AI Grows and Grows — but Still Can’t Run Your Vending Machine
More evidence that AI is getting better (as if you would need it!), but also still lacks…
Dear Friend –
Amidst the (nowadays) usual craziness of the world, CEOs are finally owning up to their expectation that AI and automation will, indeed, lead to layoffs… Ford’s CEO expects AI to replace “literally half of all white-collar workers,” and Amazon just reported that they deployed their one-millionth robot. How does the white collar workforce react? They are sending their AI notetakers to meetings instead of attending themselves. Maybe not the smartest move if you worry about your job… 🤷🏼
And now, this…
Headlines from the Future
Large Language Models Are Improving Exponentially ↗
Maybe all you need to know about where AI is heading…
—//—
Can Claude Run a Small Shop? (And Why Does That Matter?) ↗
Can AI run your business? Anthropic (maker of the Claude AI models) wanted to find out:
“We let Claude manage an automated store in our office as a small business for about a month. We learned a lot from how close it was to success—and the curious ways that it failed—about the plausible, strange, not-too-distant future in which AI models are autonomously running things in the real economy.”
The short answer: No. But there is a whole lot more to look at and learn from the experiment:
“It’s worth remembering that the AI won’t have to be perfect to be adopted; it will just have to be competitive with human performance at a lower cost in some cases.”
“An AI that can improve itself *and* earn money without human intervention would be a striking new actor in economic and political life.”
And it comes with a bunch of warnings/red flags:
“We do think this illustrates something important about the unpredictability of these models in long-context settings and a call to consider *the externalities of autonomy*.”
“In a world where larger fractions of economic activity are autonomously managed by AI agents, odd scenarios like this could have cascading effects—especially if multiple agents based on similar underlying models tend to go wrong for similar reasons.”
In summary:
“Although this might seem counterintuitive based on the bottom-line results, we think this experiment suggests that AI middle-managers are plausibly on the horizon.”
The whole study is worth perusing.
—//—
Microsoft Pushes Staff to Use Internal AI Tools More, and May Consider This in Reviews ↗
After the e-commerce juggernaut Shopify, the creator marketplace Fiverr, and the language-learning platform Duolingo’s respective doctrines from their founders enforced the use of AI across their workforce, tech granddaddy Microsoft is joining the fray:
“AI is no longer optional,” Developer Division President Julia Liuson told managers. […] “AI is now a fundamental part of how we work,” Liuson wrote. “Just like collaboration, data-driven thinking, and effective communication, using AI is no longer optional — it’s core to every role and every level.”
Aside from an overall question of how helpful AI truly is for certain tasks (depending on the day, you will see reports claiming massive productivity gains to none at all), it is a good idea to have everyone learn the ropes when it comes to AI. As Wharton professor Ethan Mollick pointed out, AI has a “jagged frontier” – its use is not necessarily intuitive, nor are the results always even; and the best way to learn what this jagged frontier looks like is to experience it.
—//—
Daytime Napping Fosters “Aha-Moments” ↗
New study shows that napping measurably leads to deeper insights and more “aha moments.” Take that afternoon nap – it will make you smarter!
Sleep EEG data showed that N2 sleep, but not N1 sleep, increases the likelihood of insight after a nap, suggesting a specific role of deeper sleep.
—//—
AI’s Consumer Tipping Point Has Arrived ↗
Menlo Ventures just published their latest AI report – assuming that their data is correct (as with all statistics, one ought to question the data, always!), what is going on in AI-land is truly bonkers…
Widespread Use: 61% of American adults have used AI in the past six months. This translates to an estimated 1.7–1.8 billion global users.
Daily Habit: Nearly one in five U.S. adults (19%) use AI every day, totaling an estimated 500–600 million daily global users.
But also:
Low Conversion: Only about 3% of AI users pay for premium services.
Revenue Concentration: General AI assistants (like ChatGPT and Gemini) capture 81% of the $12 billion in consumer spending.
ChatGPT Dominance: OpenAI’s ChatGPT accounts for approximately 70% of total consumer AI spend.
There is a ton more – here is the link to the report.
What We Are Reading
🐌 The AI Revolution Won’t Happen Overnight Silicon Valley is selling AI dreams at startup speed, but the corporate reality check is about to be brutal and expensive. @Jane
🤖 The Dawn of the Posthuman Age What do we get when we read demographic and technological megatrends together? A glimpse of a possible post-human future where work, creativity, and connection are all profoundly changed. @Jeffrey
🐍 Demon Copperhead by Barbara Kingsolver A modern novel inspired by Dickens’ David Copperfield, about an impoverished boy who’s fueled by grit and resilience. This book is worth a read before the adapted screenplay hits theaters. @Kacee
🎯 Engineered Addictions By now, it doesn’t come as a surprise anymore to learn that Silicon Valley is hijacking our attention. It does come at an enormous price, though – it’s time to change the rules of the game. @Pascal
Rabbit Hole Recommendations
The new skill in AI is not prompting, it’s context engineering
Reinforcement learning, explained with a minimum of math and jargon
Microsoft CEO Satya Nadella: The output of AI needs to be socially useful to justify its energy use.
Germany tells Apple, Google to block DeepSeek as the Chinese AI app faces rising pressure in Europe
Scientists invent photosynthetic ‘living’ material that sucks CO2 out of the atmosphere
Large CO2 removal potential of woody debris preservation in managed forests
Happy Distractions
👶🏼 This might have sounded like a good idea once upon a time: A brief history of children sent through the mail
😱 Frequent dreams and nightmares surged worldwide during the COVID-19 pandemic