For about a year now, I’ve been muttering, with the casual confidence of someone who has absolutely no way to stop it, that when AI and quantum computing mature and inevitably merge, we’re looking at the beginning of the end of the human experiment. This article is Chapter 1 of that story — and it reads uncomfortably like the opening act of Terminator 2, where Sarah and John Connor race back in time to find Miles Dyson before he builds the microprocessors that become Skynet. Let’s hope the rest of this movie has a better story arc than the one currently forming in my head.

Article Excerpt:

Researchers have developed a method to reduce uncertainty in artificial intelligence (AI) systems by tapping into the power of quantum computers. They say their work represents the first demonstration of ‘quantum enhancement’ in a production-scale, pretrained large language model (LLM).

Article Summary:

IBM researchers have pulled off what they’re calling the first-ever demonstration of “quantum enhancement” in a production-scale, pretrained large language model — meaning they actually used a real quantum computer to improve a real AI system, not just run a simulation in a lab while wearing optimistic expressions.

The key metric here is something called “perplexity” (PPL), which measures how well an AI predicts what comes next in a sequence — essentially a proxy for how confused or confident the model is. Lower perplexity generally means better reasoning. The quantum-trained model showed measurable improvement over its classical baseline, answering questions correctly that the unenhanced model got wrong.

The especially unsettling part: the quantum boost required adding only a relatively small number of parameters. That’s the detail that should make you put down your coffee. It suggests this isn’t a brute-force quantum overkill situation — it’s an efficiency play, which means it scales.

This is early days. Quantum computing is still finicky, expensive, and not ready to run on your laptop. But “first demonstration” has a way of becoming “standard practice” faster than anyone expects. Chapter 1 rarely feels like the scary part until you’ve read Chapter 10.

Read the Full Article

The post AI + Quantum Computing = Yikes! (Chapter 1) first appeared on Sevenelles.

Two countries that can barely agree on the time of day are apparently being nudged — by at least one NYT opinion writer — to set aside their considerable mutual contempt and cooperate on the one thing that might actually end both of them. That’s either the most sensible geopolitical proposal of the decade or a setup for the worst buddy-cop movie ever made. Advanced AI is the kind of problem that doesn’t much care about trade deficits or territorial disputes when it’s rewriting the rules of everything. Strange bedfellows have collaborated before when the alternative was mutual annihilation. Whether these two can manage it while actively trying to kneecap each other on every other front is, to put it generously, an open question.

Article Summary:

Thomas L. Friedman, in a New York Times opinion piece, argues that despite deep geopolitical rivalry, the United States and China share a common and genuinely existential concern: the unchecked development of advanced artificial intelligence. The piece uses the framing of a potential Trump-Xi summit as a hook, suggesting that whatever else divides the two superpowers, AI governance might be the rare table where both have reason to sit down.

The core argument is familiar but underappreciated — that transformative AI poses risks that don’t respect national borders or political allegiances. An AI catastrophe, whether through misalignment, misuse, or an accelerating arms race with no guardrails, is bad for Washington and Beijing alike. The Cold War analogy lurks in the background: even the U.S. and USSR managed nuclear arms treaties at the height of mutual loathing, because the alternative was worse.

What makes the argument worth bookmarking is less the diplomatic prescription than the underlying premise: that AI may be the first genuinely species-level challenge where even rivals have aligned incentives, if they can stop posturing long enough to notice.

Read the Full Article

The post The U.S. and China Should Gang Up on AI first appeared on Sevenelles.

I have spent an embarrassing portion of my life optimizing things that did not need optimizing — researching the fourth-best blender on Amazon for three hours, then buying the third-best anyway out of spite — so Herbert Simon’s concept of ‘satisficing’ hits me somewhere tender. The idea is simple: pick something good enough, and move on. Simon wore the same socks, ate the same breakfast, lived in the same house for 46 years — and won a Nobel Prize, presumably because he had freed up his brain from deciding what to put in his bowl every morning.

Article Excerpt:

By making up one’s mind to invest in a choice, regardless of more attractive options that may come along later, ‘a great deal of energy gets freed up for living, instead of being spent on wondering about how to live.’

Article Summary:

Herbert Simon, Nobel laureate and pioneer of cognitive psychology, spent his career arguing that humans are not built for maximizing — there are too many options, too little information, and not enough brain to weigh it all. His answer was ‘satisficing’: consider a manageable set of options, pick the one that clears a good-enough threshold, and get on with your life. He lived it, too. Same socks. Same breakfast. Same house for 46 years. One beret at a time, sourced from a specific European haberdashery.

The research backs him up. Psychologists developed a maximizer/satisficer scale after Simon’s death and found, unsurprisingly, that maximizers are less happy, more prone to regret, and spend a disproportionate amount of time comparing themselves to everyone else. Their standard — ‘the best out there’ — is a moving target that makes contentment structurally impossible.

The problem is getting worse. Consumer choices today outnumber those of preindustrial societies by a factor of roughly 100 million. Social media functions as a permanent highlight-reel comparison engine. Dating apps have industrialized the ‘what if something better is out there’ loop. And AI, the article warns, may make all of this dramatically worse by optimizing everything and expanding the menu of options indefinitely.

The piece closes with a Murakami story — two people who are genuinely perfect for each other, who let doubt talk them into one more search, and who never find each other again. Simon would not have been surprised. The lesson: set a good-enough standard, stop when it’s met, and spend what’s left on something that actually matters.

Read the Full Article

The post Want to Be Happier? Give Up Choices first appeared on Sevenelles.

That’s social media right now. The whole thing. All of it.

A deeply thorough piece from Ars Technica this week serves as the most honest eulogy I’ve read for a thing that isn’t quite dead yet. which, if you think about it, is the perfect condition for a eulogy. Nobody writes a good one after the grief has settled.

Researcher Petter Törnberg from the University of Amsterdam has been studying the structural mechanics of social media. His findings are roughly: echo chambers aren’t a bug, they’re load-bearing walls. The toxicity isn’t caused by algorithms or human nature alone — it’s architecturally embedded. You can’t renovate your way out of it. You’d have to demolish and rebuild.

So here we are. Lingering in the doorway.

The Room Is Full of People Who Aren’t Really Here

What fascinates me is the texture of this moment. Not what’s happening commercially, but what it feels like to be a person inside it.

Because the few people I know who have not already left have one foot out the door. They’re still scrolling, still posting, still checking notifications with the compulsive regularity of someone touching a zit to see if it healed yet.

Törnberg’s data backs this up with uncomfortable specificity. Posting on Facebook, YouTube, and Twitter/X has declined sharply among actual humans. But the platforms don’t look empty because the bots moved in. “We don’t need the users anymore,” is apparently the quiet logic now. The platforms can simulate the bustle of a thriving public square, while the actual public has wandered off to AI conversations and rediscovered humans in real life.

Let that sink in for a second. You’re performing for an audience that is increasingly composed of things that are not people.

So Where Do We Go From Here?

Törnberg identifies three emerging replacements for the thing we used to call social media: private group chats (WhatsApp, Discord), algorithmically curated broadcasting platforms (TikTok, Reels), and people just straight-up talking to AI chatbots instead of each other. He notes — with what I imagine is the measured tone of a man trying not to scream — that roughly twice as many people are now talking to a chatbot as are posting on social media.

Twice as many.

We replaced the illusion of connection with something that doesn’t even pretend to be a person who give two fucks. And somehow that feels more honest, which is either progress or the most depressing thing I’ve written today.

The Long Irish Goodbye of Western Civilization

Here’s what I find genuinely strange about this particular threshold: we’re not grieving. Not really. If anything, most people seem relieved, in the same way you feel relieved when a relationship that’s been dying for two years finally, mercifully ends. You’re sad, sure. But mostly you’re just tired. And maybe a little embarrassed about how long you stayed.

The party metaphor keeps working because parties have a social physics of their own. There’s a tipping point where enough people leave that the remaining guests suddenly feel self-conscious about still being there.

Social media hit that tipping point a while back. The people still posting with full sincerity — the ones who still believe the room is real and the audience cares — increasingly look like they’re performing karaoke to an emptying bar.

Törnberg’s most unsettling finding, to me, is about what happens when the casual users leave and only the most partisan, most extreme voices remain. It’s the boiling-the-frog effect. The temperature rises so slowly that the people who stay don’t notice they’re being radicalized by the very act of staying.

Sound familiar?

What Comes Next Isn’t a Better Party

The tempting narrative is that we’ll find something better. Smaller, kinder, more human-scaled (I’m looking at you, BlueSky). A digital coffeehouse. A virtual pub where everybody knows your name, the conversation is civil, and the beer is metaphorical but ever so fucking satisfying.

Törnberg pours cold water on this. Private groups aren’t constrained by geography the way a real coffeehouse is. A WhatsApp group can tip into an echo chamber just as fast as a subreddit — faster, maybe, because there’s no public scrutiny.

So we can’t stay at the party. We can’t go home alone. And the next party might be worse.

Cool. Great. Love it here.

I keep coming back to this: it’s always been and remains performative. And now the performance is standing in the doorway with your coat on, saying “well, we should do this again sometime,” knowing you won’t, knowing they know you won’t, and doing the bit anyway because the alternative — just turning around and walking into the dark without a script — is terrifying in a way that small talk never is.

Social media trained us to perform the ideal version of ourselves. Now the stage is collapsing, but we’re still hitting our marks and running our lines.

The only honest thing left to do is put your coat on, step outside, and figure out who you are when nobody’s watching.

Which, of course, is the one thing the last fifteen years made sure we never learned how to do.

Read the Ars Technica Article

The post Are We the Last People at the Social Media Party? first appeared on Sevenelles.

There’s a scene in Pluribus— one of the most thoughtful series on Apple TV about a hive mind that absorbs humanity — where all the consumed humans are practically begging Carol to write another book. Not because they’re bored. Because they’ve lost the ability to surprise themselves. They’ve pooled every thought, every memory, every scrap of creativity into one enormous collective intelligence, and it turns out that an ocean of shared knowledge is about as creatively fertile as Oklahoma farmland in 1935.

I thought about that scene a lot while reading Bright Simons’ essay “The Social Edge of Intelligence” in The Ideas Letter. It’s one of the most carefully argued pieces I’ve encountered on AI — not because it screams that the robots are coming (they are) or that everything will be fine (it won’t), but because it identifies a dependency so fundamental that most of Silicon Valley hasn’t bothered to look at it.

Here’s the thesis, and it’s worth sitting with: AI doesn’t really think. It remembers how we thought together. And we’re rapidly creating conditions where we’ll stop giving it anything worth remembering.

The Devastation Is Real — Let’s Not Pretend Otherwise

The economic carnage ahead is not speculative. IBM announced plans to replace 7,800 roles with AI. Duolingo cut a tenth of its contractors. Klarna’s AI assistant now does the work of 700 customer service employees, and the company’s stated goal is to shrink its workforce below 2,000. Jack Dorsey wants Block’s headcount flat while AI carries the growth.

This is not a drill, and it’s not a blip. The internal logic is merciless: routine cognitive work gets automated, junior roles evaporate, productivity gains compound. For any board reviewing cost structures, it’s the cleanest investment case since the internal combustion engine retired the horse. There’s even a moral momentum to it — hesitate and you fall behind, and nobody wants to be the last company still paying humans to do things a model can do for pennies.

If you’re in your twenties right now, entering a workforce that’s being systematically thinned at the bottom, I won’t sugarcoat it: the next decade is going to be brutal. Entire career ladders are being pulled up. The entry-level positions that used to teach people how to think inside an organization? Many of them are already gone.

But Here’s the Thing Nobody’s Talking About

Simons describes a 2024 experiment where roughly 300 writers were asked to produce short fiction — some with GPT-4’s help, some without. On the surface, the results confirmed the AI hype: AI-assisted stories were rated more creative by independent judges. Individually, every writer got a boost.

But when the researchers looked at the full body of stories rather than individual ones, a different picture emerged. The AI-assisted stories were more similar to each other. Each writer had been individually elevated. Collectively, they had converged.

The researchers called it a tragedy of the commons. I’d call it something more plainly terrifying: the slow-motion extinction of surprise. We will end up with the hive mind from Pluribus. Desperately craving Carol’s next novel because it wants something it hasn’t already seen.

AI Eats Its Own Tail

Simons builds his case on a chain of research that’s hard to argue with. In 2024, a team led by Ilia Shumailov published in Nature that AI models trained on AI-generated data start to collapse. The distribution narrows. Minority viewpoints, rare knowledge, unusual formulations — the weird, edge-case stuff that represents actual intellectual diversity — vanishes first. What’s left is statistically average. Fluent, plausible, and hollow. Or, as Bruce Cockburn put it, “the Trouble With Normal is it only gets worse.”

Meanwhile, Epoch AI estimates that the total stock of quality human-generated text available for training will be exhausted between 2026 and 2032. Most people frame this as a resource problem, like running out of oil. But Simons sees something deeper: the reservoir isn’t just being drained. The springs feeding it are drying up.

Because here’s the feedback loop from hell: AI replaces human workers. Fewer humans doing complex cognitive work means less diverse, friction-rich human language production. Less diverse language production means less valuable training data. Less valuable training data means AI systems start to degrade. The technology is quietly consuming the very substrate it depends on.

We’re not just outsourcing tasks. We’re outsourcing the effort of thinking. And effort, it turns out, is where the interesting stuff happens.

Is the Thesis Airtight? Let’s Push Back

Before we get too comfortable with this narrative, it’s worth stress-testing it.

First, the synthetic data problem may be solvable. Researchers are already working on techniques to filter AI-generated text from training sets and to generate synthetic data that preserves distributional diversity. It’s not unreasonable to think that clever engineering could mitigate model collapse — at least partially, at least for a while.

Second, the “exhaustion of human text” timeline assumes current architectures and training methods. Breakthroughs in reasoning models, multimodal learning, or entirely new paradigms could change the equation. We’ve been surprised before.

Third, there’s an argument that AI could increase cognitive diversity by lowering barriers to entry — giving more people from more backgrounds the tools to participate in complex knowledge work. That’s not nothing.

But here’s why I ultimately find Simons’ argument more convincing than these objections: every counterargument assumes human behavior will remain unchanged in the presence of powerful cognitive offloading tools. And we have decades of evidence that it won’t. A Microsoft and Carnegie Mellon study of 319 knowledge workers found that in 40% of AI-assisted tasks, participants exercised no critical thinking whatsoever. Anthropic’s own research shows that users pause to double-check AI output only 8.7% of the time.

We’re not just outsourcing tasks. We’re outsourcing the effort of thinking. And effort, it turns out, is where the interesting stuff happens.

The Hope Buried in the Wreckage

So here’s where the Sevenelles brain kicks in — the part of me that refuses to let a clear-eyed assessment of reality collapse into the nihilistic need for another gin & tonic.

If Simons is right — and the research increasingly suggests he is — then the very things that make humans inefficient, frustrating, and expensive are also the things AI literally cannot survive without. Disagreement. Friction. The stubborn insistence of someone who sees the problem differently. The junior employee who asks the dumb question that turns out not to be dumb. The messy, ego-bruising, time-consuming process of humans actually engaging with each other.

The organizations that figure out how to use AI to create more human interaction — more debate, more cross-pollination, more productive friction — will be the ones that thrive. This isn’t wishful thinking. It’s a logical consequence of the dependency Simons identifies. If AI’s intelligence is a function of the social complexity of the civilization that feeds it, then protecting and enriching that social complexity isn’t a nice-to-have. It’s the whole game.

What This Means for You and Me

The transition is going to be ugly. Let’s not pretend otherwise. Millions of jobs will disappear before the correction kicks in. People will suffer real economic pain while executives learn the hard way that you can’t automate the source of your own intelligence.

But the correction will kick in. And when it does, the premium won’t be on people who can do what AI does — process, summarize, generate plausible output. The premium will be on people who can do what AI can’t: think in genuinely novel ways, hold productive disagreements, bring perspectives that haven’t been averaged into the training data, and do the unglamorous, essential work of keeping human knowledge diverse and alive.

Every consumed human in the Pluribus hive mind wanted Carol to write that book. Not because her prose was technically superior to what the collective could produce. Because she could still surprise them. Because surprise requires a mind that hasn’t been averaged.

That’s the edge the human mind holds. Not efficiency. Not productivity. The capacity to be unpredictable in ways that matter.

Read Bright Simons’ essay “The Social Edge of Intelligence” in The Ideas Letter

The post AI Will Devastate the Workforce. But the Human Mind Will Always Have the Edge. first appeared on Sevenelles.

In another great newsletter, my friend Curt Buermeyer summarizes Peter Diamandis’s three-phase roadmap from AI labor disruption to post-scarcity abundance. I’ll be honest — I want to believe it. Curt and I agree that the next five years are going to be rough. But where he sees a ten-year arc bending toward something beautiful, I see a timeline that assumes robotic construction, autonomous transport, AI healthcare, and vertical farming all scale and democratize faster than any infrastructure transformation in human history. But here’s the thing about being a pessimist: often you’re wrong, and that is a beautiful thing. Curt’s optimism is grounded and honest, not naive, and Diamandis has a track record of being right about exponential curves before the rest of us could see them. So I’m bookmarking this roadmap not because I’m convinced, but because I hope like hell they’re right and I’m wrong.

Article Excerpt:

I believe we’re facing a reality where roughly 40% to 50% of white-collar jobs—the knowledge work that college students and recent grads have spent years studying for—will likely be displaced almost overnight.

Article Summary:

Curt Buermeyer’s latest Smash Your Thinking newsletter summarizes a roadmap by Peter Diamandis (XPRIZE founder, exponential tech evangelist) for navigating the AI-driven collapse of the labor market without also collapsing the social contract. The core argument: the AI disruption won’t look like the Industrial Revolution, which reshuffled the workforce over 150 years. This one is compressed into a decade, and there’s no adjacent sector to retrain into because AI is eating every sector simultaneously.

Diamandis lays out three phases. Phase 1 (2025–2028): the fracture. Jobs disappear, identities crater, and the proposed floor is a $3,000/month Universal Basic Income paired with a 32-hour workweek to keep people attached to purpose and social structure while the shock absorbs. Phase 2 (2028–2031): the automation dividend. AI companies have extracted enormous value from public infrastructure — our data, our research, our systems — and the public deserves a share. Think Alaska Permanent Fund, but for the robot economy. Phase 3 (2031–2035): the great deflation. Housing drops from $2,000/month to $600. Transport falls to $225. Food, energy, and healthcare trend toward near-zero marginal cost, driven by robotics, solar, autonomous fleets, and AI diagnostics.

Curt adds his own caveat: he’s skeptical the Phase 3 deflation happens as fast as Diamandis projects. Prices in housing, healthcare, and food have never actually gone down for normal people, and he’d bet the timeline slips. The actionable close is straightforward: share the roadmap, audit what you’d do with your time if survival wasn’t the point, and pressure policymakers to build the bridging mechanisms now — before “the Valley” between 2026 and 2031 turns into something uglier.

Read the Full Article

The post Will AI Force US to Adopt Universal Basic Income? first appeared on Sevenelles.

There was a version of the ultra-wealthy that at least felt some obligation — real or performative — to the world that made them rich. That version is being quietly escorted out. What’s replacing it is a cohort of ideological mercenaries who’ve convinced themselves that hoarding capital and influencing elections is philanthropy, and that Warren Buffett passing around a philanthropy pledge card was somehow the real corruption. What happens when “giving back “is no longer fashionable? I’m guessing “not good things” , and we will all soon find out.

Article Excerpt:

Peter Thiel, the tech billionaire and a frequent Gates critic, said in an interview that he had privately encouraged around a dozen Giving Pledge signers to undo it. ‘Most of the ones I’ve talked to have at least expressed regret about signing it,’ he said.

Article Summary:

In 2010, Warren Buffett and Bill and Melinda Gates launched the Giving Pledge — a moral commitment for billionaires to donate more than half their wealth to charity. It was, briefly, the fashionable thing to do. Oval Office visits happened. Fortune covers happened. Over 250 families signed on, including MacKenzie Scott, Mike Bloomberg, and Sam Altman.

The vibe was: big capitalism and big philanthropy can coexist, and being seen as a ‘good billionaire’ actually mattered. Fast forward to now, and the whole thing is being treated like an embarrassing yearbook photo.

Signups have cratered — 113 in the first five years, down to just 4 in 2024. The Trump administration views the Pledge as roughly a punchline. Peter Thiel has been actively lobbying signers to bail, calling it an ‘Epstein-adjacent, fake Boomer club’ (again — the man has his own Epstein ties, so that’s a bold rhetorical swing).

One signer actually unsigned it, which the article notes is ‘without precedent,’ which tells you something about the current climate. The new dominant ideology among ascendant tech billionaires holds that philanthropy is basically a PR scam, and that the real gift to humanity is just making more money and letting it trickle somewhere.

Elon Musk has said his “businesses are philanthropy.” That sentence exists. Meanwhile, the Gates Foundation’s causes — global health, gender equality — are being actively dismantled by the administration that many of these same billionaires helped elect. Also worth noting: the Pledge has no enforcement mechanism whatsoever. It’s a moral commitment.

Which, given the moral inventory of some of its critics, may be precisely the problem.

Read the Full Article

The post Fuck Philanthropy. Our Billionaires Want More Toys! first appeared on Sevenelles.