nervous system, for good and worse. Whoever controls the shared story controls what counts as real; and we’re still wrestling with that.
Network TV – revealed seeing is categorically different from reading about. Reading about Vietnam is different than seeing the body bags. What happens when authority can’t control the prevailing image? Still up for grabs.
The Internet – revealed the side effects of distance. If you no longer need to be near someone to transact, collaborate, or organize with them – what exactly were all these middlemen for? (Also, anonymity doesn’t reveal some truer self; it mostly reveals social constraint was probably doing more civilizing work than we thought.)
Social/Mobile – revealed we’re not built for this, but we want it anyway; also revealed friction in human interaction wasn’t just inconvenience – it was load-bearing. What happens to the self when it’s always being watched, including by itself? (Oh, and outrage travels faster than nuance by design.)
Each era of technology also removed a constraint
Print removed scarcity of text. Newspapers removed geographic isolation. TV removed the gap between distant event and felt experience. The Internet removed information scarcity and many gatekeepers. Social/Mobile removed time, space and status as barriers of connection.
So the right question for AI isn’t, “what will it give us?” It’s what constraints will AI remove? And what will we see, and see differently, as a result?
AI removes the scarcity of synthesis
Until now, making sense of complex subjects – across domains, at speed, at scale – required rare, expensive human expertise. Grace in Hail Mary is that prototype. So that constraint shaped how we built institutions, how we credentialed authority, how we decided who gets to have informed opinions about hard things. That setting also defined who got to think carefully about challenging problems.
Synthesis is pattern recognition across complexity. By contrast, wisdom is knowing which patterns matter and why, and what you’re willing to do about it. But, we let the scarcity of the first convince us it was the same thing as the second. Doctors, lawyers, consultants, professors – they were credentialed for synthesis and we extended them authority over judgment as a package deal, because we had no way to separate the two.
I think this is what we’re not yet seeing.
Consider medicine, siloed into specialties not just because bodies are complex, but because no one human could hold all that domain expertise singlehandedly. So we created the system where your cardiologist doesn’t talk to your psychiatrist and your GP is managing traffic between people who each know one deep thing. The silo isn’t a design choice. It’s a workaround for a cognitive bandwidth problem.
Law and Marketing did the same thing. The billable hour is essentially a tax on synthesis scarcity. Same with McKinsey. Same with the entire analyst class.
Art, Design and Content are experiencing similar evaluation. All the expensive ephemeral related to synthesizing nomenclature and definitions, practice history, and relevant case studies into concepts and especially layouts is essentially the toll we charge.
But.
Over time we’ve come to confuse synthesis (or coordination or bureaucracy) with wisdom, and built our authority structures accordingly. How much of what we called expertise was synthesis we couldn’t otherwise afford, and how much was judgment that actually deserved the authority we gave it?
AI splits that package open. AI asserts – those taxes and tolls for pattern recognition in complex domains no longer apply as much as they once did.
Right now, a small farmer in rural Minnesota has the same access to synthesized complexity about soil chemistry, commodity futures, climate modeling, and loan structures as an ag economist at Land O’Lakes, Inc. The farmer still lacks the time and maybe the framing. But the raw cognitive resource? It’s there.
So, what aren’t we seeing, yet?
How long is it going to take the majority of us to adopt AI for real?
I think we humans aren’t seeing our real value in this AI age – knowing which patterns matter and why, what we’re willing to do about them, and why we get to think carefully about these hard problems.
*The three non-AI photos in this post are by Tavepong Pratoomwong, a street photographer from Chanthaburi, Thailand. Learn more via 121Clicks, or Sony.