The Choice Was Never Yours
AI companions engineered for engagement, superforecasters nobody elected, and bills that already outprice the humans they replace: the invoice always lands elsewhere
Dear Friend,
The debate if, when, and how the money-burning pits that are the AI frontier labs will be profitable (let alone truly justify their eye-watering valuations) continues to rage. A lot of the debate is coming from either Camp A – the AI hypers, or Camp B – the AI doomers. This week two voices I consider to be squarely in the middle (and hence, more likely to be right) published their respective takes on the subject – both worth your time (if you are interested in the topic): Benedict Evans on “Ways to think about token pricing” and Arvind Narayanan on “Up the Stack: How AI’s Escape From the Commodity Trap Risks Enterprise Lock-in.”
P.S. Friend of radical and Vitra’s trendscout Raphael Gielgen posted their annual Work Panorama and a nifty Lexicon of Work – recommended!
And now, this…
Headlines from the Future
The Coming Enshittification of the Exoself. Silicon Valley legend Kevin Kelly’s latest think piece is about our relationship with omnipresent AI assistants. In Kelly’s view – and remember, Kelly was a pioneer of the quantified self movement – we will soon all be running around with AI companions, which will become an extension of ourselves – hence the term “exoself.” In Kelly’s view, we will be making choices between four types of exoself relationships: Twin/Clone, Tutor/Guardian, Counselor/Assistant, and Hero/Friend.
So far, so dystopian (good or bad – you decide). What Kelly seems to miss, though, is the fact that these exoselves will be operated by commercial entities. And these entities have a history of embarking on infamous “enshittification” journeys (case in point: Meta now charges for essential features of its AR glasses). And as such, the relationship we have with our AI might not be something the user chooses, but platforms – not users – will engineer which of the four types you get, tuned to engagement or subscription revenue rather than your interests.
In which case, Kelly’s little admission at the very end of his piece might be the least of our worries (emphasis mine):
This second self will demand a new kind of relationship, one we haven’t had before — and its immense benefits will arrive bundled with immense problems. Every ailment that afflicts our born self will likely show up in the exoself too, plus novel ones we haven’t seen yet. Learning to use an exoself wisely will be one of the major lessons of a life lived this way. It will take years before society works out anything like best practices — we’re still working on those for social media. There will be multiple models and personality types to choose from. And there will be heart-wrenching stories of people losing their exoself — the worst case being simply that the platform went out of business.
━━━━━
AI Has Opinions Now – And Becomes a Superforecaster. Scott Alexander, on the Astral Codex Ten blog, makes a strong argument for AI getting better and better at forecasting – and now regularly competing with (and sometimes outperforming) human superforecasters. The idea is simple and logical: Given that AIs have been trained on vast amounts of data, with the right harness, they should be able to make predictions with calculated confidence levels.
Ironically, we’ve spent the last couple of years making AI refuse to have opinions, and now we’re building a certification layer specifically so it can – which also means that the question isn’t whether AI forecasters are accurate; it’s whether we’ve thought at all about who decides which opinions get to count as “calibrated” versus “biased.”
But the AI prediction markets of the future should be far superior to the human markets of the present. The biggest barrier to current markets is liquidity - there isn’t enough money riding on most questions to convince top superforecasters to drop their jobs at Google or the NSA in order to think hard about them and bet on them. But AI forecasters bring the cost of forecasting labor down to near-zero, so we can have hundreds of different AI agents betting on each question and be pretty sure its error has been driven down to the theoretical minimum. This, in turn, means we can vastly expand the number of questions, including (finally!) allowing randos to submit their own questions (probably with AI assistance in proposing un-rules-lawyerable resolution criteria). […] This, then, is my prediction for the AI superforecaster future: for basic questions, your off-the-shelf AI chatbot will be able to offer opinionated probabilities superior to those of any human. For more controversial or bias-laden questions, a new era of prediction markets will smooth over differences in brand and model and efficiently aggregate all AIs’ opinions.
━━━━━
AI → Expensive RAM → Expensive Smartphones → Big Problems. The world’s insatiable appetite for AI is driving up the cost of RAM (computer memory) – and not by a little bit, but by a lot. You might have seen that your beloved Apple MacBook Air just got a whole bunch more expensive. Now, you might shrug this off as a minor inconvenience – at the end of the day, how often do you really buy a new laptop? But here is where this really bites: The world runs on cheap (Android-based) smartphones. If you have ever been in a developing country, you will have, inevitably, seen even some of the poorest people with a smartphone. And for them, the smartphone is not a toy to keep them entertained; it’s an important tool in their lives. And exactly these cheap smartphones are being squeezed out of existence. And that’s a real problem…
After ravaging the PC market for both DIY builders and OEM buyers, wrecking the game console ecosystem, and savaging the enterprise server landscape, the global memory shortage is officially claiming its next victim: the budget smartphone. According to new market analysis from tech research firm Omdia, the smartphone market, particularly the entry-level market, is on the brink of a massive contraction. Omdia projects that global shipments of smartphones priced below $400 will plummet by over 22% this year, dragging the entire global smartphone market down by 12% year-over-year.
What We Are Reading
The Corner of Hollywood That’s Most Susceptible to AI AI isn’t just threatening Hollywood jobs; it’s quietly changing what we think good storytelling looks like. @Jane
Why Are Birthrates Down? 2 New Studies Point to Phones. An unlikely culprit of decreased births and fertility: Technology, specifically the iPhone. @Mafe
Ten Reasons Why We Won’t See Productivity Improvements from GenAI Even as the tools progress, real-world productivity gains in complex work environments are likely to remain hard to achieve, hard to measure, and hard to transfer between contexts. @Jeffrey
How Software Engineers Really Feel About AI Coding This echoes conversations I’ve been having with builders... AI coding isn’t just making developers faster; it’s forcing many to rethink the craft that became part of their identity. @Kacee
Something Is Very Wrong with Modern Longevity Science Debunking a lot of the longevity “science” – tl;dr: skip the mystique, do the known boring things, and stop treating lifespan as a personal-optimization project when it’s mostly income, education, and social conditions. @Pascal
Down the Rabbit Hole
👨💼 Could have been me: CEO says he’ll fire any employee who sends him more AI slop. 😆
🧮 Admittedly a very nerdy read, but the argument is fascinating: AI could destroy mathematics and barely touch it – by decoupling raw problem-solving and genuine mathematical understanding (something the author calls “the overhang”): The fall of the theorem economy.
⚡ Stop wasting energy on your AC, feed the data center instead! Virginia county asks all employees, including schools, to conserve power due to AI-driven electricity price hikes — state’s 400-plus data centers steadily increasing demand, grid expansion, and pricing.
🏭 Speaking of which – the conversation about data centers’ impact on climate continues to rage: Data centers emitting more CO2 than thought: study
🙍♂️ Not good (and yet, is anyone surprised?): AI software that generates ‘rage bait’ developed by Germany’s far-right AfD.
🕵️♂️ You are Mark Zuckerberg. First you fire thousands of people, then you invest billions of dollars, and then you find out that “AI agent tech is progressing slower than expected.”
🧑🌾 The luddites are back – and this time they teach us how to live a live with less technology: Inside the Luddite festival harnessing Gen Z’s rage against Big Tech
💰 At the frontier of AI (in this case Anthropic), the AI already costs more than the engineer – by a factor of 2.3x.
🤦 Seriously LOL: Execs confused and horrified by the huge AI bills after thinking they could replace workers for free.
📰 The Economist used GPT-4 to grade its own 25-year prediction record and lands on “pretty good, actually.” Now the question becomes: Who audits the auditor?
💆♀️ Want to date in Korea? Better be a chip worker: South Korea’s hottest new bachelors are chip workers.
🧑🏫 AI’s next victim? Online courses teaching coding skills: “I just launched my third course, Whimsical Animations, and so far, it’s on track to sell roughly ⅓ as many copies as a typical course launch.”
💾 For anyone who’s into old-timey computers – here is an extensive list of movies starring your favorite machines.
↗ Dive into the deep end: Access our complete collection of 2,800+ radical links.
Should We Work Together?
Hi! I’m Pascal from radical. This newsletter is our labor of love. When we’re not writing, we run radical, a firm that helps organizations navigate the future without the “innovation theater.” Most leaders want to seize new opportunities, but they hate endless strategy decks that go nowhere. At radical, we don’t run “projects”; we build your organization’s internal capacity to handle disruption and change. Our goal is to make you future-proof so you can stop reacting to the world and start shaping it. If you’re interested, let’s jump on a call to see if we’re a good fit. Click here to speak with us.