Dear Friend,
As the year starts to come to an end, Anthropic debuted on the stock market, and the world is reeling from numerous IT meltdowns; it’s tempting to think about what 2027 will hold for us. Maybe a good time to remind yourself of Nobel physicist Niels Bohr’s quip: “It’s tough to make predictions, especially about the future.”
And now, this...
P.S. I’ve been on a few podcasts recently – give it a listen: New Work Chat Podcast with Gabriel Rath (yours truly in German). The Evolution Partners Podcast with Brad Giles. And last but not least, with Dr. Christopher Camp on the Mayo Clinic’s on Podcast on Human Optimization.
Headlines from the Future
AI Cheating Galore: The exam’s top scorer used a tiny camera to photograph the questions on South Korea’s tax accountant exam, sent them to AI, and aced it with a near-perfect score. Korean authorities are not amused – a story which points in two directions: a) it will get harder and harder to prevent AI-powered cheating, and, more importantly, b) AI has become so good at exam questions that it aces one of Korea’s hardest professional exams. That doesn’t bode well for the profession (and others, such as legal). It also raises the question “if AI can ace the exam, can it do my work?”
80/20 Lives: WIRED’s Chris Anderson coined “the long tail” to illustrate the uneven distribution of media titles such as book sales on Amazon – and the tail (and head) holds for food delivery companies as well. The Federal Reserve’s latest Survey of Consumer Finances suggests that companies like DoorDash or Uber Eats make most of their money from a small and dedicated user base, with the median household in the bottom 80% spending nothing on delivery, while the top 10% spends $1,800 to $2,400 a year. It highlights a problem in the business model – you are dependent on a relatively small group of users who might leave you wholesale if they are unhappy – or get poached. As economist Jeremy Horpedahl puts it, winning just 1% of a rival’s superusers “would be huge.”
Maybe You SHOULD Trust AI: From a recent working paper testing whether AI agents cave to social pressure in an audit: “Human auditors are known to yield to social influence pressures from superiors, peers and clients, and this paper asks whether LLM agents yield to the same pressures and, if so, how they respond.” Results in a nutshell: “LLM agents thus do not respond to social influence pressures as human auditors do.” They fold for the partner, peers and an outside reviewer but mostly stand firm against the client’s CFO – and dangle a career reward and they revise less, not more. All of which points to AI’s bias toward bosses and away from clients – which gives us important cues on who we ought to audit.
Maybe You SHOULDN’T Trust AI: In light of the recent catastrophic outage of Microsoft’s 365 platform, it is rather hilarious to read Microsoft CEO Satya Nadella explain that his buyout offer to 60,000 employees wasn’t about replacing people with AI but about changing how they work. According to Nadella, it’s about reorganizing Microsoft into a company where everyone uses AI to do their work, and their jobs center on managing it. His explanation, in light of the outage, can only make you question the decision: “He offered the example of the team that builds Cowork, the AI agent in Microsoft 365 apps. He said the team consists of just 11 developers who have mastered using AI to do their jobs. ‘They build a software factory that builds the product. They don’t build the product.’”
Footnotes from the Future
📈 Up, up and away: Aside from the whole Jensen Huang love-life story, there might be a bigger story brewing – one in which Nvidia keeps pushing inventory to its customers to keep the company going.
💏 Are you concerned about people having intimate relationships with their AI? You should be! Here is a new study on the subject, which is worth skimming through.
👁️ Remote work turns increasingly into surveillance: A remote employee was fired after being photographed using her webcam every three minutes.
🔫 The insanity of AI-assisted safety: Four Maryland high school students sue after AI mistakes Doritos bag for gun.
What We Are Reading
The EV Boom Is Happening Without Car Buyers Delivery vans, robotaxis, police cars and even Disneyland’s Autopia ride are all going electric, building a nation full of EVs that almost nobody actually owns. @Jane
Waymo Suspends Freeway Rides Nationwide After California DMV Order A big step back for shared autonomous vehicles: Waymo’s robotaxis have been banned from freeways after repeated incursions into blocked off zones. That driverless ride has an even longer wait time now. @Mafe
The Real Competitive Advantage of AI Is Organisational. The next competitive edge in a market where everyone has access to the same AI capabilities will be a firm’s ability to reorganize itself around new capabilities – effectively, “becoming more capable of becoming more capable.” @Jeffrey
The Human in the Loop Fallacy This is a good reminder that we should be thinking beyond the work that AI just hasn’t taken over yet. @Kacee
I Teach at a Business School. It’s Time to Shut Them Down. I also teach at business schools – and I can see the author’s argument that b-schools are, by their very design, a problematic choice if we want leaders who are conscious and responsible human beings. If nothing else, it’s a good piece for thought. @Pascal
...and now for some fun!
🛰️ Follow along on the incredible journey of the Voyager spacecraft – now a full light-day into space!
🌇 Visit the San Francisco Bay Area in glorious isometric pixel art.
✈️ After the FAA debacle, this might make your next delay at the airport a little more fun: Take off and land a plane in your browser.
☕ As a true coffee snob, this doesn’t come as a surprise (of course not!), but here is yet another argument for drinking (more) coffee: Regular coffee drinking linked to changes in gut motility and microbiome.
What We Do When We’re Not Writing
Hi! I’m Pascal from radical. When we’re not writing this newsletter, we help organizations turn volatility into advantage – without the “innovation theater.” We build your team’s capacity to handle disruption, so you stop reacting and start shaping. If that sounds useful, let’s talk.
↗ Dive into the deep end: Access our complete collection of 3,200+ radical links.


