Bertrand Duperrin

• Earlier this month, former Google CEO Eric Schmidt was met with jeers when he brought up AI during his commencement speech at the University of Arizona
• ootage of real estate executive Gloria Caulfield being booed at her commencement speech at the University of Central Florida after mentioning AI went viral online.
• “What happened?” Caulfield asked the raucous crowd, incredulous. “OK, I struck a chord! May I finish?
• , the two incidents perfectly highlight massively growing backlash to the controversial tech, with millions of students who are about to enter the workforce
• Ronny Chieng sang a dramatically different tune during a profanity-laden commencement speech at Harvard’s Class Day event this week.
• “I’m here to tell you the mission of your generation is to destroy AI,” he told a far more receptive crowd.
• Chieng addressed ongoing concerns that AI may lead to atrophying skills,
• “I’m talking about the accumulation of cognitive debt due to excessive use of large language models… This is why you should be scared of AI.”
• “It’s going to be people with substance versus people with shallow knowledge. It’s going to be mastery versus faking it. It’s going to be people with good taste versus tacky.”
• University students across the country are starting to speak out, arguing that the tech is being hoisted on them against their will
• Put simply, they refuse to be replaced by machines as executives continue to celebrate AI as the next industrial revolution.
• “Creating is the fun part,” he said. “Why would I want AI to take that away from me?”

10 more annotations...

• AI is like religion. Either you believe it changes everything, or you don’t believe at all. There is no moderate position; nobody believes in AGI “more or less,” just like nobody is “casually religious.”
• In 2026, the four largest cloud and AI infrastructure providers—Amazon, Google, Meta, Microsoft—committed to spending $670 billion on CapEx
• Apple, in contrast, spent $12.7 billion on capex last fiscal year and projects $14 billion for 2026, 2% of what its peers are spending.
• Siri has been a punchline for years—an internal executive called the delays ugly and embarrassing—and critics say that Apple has not been the same without Steve Jobs. It is falling behind, they say, and moving way too slowly for AI.
• Apple is the most powerful tech company in the world right now because it’s acting according to what it believes.
• Cook could have named someone who would reorient the company around the most hyped technology in a generation. Instead, he picked another non-AI guy.
• In March 2026, Bloomberg reported that Apple is opening Siri to third-party AI models.
• Apple is building a menu that lets you choose which AI powers your assistant rather than get lost in a zero-sum race with fake believers and actual advocates
• u do this if you believe AI models and agents are swappable; a commodity that will play out like electricity, not a divine machine.
• The AI industry is playing cool. And right now, roleplaying as a fanatic is the cool thing to do (meaning: it is what keeps the revenues coming, shareholders happy, and valuations intact). The moment the wind changes, we will see their true colors. But for now, let’s look at the data to compare.
• Meta’s primary consumer-facing product is a chatbot that lives inside WhatsApp.
• Google spent ~$93 billion on AI last year and then embedded Gemini into Gmail, where it suggests slightly worse replies, and into Search, where it makes up responses.
• Microsoft CEO, will invest in OpenAI before anyone else, but then forces its Office clients to stick to Copilot.
• Elon Musk will build a giant AI datacenter and then rent it at a profit to a competitor.
• The entire AI ecosystem is made up of entrepreneur types trying to inflate the pie to take a portion before it explodes
• Even Sam Altman. Sam Altman doesn’t believe in anything. Or rather, he believes in everything at different times, whatever the public wants to hear, he obliges. And so on.
• They are bolting AI onto existing products the way they once bolted social features onto existing products.
• AI is, if lucky, a new dot-com boom.
• They believe what Apple believes but behave like cowards; if they genuinely believed God was watching them, they wouldn’t hedge.
• They can’t prove AI will be transformative. But if it is and they don’t invest, they’re dead.
• . However, if it isn’t transformative and they invest, they only lose some cash they will recoup anyway by doubling down on ads
• Apple is willing to let its actions match its actual beliefs. That’s a stance I can respect. The rest are performative sheeple overdrinking from the fountain of fear.

20 more annotations...

• In May 2025, Dario Amodei, the man who founded Anthropic, went on Axios and warned of an AI ‘white-collar bloodbath’ and that half of all entry-level office jobs will be gone, and unemployment will rise to 20%.
• Amodei believes that 10% of the laborers will be doing 90% of the jobs, output grows exponentially, but because AI will outgrow the speed of achieving Jevons paradox, we should expect:
• Basically, he argues that AI advances faster than the economy can absorb it, leading to an era of total chaos.

• To do this, let’s rewind to a time most of us forgot: when money was nearly free.
• Not just Meta, but every company with sufficient revenue to borrow saw the same thing: cheap capital, a pandemic that seemed permanent, an online-everything boom, and executives convinced that the surge in demand would stick around.
• When the pandemic-era supply shortages met a sharp rebound in demand, while energy prices jumped after Russia’s invasion of Ukraine, causing inflation
• Shortly, CEOs could no longer afford the commitments they’d made to white-collar employees,
• Many people predicted this would be a permanent acceleration that would continue even after the pandemic ended. I did too, so I made the decision to significantly increase our investments…”
• “We've cut costs”… and “We're restructuring teams to increase our efficiency. But these measures alone won't bring our expenses in line with our revenue growth, so I've also made the hard decision to let people go.”

• The first message tanks your share price; the latter gives your shareholders peace of mind.
• 4 years and counting, I’ve been tracking AI (especially LLM) adoptions. I’ve seen only a handful of successful implementations
• List out a job you care about (like yours); e.g., a software engineer breaks down into 10, maybe 15, discrete tasks
• Yes, AI can speed up coding, but this is a task, not a job
• The success of AI implementations is currently stuck at the task level.
• When questions get messy, the answer with the fewest assumptions is usually the answer
• While most companies claim they lay off because of AI, as the chart shows:
• The untold truth is, in the same report, 59% also admitted they emphasize AI in their layoff announcements because it plays better with stakeholders than saying “we have a budget problem.”
• Another firm that has counted American layoffs for decades logged about 1.2 million job cuts in 2025. The number of roles pinned on AI: only ~55,000. Under five percent, ranked behind government cuts, behind market conditions, and also behind plain restructuring.
• Most economists would file this under “healthy, slightly tight labor market,” rather than “crisis”
• Take Marc Andreessen (cofounder of Netscape, then a venture capitalist), who called AI the “silver bullet excuse,” and reckoned most large companies are 25 to 75% overstaffed, left over from the COVID years.
• “Businesses have a strong incentive to talk about layoffs as if they were caused by AI. Talking about how they’re using AI to be far more productive with fewer staff makes them look smart. This is a better message than admitting they overhired during the pandemic when capital was abundant due to low interest rates and a massive government financial stimulus.”

• We need to start by asking,

  why does a job even exist?
• A job exists because someone, somewhere, feels a pain sharp enough to pay to make it go away.
• The pain couldn’t be fully resolved by this bulky machine. So teller numbers didn’t collapse. They shifted to more on-site banking.
• Therefore, what ended teller jobs wasn’t ATMs but something else that provides a complete solution: a much better experience, at a lower cost, more accessible, with software that can be updated easily.
• So “will the job disappear?” was always the wrong one to ask, try “what did customers actually need? And will this new technology make the journey to satisfy the need smoother?”

• Run the following task to evaluate your own work.

  1. Break down your job into tasks,

  2. write down the pain your role exists to remove,

  3. finally, look at where AI actually is in relation to that pain.

26 more annotations...

• Classic AI’ and machine learning have already delivered significant value in our core operations, optimizing supply chain forecasting, sourcing, and in-store productivity.
• Now, gen AI is unlocking new capabilities. It is transforming how we manage our product information at scale, for instance, by automatically generating product descriptions and analyzing documentation from thousands of sellers
• On the consumer side, we are using gen AI for visual inspiration.
• Fundamentally, AI is an incredible enabler, but only if you have a solid digital foundation. AI will widen the gap between frontrunners and laggards, leaving companies that missed the first wave of digital transformation even further behind
• We have an “AI Radar” where any team can propose and test new ideas. With over 250 use cases now, we let this innovation happen organically to not block ideation. Once a use case proves its ROI, we centralize it, productionize it, and scale it across the group.
• At a foundational level, we have a global contract with Google, so every employee has access to Gemini and we have an internal chatbot for our stores that provides access to company data.
• Fully autonomous commerce feels like a threat, not a genuine opportunity for retailers
• Fully agentified commerce will remove our ability to cross-sell or upsell, turning us into a simple delivery company in a price-driven red ocean.
• The biggest challenge is not the technology; it’s the impact on our governance and people.
• For years, teams would debate which product to feature first on a webpage.
• The reconfiguration of these roles and responsibilities is the most difficult thing to manage.
• When an AI-driven pricing model goes wrong, who is accountable? The tool cannot be.
• we will go through a period of disillusionment. My job is to keep everyone grounded and focused on creating real value, not just chasing the hype.
• There is so much hype, and people see AI as a magic tool that can solve problems that have existed for years.
• I am focused on achieving consistency over time, especially as we move from experimenting with models to embedding them in long-term applications that require stability.

13 more annotations...

• For a while now, executives have raised eyebrows by justifying sweeping layoffs by arguing that AI had made thousands of roles redundant. But as reality settles in and the tech’s real-world capabilities are coming into focus, some in the industry are starting to sing a dramatically different tune.
• “The narrative that connects AI to job loss, for many of the CEOs that are doing it — it is just too lazy,” Huang told Channel News Asia. “AI has just arrived, how is it possible they’re already losing jobs?”
• “How is it possible that AI became productive and useful only six months ago, and they were somehow laying people off two years ago because of AI?” he added. “It doesn’t make any sense.”
• “It was just a way for them to sound smart and I really hate that,” the CEO argued. “I think we’re scaring people, and that’s irresponsible.”
• We’ve long suspected that CEOs have been trying to mislead investors by claiming that much human labor was simply no longer needed in the age of AI.
• Instead of a dramatic rise in productivity, what’s far more likely is that companies are draining their pockets by making enormous investments in AI.
• “It’s more likely that the companies with ambition will be more productive, they will do things faster, their company will increase in velocity,”

• “Of course, they’ll use more AI, but they will also hire more people,” he added.
• It’s a shift in messaging. Just last year, Huang warned in a CNN interview that “if the world runs out of ideas, then productivity gains translates to job loss” and that “everybody’s jobs will be affected” while “some jobs will be lost.”
• Just last week, Google DeepMind CEO Demis Hassabis accused leadership at other companies of a “lack of imagination” for blaming layoffs on AI.
• “This is the people selling the shovels telling the miners to stop blaming the shovels for the cave-in.”

9 more annotations...

• As AI accelerates individual execution, a new bottleneck to progress is emerging: managers. Traditional management was designed for a world where execution took time. You could delegate a task and then comfortably check in the following week.
• But AI has collapsed that timeline. Increasingly, the pace of progress is limited by how quickly managers can offer feedback. “If I discuss a project with a direct report in a 1:1,” a manager shared in an interview we conducted for this article, “It’s already shipped and running the next day.”
• Research from Atlassian, where Liz works, shows that 89% of leaders agree that AI has accelerated the speed of work, creating an always-on review environment
• AI has created abundance, meaning it allows teams to pursue more opportunities and ideas, but has also skyrocketed feelings of scarcity, as the new pace strains time, energy, and attention.
• The Atlassian research shows that 87% of knowledge workers say that with everyone in execution mode, teams lack the time or capacity to coordinate. That makes it even more crucial for managers to ensure their teams are aligned.
• That means clarifying your mission, objectives, and key results.
• “Driving accountability for metrics has been the single most useful principle for ensuring that my teams stay focused on the right work.”

• Before moving forward, verify that every one of your directs can answer four questions:

   
   
  • What is the core problem we are solving?
   
  • What specific change are we driving?
   
  • How will we measure success?
   
  • Who is responsible?
   
   
• According to a recent BetterUp survey, 54% of managers report receiving AI workslop, content that appears polished but lacks real substance.
• “I have a rule that my team cannot send me something AI-generated that they didn’t read or edit,”
• “Before you work on a full draft, I’d like to review a half-page outline to ensure we’re on the right path,” or “I need to approve the final version of anything executive-facing.”
• Fernando Garcia Valenzuela, Head of Engineering, Atlassian Cloud Storage, built an agent that scans his DMs with his direct reports. Every two weeks, it generates a summary flagging poor tone, missed recognition, and relationship-building opportunities.
• Managers need broad context, but they also need depth on the work that matters most. Relying too heavily on AI summaries can be a trap. “I’ve found AI summaries to be useless in most contexts,”
• Instead of solely asking AI to condense information, use it to understand which work merits more careful review.
• “I have AI find me the five individual, specific nuggets that I know will be useful for me to go deep in.”
• Automate status updates to elevate 1:1s. Dr. Caribay Garcia Marquez, Principal Manager & Head of People Science R&D at Microsoft Viva, relies on AI-generated status updates so she can spend 1:1s on conceptual coaching
• Increase frequency to course-correct faster. “We can prototype things much faster,” says MSCI’s Faquiryan. “But that means people can waste real money on tokens or lose enormous amounts of time running far in the wrong direction.
• The most effective managers will be the ones who realize their job is no longer to police the work, but to clear the path.

16 more annotations...

• One aspect of today’s boom is already much larger than the TMT bubble ever was. In 2025, US businesses invested almost $1.5tn in IT equipment and software. At the peak of the TMT bubble, it was $466bn or $829bn when adjusted for inflation.
• Indeed, the US economy is growing solely because of the tech boom
• 93 per cent of US GDP growth was explained by tech investments. Even at the peak of the TMT bubble, it barely reached 60 per cent.
• preparing for blockbuster initial public offerings later this year to benefit from investor optimism about their growth.
• Meanwhile, the hyperscalers Microsoft, Alphabet, Amazon, Meta and Oracle plan to invest hundreds of billions in the next five years in data centres to provide the computing power to run these models.
• In these five years, capital investments are expected to rise by 20 per cent a year, a growth rate never seen before in this industry.
• Meanwhile, revenues are expected to grow 15 per cent annually.
• Yet, even under these extremely optimistic assumptions, I calculate the implied return on investment is highly negative for all of them except Amazon.
• the AI boom will become a story of one of the largest destructions of shareholder value in history. But there are two ways out of this corner.
• If the hyperscalers want to generate, say, a 10 per cent return on investment, they would have to find an additional $2tn to $5tn in revenue a year. A tall order for a group of companies that currently generates revenues of just $1.5tn per year
• The other option is that the planned investment in data centres, computer chips and other areas never materialises — maybe as equity investors turn more cautious on the sector or if debt funding for data centres becomes harder to get.
• The share prices of the largest companies on every continent from Nvidia to ASML, Samsung and TSMC are supported by these investment plans and the resulting demand.
• And remember that US GDP growth currently is driven exclusively by rising tech spending. If these start to drop, the US economy will enter recession very quickly
• A repeat of the tech crash in the early 2000s is a real risk with stock market drops of 50 per cent or more in the first year.
• The efforts by OpenAI, Anthropic and others to keep the hype going at least until after their respective IPOs are likely to support the boom for now.
• And eventually, reality will kick in. Probably not in 2026, but possibly in 2027 or 2028.
• the IPO of these AI companies is probably nothing more than a major transfer of investment risk from the current owners to retail investors, pension funds and others who are willing to buy the hype. 

15 more annotations...

• s I wrote then: tens of thousands of years of manufacturing knowledge disposed of in Evansville, replaced by workers with a few weeks of experience, while severance charges got an easier reception from short-term shareholders than a long-term investment ever would. Turnover costs chronically underestimated.
• That is some wacky accounting. Actually it's traditional accounting, and it makes perfect sense to a traditional accountant. That's the problem.
• Under both US GAAP and IFRS, people appear on the P&L as pure expense: salaries, benefits, training, all of it flows straight to the income statement.
• They appear nowhere on the balance sheet as an asset.
• When a company lays off people, there's no corresponding balance sheet entry. The income statement shows an improvement
• he asset — years of accumulated knowledge, problem-solving capability, customer relationships, institutional memory — simply disappears from the books without trace, because it was never there to begin with.
• Equipment depreciates. Human capital, in most cases, does the opposite.
• Wharton professors Peter Cappelli and Daniel Taylor have petitioned the SEC to require three modest disclosure changes within the existing GAAP framework
• require companies to identify what proportion of workforce costs represent investment in future growth
• treat workforce costs as a standalone line item rather than burying them in administrative expenses
• disaggregate labor costs in income statements so investors can see their contribution to major expense categories
• The SEC took a small step in 2020, amending Regulation S-K to require public companies to disclose "material" human capital information in their annual filings.
• The World Economic Forum and Willis Towers Watson have published voluntary frameworks for human capital accounting
• What if employees thought of their knowledge, creativity, and experience as their own asset, and companies entered into a long-term agreement for access to that asset
• Under that framing, a hire isn't a cost transaction; it's a lease.
• . Early termination of the lease — a layoff — would require recognizing a loss, not booking a gain.
• The favorable P&L impact of eliminating salary expense would be offset by the balance sheet cost of terminating the agreement.
• But as a mental model for what the current treatment obscures, it's clarifying.
• people actually owning their intellectual assets and seeking the highest-value deployment of them.
• Companies that treat people as interchangeable get what they're paying for.
• Companies that offshored manufacturing for labor cost savings are now discovering that the knowledge they discarded doesn't reconstitute when they come back.
• What that analysis still doesn't fully capture is the tribal knowledge — the undocumented expertise built over years of experience — that evaporated when the decision was made.
• Whirlpool's Bob saw it in 2006. The companies now struggling to reshore are learning it the hard way in 2025.
• So the question for leaders is whether they're willing to do the calculation in their heads that the ledger refuses to do for them.
• The balance sheet doesn't record what walks out the door. That doesn't mean it isn't real.

23 more annotations...

• Nobody is more exposed than the Japanese investment company SoftBank, which has poured an eye-watering $60 billion into OpenAI over the past few years.
• According to explosive new reporting by Bloomberg, even certain insiders at the company are rattled
• What’s clear from the reporting is that Altman has done what he does best: turned Son into a true believer in his vision
• But in recent years, the company is probably best known for Son’s dogged financial support of WeWork, the would-be coworking space startup with an Altman-like charismatic founder named Adam Neuman — and which imploded in spectacular fashion in 2019.
• is Altman a visionary ushering in a new world order, or is he a con man taking Son — and many other financial luminaries around the world — for a wild ride that’ll soon come crashing back to reality?
• . SoftBank has already sold top assets, including shares in fellow AI company Nvidia, to pay for its OpenAI commitment. And insiders are reportedly jittery about signs that OpenAI is losing ground,

4 more annotations...

• Al-Dahle notes that recent hiring of new graduates by major tech companies has declined by half since 2019
• uccessful reinforcement learning examples, like AlphaZero, thrive in stable environments with definitive rules and rewards. In contrast, knowledge work is characterized by continuously evolving rules subject to change, necessitating ongoing human guidance in AI evaluation.
• Many current AI systems are trained on the expertise of experienced workers. However, the automation of entry-level jobs, which traditionally cultivate such expertise, limits the next generation’s capacity for judgment.
• When the organizations no longer need such expertise for day-to-day operations, the incentive to pursue these careers diminishes, leading to a reduction in skilled professionals and an eventual decline in innovative capabilities.
• This creates a “hollowing out” effect, where superficial performance remains despite an underlying loss of human expertise capable of contextual validation and correction.
• urrent evaluation methods are primarily rubric-based, which capture only measurable criteria.
• Al-Dahle believes there is potential to close the evaluation gap through future advancements, but stresses that such solutions are currently unavailable.
• “The thing AI most needs from humans is the thing we’re least focused on preserving,” Al-Dahle stated, highlighting the risk of ignoring this critical aspect of AI development.

6 more annotations...