Mind and Iron: The Forecast Issue
From AI assistants in our pockets to nuclear weapons on the battlefield, we've got the gamut of 2024 covered.
Hi and welcome back to another barn-burning edition of Mind and Iron. I’m Steve Zeitchik, veteran of The Washington Post and Los Angeles Times and Penny Lane of this Stillwater entourage.
(No? “Almost Famous?” I thought it was iconic.)
Every Thursday we sort through the noise to understand when science, technology and business are actually improving our lives and when they’re just messing with them. That’s super-important, but super not-cheap. Please consider pledging a few dollars to our cause here. You’ll get all the content when our paywall drops and know you’re sustaining an important mission to boot.
This Thursday we take a partial step back from reporting on the week’s news to look at what’s coming in the year ahead. 2023 was a time of disorienting change, from AI bursting into the mainstream to CRISPR therapies curing diseases. 2024 promises to be just as fateful.
With most of the year still stretched out ahead of us (hey, the last two weeks were news-packed), this issue will offer some plausible forecasts of where we’re headed.
-We asked a revered futurist to tell us what he expected out of the next 12 months.
-We take a (skeptical) look at a CES smash that could soon infiltrate our lives.
-And I’ll offer my own projection of four events that are absolutely, inviolably, take-it-to-the-bankishly happening this year.
First, the future-world quote of the week.
"To me, 2024 feels like a ‘chickens come home to roost’ year — issues and topics that futurists and pundits have been talking about for years finally moving from the potential to the actual.”
—Renowned futurist Jamais Cascio tells Mind and Iron where he thinks the next 12 months are taking us
Let's get to the messy business of building the future.
IronSupplement
Everything you do — and don’t — need to know in future-world this week
CES shows us the AI assistant in our pocket; a savvy crystal-baller tells us what’s ahead; four 2024 predictions you can take to the (piggy) bank
1. CES IS GENERALLY A PLACE OF GADGET FROTH, with either incremental advances on equipment that already works just fine or impossible hype that will never get to the real world. If you want to know what the future looks like, you’re better off just walking outside into the neon glow of capitalism — or to The Sphere.
But amid the smart toilets, sound-wave erotic toys and cars that go sideways on display at last week's Vegas jamboree, something else appeared: the Rabbit R1. The pocket-sized product is worth spending a minute on, because it comes with all sorts of big promises and implications.
Produced by an L.A.-based startup, the Rabbit R1 is an AI device that smartly liaises with your phone — or, really, with all the things your phone can do. Its premise is that every message we send, delivery we order or plane-ticket we book can be handled by one simple device that we just speak to — Joaquin Phoenix from "Her" on steroids.
"Our mission is to create the simplest computer — something so intuitive that you don't need to know how to use it," founder Jesse Lyu told the assembled. (You can watch his sales pitch here.)
Rabbit employs something called a "Large Action Model." You've heard of Large Language Models, which basically scour a mega-trove of text to figure out how to write that long-overdue apology letter to your 10th grade teacher? This also scours, but to figure out how to accurately take actions on your behalf.
So: speak in natural language and the hockey tickets are ordered, the restaurant is booked, the Target shopping cart is filled and on its way to you.
How much the R1 will actually learn our preferences and autonomously make decisions for us is unclear. But at the very least it’s doing an extreme amount of filtering that cuts out much of the information we see and buttons we tap. If it’s not thinking for us, it’s eliminating a lot of our need to think.
Rabbit says it has just sold out 20,000 (!) preorders of these devices at $200 a pop (they arrive in spring), with many users lining up to buy the next batch. The YouTube influencer AI Revolution hailed it as something that “might change the way you use technology forever.” Hype hype hooray.
The Verge’s comment board is currently a hothouse of whether the R1 is the future or fool's gold.
Dmitri: "If they’ve cracked ‘Just talk and it’ll happen’ (I believe they have NOT, but that’s a different story), then the mode of use will be very compelling. Like the iPod was."
My Name Is Not Bob: “They’ve done nothing to push back on the wild speculation. Their presentation just fed into it. They are being just vague and suggestive enough to feed into the hype. Disappointment is going to be the only result.”
leedSt: “So many on here trashing some ambition and wanting only a select few companies to deliver the same reheated glass slabs every year. Rabbit have come up with something a bit different, something exciting and I think it’s great. More please!” (A device like this might actually one day get merged with a new iteration of an Android or iPhone but that’s for later.)
Arjemon: "Ah yes, such an innovative concept: a box that does like a third of the things a smartphone does, but filtered through tech's dumbest party trick."
DarKnightofCydonia (is an AI coming up with these names?): “The keynote was crazy impressive but I'm happy to wait until this gets properly released and reviewers can shed light on limitations.”
On the Lifewire site, a reviewer described the R1 as "how you might imagine a pocket computer if you had never seen a smartphone. You just tell it what you want it to do, and it does it."
Certainly the idea of reducing the friction in our daily lives seems welcome. If Uber, Airbnb, Stuhub and DoorDash fundamentally allowed us to access a whole new economy from our pockets, this would enable all of that without the searching, scrolling, tapping, clicking, waiting and crashing that we accept as the cost of doing its business. (Not to mention weed out mistakes and problematic listings.) On-demand culture is a major misnomer given how much demanding we actually do.
Of course, the naysaying commenters are right — there ARE a slew of obstacles, from whether the R1 could really correctly carry out what we want to how the apps get paid (and how much we might have to pay) to whether a program can do all of this without hallucinating. (Verge commenter Vogon5: "When an image generator hallucinates, you get nonsensical imagery...What happens when an action model hallucinates?? It orders the wrong pizza? Orders 20 of them? Instead of calling Pizza Hut calls the cops?") Like so much new tech, there's the massive question of whether this will work.
More interesting, though, is what if it does work — what happens then?
If a device can initiate and complete our everyday tasks pretty much as well as we can, should we want it to? If these 20,000 orders of the R1 become 200 million and we all walk around with dedicated AI Assistants in our pockets, what does the world look like?
This is where the forecast gets muddy — and fascinating.
First come the privacy concerns. They're not small. What a single company might do with such a detailed tsunami of information about our lives is anyone’s guess, no matter what reassurances they provide.
But even if regulation and (ahem) responsible corporate behavior can get the privacy risks down to non-toxic levels, is all this to the good? Do the benefits that come from machines executing these scores of small everyday decisions outweigh what we'd have to give up in return?
Or let’s spin it around: Do the actions we currently take in these Uber and DoorDashy regards make our lives meaningful — give us a feeling of agency — that we shouldn’t be in a rush to give up? Does the serendipity that can happen when we carry out these choices ourselves have value? Or are such thoughts a myth, an elaborate self-justification for our daily humdrummery, and the sooner we ditch these tasks the better?
A Verge commenter named apparitchiki made a forceful case for the former on the subject of restaurant apps: "I like to take a look at different options with whomever I'm at the moment (usually my gf), we look at current deals, ratings, delivery times (It's a quick 5 min process but we do like to see a list of whats appealing or available, new places come up sometimes), I don't want a LAM picking the best one for me."
Is a sentiment like this self-evident or destined for history's dustbin?
At heart the R1 is one more example of how tech, in the name of comfort and efficiency, promises to make our lives better without calculating the costs of these improvements. Over the last century this question has underlied pretty much every advancement. And been answered in every direction.
Few of us, for instance, would argue that the TV remote control was unwelcome. "What happened to that grand old feeling of getting up from the couch every time you want to change the channel" is a sentiment even your Facebook-loving uncle isn't expressing. But at the other end of the spectrum comes a bunch of innovations that in hindsight proved to be disastrous. PFOA's immediately come to mind. Easier cooking obviously isn't worth deadly toxins.
The Rabbit R1 falls between these poles. Perhaps a grandiose analogy applies: the gas-powered car, an innovation that ushered in a spectacularly complicated set of effects. The idea of a machine that could transport people great distances with no physical effort made the world a dramatically smaller and more accessible place, an undeniably humanity-improving boon. But it also made for massive health and environmental fallouts still felt to this day.
We are now, in a way, beginning to do the same for mental effort. And the truth is that there's no way of knowing until it's operationalized just what effect this change will have. The idea of a machine in our pocket doing something for us that we're only vaguely aware is being done seems to fall into that automobile-ish crevasse of uncertainty. It might strikingly improve our lives. Or it might slowly leech off our humanity, leaving us to hop around like directionless bunnies.
2. IF YOU WANT TO KNOW WHERE WE’RE HEADED, YOU ASK A FUTURIST.
AND IF YOU WANT TO KNOW IF WHERE WE’RE HEADED IS A GOOD PLACE, YOU ASK JAMAIS CASCIO.
One of the country’s most decorated forecasters (Foreign Policy once called him “our moral guide to the future”), Cascio boasts an impressive resume. He is a veteran of Nick Bostrom’s think-tank Institute for Ethics and Emerging Technologies and the founder of Open The Future, which examines environmental and other “foresight scenarios” to figure out which curves we should steer into.
I asked him to contribute a few thoughts on what awaited us in the year ahead, across a range of areas. Here’s what he wrote:
“To me, 2024 feels like a ‘chickens come home to roost’ year. Issues and topics that futurists and pundits have been talking about for years finally moving from the potential to the actual. 2023 already saw the shift from ‘this might happen’ to ‘this is happening’ for CRISPR, AI, and Climate Impacts (in the broader public mind and, importantly, for policymakers; specialists and people who pay attention have known these are happening for quite some time). These three will only get bigger and more complicated, of course, but I suspect that they’ll be joined by:
• Unintended consequences of EV adoption, such as road damage from heavier vehicles combined with loss of road maintenance funds from gas taxes, power grid not being able to support charging surges around holidays, lack of plans/preparation for large-scale snowbound EV traffic during Winter (the I-95 jam a year ago lasted for over 24 hours), pushback on EVs as electricity rates climb (especially in California), etc. Probably not at overwhelming levels, but visible to those who are looking.
• Political violence in the US, which I believe at worst will seem like the “Troubles” in Northern Ireland far more than like the US Civil War. I’m unsure whether this is more likely if Trump loses than if he wins. There are all sorts of wildcards that could make this situation even more precarious, such as Biden dying or becoming incapacitated prior to the election (but too late for a primary race) or Trump dying/becoming incapacitated, for that matter, blatant election tampering by Republicans in various states, and/or some kind of big event (e.g., major earthquake in CA, major terrorist (domestic or international) attack, a renewed pandemic surge) that distracts attention. The pieces have all been in place for some time, just waiting for the right source of ignition.
• It’s possible that we’ll see the first serious, high-level discussions of the use of geoengineering. No actual implementation yet, but diplomatic and practical efforts to start working through the details of how it could be done, legal issues, etc. Massively controversial when the conversation starts. I’m not definite on it happening in 2024, but very very likely by the end of the decade.
• Even more uncertain, but distressingly plausible, is the use of nuclear weapons in war. Not a full-on World War III style nuclear exchange, but an act of desperation using just one or a small number of relatively small devices. It could be a non-state actor using one (e.g., against Israel), or it could be an “escalate to de-escalate” situation where Russia or China sets one off in the midst of increasing tensions to demonstrate resolve. Probably not North Korea, at least not yet. I don’t think that Modi is willing to do so against Pakistan, but there are some on the Indian right-wing who would at least consider it.
This last one is more gut than analysis, but the detail that there really aren’t any major global figures left who were even teenagers when the bombs were dropped on Hiroshima and Nagasaki kind of sticks out for me — there’s nobody left (with any influence) who saw or had the visceral reaction to the use of these weapons. In some ways, the popularity of Oppenheimer might have made things worse — associating the massive destruction of a nuclear weapon with Hollywood.
Wow, these are all awful. So add to this: human tests of treatments for specific cancers that have a near-100% success rate; the development of “social tune” apps to help people be more socially appealing online (largely by making them less hostile); and a very successful novel treatment for chronic kidney disease in cats.”
Whatever the German word is for being at once heartened and scared.
3. OK SO HERE WE GO, FOUR THINGS THAT ARE SURELY, UNDENIABLY, DEFINITIVELY GOING TO HAPPEN IN THE NEXT 12 MONTHS. (Please misplace this newsletter by October.)
1. AI deepfakes turn the U.S. presidential election into a hot mess.
This is a horrible time, democracy-wise, for where we sit with AI deepfakes — the nightmarish in-between of having the tech tools that can cause mass confusion but not the sophistication to counteract them. What has been a conceptual worry becomes a sloppily real problem in 2024, right in the middle of choosing the leader of the free world. So which subverting force weaponizes this and who do they depict and for what purpose? Yes.
2. AI Medicine is a ferocious battleground.
A new chatbot called AMIE that in early studies diagnoses patients better than human doctors becomes the tip of the spear for a massive multi-party fight involving researchers, regulators, clinicians, insurers and other powerful parties. We’re not going to resolve any of this in 2024 — not even close. But this year, trials for AI-discovered drugs will be conducted, breakthroughs will happen, hypesters will hype-ify, stakeholders on all sides will scream and government agencies will be caught between the criticisms that they’re regulating too much and not enough. While patients will have their heads spin amid the competing claims that AI either could ruin their lives or save them.
3. AR/VR Glasses start to take off.
No, we didn’t lift this one from eight different prediction lists over the last two decades. And it’s not April 1. But starting tomorrow users can pre-order the Apple Vision Pro, the company’s mixed AR/VR headset (which officially drops next month). Call it the soft launch of a new era. Yes, there will be fits and starts. Many of them. But by this time next year, many more people are putting on these and other headsets — watching, playing, scrolling, shopping. Get ready for our view of the world to be overlaid with a lot more screen data. And for people to walk into walls.
4. Americans finally start to climb on board high speed rail.
Across the country, the cost overruns are spectacular and the political interests entrenched. But 2024 is finally the year when we see high speed rail get traction with Americans, making inter-city travel environmental — and relaxing.
This is partly due to Brightline West — a private outfit that already has a slower train running between Orlando and Miami — saying that after a $3 billion federal grant it’s close to breaking ground on its Vegas-San Bernadino County route. (The ambitious goal is the 2028 Summer Olympics in LA.) And it’s partly because the project could create jobs for more than 10,000 people. Nothing breeds popularity like employment.
The Mind and Iron Totally Scientific Apocalypse Score
Every week we bring you the TSAS — the TOTALLY SCIENTIFIC APOCALYPSE SCORE (tm). It’s a barometer of the biggest future-world news of the week, from a sink-to-our-doom -5 or -6 to a life-is-great +5 or +6 the other way. Last year ended with a score of -21.5 — gulp. But it’s a new year, so we’re starting fresh — a big, welcoming zero to kick off 2024. Let’s hope it gets into (and stays) in plus territory for a long while to come.
A CES SENSATION HAS OUR TASKS EXECUTED BY A NEW-AGE PHONE: It could be ground-breaking — if it can actually work. +2
A RENOWNED FUTURIST SAYS POLITICAL DIVISIVENESS AND OPERATIONALIZED NUCLEAR WEAPONS AWAIT BUT SO DOES SOME PRETTY LIFESAVING MEDICAL RESEARCH: Big developments, (nearly) offsetting. -1
ELECTION DEEPFAKES (-7), AI MEDICINE BATTLES (0), AR GLASSES (+1.5) AND HIGH SPEED RAIL (+2) ARE ALL ON THE 2024 HORIZON: -3.5