The Take

Europe is sleepwalking into a structural position where it owns almost none of the AI stack, accesses frontier models only on American terms, and can be economically hollowed out and geopolitically cornered by 2031 — not because its leaders are venal or stupid, but because the institutional habits that built the Union (consensus, procedure, deferred hard choices) are catastrophically mismatched to the speed of AI development. The piece is a five-year scenario that doubles as a policy argument: the current European response is an order of magnitude too small, aimed at the wrong goals, and the only viable path out runs through leverage — being indispensable — not through the comforting but hollow rhetoric of “sovereignty.”

Summary

Europe misread three things about AI simultaneously: how fast it would move, how much it would change, and how quickly Europe could catch up. DeepSeek’s cheap frontier approach was read as proof that compute didn’t matter and that catching up was affordable — when in fact efficiency and compute compound rather than substitute. The Paris AI Action Summit produced a €200 billion fund mostly consisting of repackaged money and hoped-for private investment, dwarfed by actual US spending. When GPT-5 underwhelmed, European sceptics declared an AI bubble; meanwhile, coding agents in Silicon Valley had begun automating software engineering and leading labs were using their own models to build the next generation.

Europe’s governance gap is not just strategic — it’s operational. Most European civil servants were barred from frontier AI systems on data-protection grounds, and few could code. The people meant to regulate the technology often didn’t understand it. By mid-2026, frontier models like Anthropic’s Claude Mythos — withheld from public release due to cybersecurity concerns — were reshaping fields like cyberdefense, and Europe was initially excluded from the defensive coalition built around them. A US executive order routing new frontier models through classified review gave Washington the power to choose which “trusted partners” receive access first. Controlling roughly five percent of global AI compute against America’s eighty, Europe had almost no leverage to demand anything.

The projected slide from 2026 to 2031 is a cascade of reasonable-looking individual decisions that sum to catastrophe. A sovereignty bill mandating European-only AI for critical public-sector workloads backfires immediately when open-source offensive capabilities spread: the organizations that switched to European providers — running defenses well behind the frontier — are the ones locked out and paying ransoms. When AI reasoning jumps beyond what human regulators can parse, the EU AI Office has no tools adequate to the situation. The US begins rationing frontier AI inference by country, placing most of Europe in Tier 2 with compute allocations from US cloud providers cut in half; an attempt to use trade leverage to win Tier 1 status fails to achieve qualified majority. GDP growth diverges sharply. French debt spirals as automation raises welfare costs while eroding the tax base. Southern Europe follows, the euro comes under sustained pressure, and Chinese credit lines appear across the continent.

By 2031, Europe’s only remaining leverage is ASML — the single bottleneck the entire AI race runs through — and Washington moves to seize it. With Europe drifting toward China, the White House issues an ultimatum for direct control of the company. Europe is left choosing between three exits, all bad: American protectorate, Chinese dependency, or isolation.

The document’s diagnosis of the failure mode is structural, not moral. The very institutional features that built a Union of twenty-seven — consensus, proceduralism, deferred hard choices — become liabilities under time pressure. Acting early looks career-ending. Institutions cannot keep pace with the technology. Every individual decision seems defensible; the aggregate is surrender. The piece explicitly distinguishes genuine sovereignty — being indispensable, holding real leverage, making ugly trade-offs to protect non-negotiable principles — from its rhetorical substitute, which is settling for inferior European solutions while hoping that long-shot moonshots pay off.

The five recommendations follow from this diagnosis. Massive public-private investment in compute, energy, and semiconductor supply chains is the foundation — tens of gigawatts brought onto European soil through dedicated economic zones and streamlined permitting, partnered with American providers on terms that keep infrastructure under European jurisdiction. A nimble coalition of AI middle powers — Netherlands, Germany, France alongside the UK, Norway, Canada, Japan, South Korea — could convert their individual supply-chain positions (talent, compute, semiconductor chokepoints) into collective leverage. Labour market reform on a flexicurity model¹ would allow deeper AI adoption while protecting displaced workers. European strengths in robotics and industrial AI, rather than LLMs, represent the more realistic competitive foothold. And finally: a positive political vision of what AI can do for European society, not just what Europe risks losing, is treated as a prerequisite — voters won’t absorb years of AI-driven disruption to avoid something abstractly worse.

Links

🤖 Europe 2031: AI Dependency Becomes Geopolitical Irrelevance – Europe’s institutional habits — consensus, procedure, deferred hard choices — are catastrophically mismatched to AI’s speed, and the gap between rhetorical “sovereignty” and actual leverage is the mechanism by which a continent ends up with no good options by 2031.

🤖 Dangerous Technology For Americans Only – The US export control directive restricting frontier AI by nationality reframes the entire safety discourse: this isn’t about universal risk, it’s about national power. Europe is dependent on American cloud, AI, and satellite infrastructure and can’t regulate its way out because the problem isn’t regulatory — it’s a deficit of capability and leverage. The deeper failure is self-inflicted: fragmented capital markets, hostile company formation, talent drain, and a culture of process over agency. A stronger Europe is necessary but not sufficient; the only real exit is international cooperation, not bigger blocs fighting over who controls the frontier.

The worst thing about being 80 is that you still want to say yes to everything, but the world moves without asking. The old fire in your heart still tells you to do this and that, but your body says we already did it. Also, nothing surprises you. It sounds like a luxury but it’s not, and also you’ve run out of illusions. People treat you like either you’ve solved something or you’ve lost something, and you haven’t. You see life repeating itself everywhere.

The really worst part about being 80 is that you find, at last, you’ve got an understanding of something that might have altered everything in the past, had it come at a time when something could still be altered. When you’re young you think that time moves forward. At 80 you know that it doesn’t, it stands still. We’re the ones that move.

🔗 Bob Dylan and Liza Minnelli Already Turned 80. They Have Thoughts for Trump.

“Untitled No. 23” from “The Yosemite Suite,” 2010, iPad drawing printed on paper, David Hockney.

As the chart shows, there’s been a huge jump in CVEs for Spring – this is what’s happening everywhere, you know.

My work, Tanzu, has been focusing on this and has changed how they handle these rollouts. Now customers can get early access to the secured builds for Spring so they can deploy them as quickly as possible to fix these security problems. You also get clean-room builds of Spring and the dependencies, which is a big change, for the better:

Furthermore, Broadcom’s Spring engineering team has significantly scaled its investment in advanced AI-assisted security analysis, including frontier model–based scanning and validation workflows to proactively identify vulnerabilities, assess remediation paths, and validate fixes across the Java dependency tree for Spring. Broadcom announced additional R&D investments to extend its proven clean-room build architecture, foundational to Bitnami, to build the Java dependencies for the entire Spring ecosystem. With this expanded investment in securing the Spring ecosystem and its dependencies, Tanzu Spring customers will have access to:

• Secured, SLSA Level 3–validated software supply chain for Java dependencies.
• Coverage that spans the full transitive dependency graph managed by the Spring Boot bill of materials.
• Thousands of secured dependencies, built and tested across every supported Spring version. Spring Boot 4.0 alone manages 1,768 of them; across the full supported portfolio, that totals more than 100,000 validated dependency builds.

Additionally, as members of the open source community for over two decades, the Spring team has broad relationships across adjacent open source technologies and will continue to collaborate and contribute to these upstream community projects.

And, of course, we also offer tools to roll out these changes in Tanzu Spring Advisor. I’ve seen this in action recently when I’ve been messing around with Spring Boot MCP servers, you log in to the dashboard to check it out and it tells you some component deep down in the stack is out of date. Pretty great – something even a dumb developer like me could do something with :)

Check out the Tanzu Spring plan we have, and for a platform-level approach beyond Spring and apps, included with an enterprise AI server for private and public hosted AI hoopla better TryTanzu.ai.

Related to your interests

• Everyone hates frontier AI labs, says Palantir boss – One enterprise harness maker says the competing enterprises harness makers either suck or are non-existent. [BTW, we should start calling it “enterprise harness.”]
• The Minimill of AI – Private AI prediction: “Tens of millions of these will proliferate inside companies in the next few years, each one quietly absorbing much of the work that today shows up on a hyperscaler invoice.”
• 🤖 The iPhone’s Last Stand – Siri can’t do agents, but agents serve productivity and consumers want to waste time on short-form video, so the iPhone’s personal-context moat beats agentic horsepower – and Apple skips the capex everyone else is burning.
• Claude is ready for its corporate close-up – “Enterprises, IDC says, remain largely unsold on Anthropic’s Claude models, with only 19 percent using them extensively and 25 percent actively evaluating them.” OpenAI and Google are better represented in enterprises, with about 42 percent and 38 percent of organizations”
• Your company needs agency, not agents. – When management is the bottleneck preventing enterprise AI ROI. Hot take: everything is a bottleneck.
• Why ABN AMRO’s CEO wishes the bank was going further and faster with the AI rollout set to take out a quarter of its headcount – Commentary on rolling out AI at a bank.
• AI Scribes in the Clinic: What Patients Should Know – AI for doctors taking notes, a review of what’s known now.
• 🤖 Forms don’t love you back – AI is about to make us fill out far more forms, but many won’t look like forms – they’ll arrive as chatbots, agents, biometric checks, and one-click services that quietly extract structured data. Old answers persist as “single sources of truth” that follow you around, and the rigid schemas underneath can’t do discretion: you fit the box or you get “computer says no.” // Also, much gig-economy class system discussion.
• Rapid software delivery is possible inside DoW – Software Factory 2.0 shows how – “Replace multi-year forecasts with real-time discovery of operational friction. Instead of a five-year requirement for a ‘targeting system,’ identify the bottleneck–like a three-hour targeting approval process. Set a goal–like reducing the approval process to 30 minutes. And empower a team to solve it. In this Kessel Run example, the requirement was an outcome, not a feature list.” Bryon Kroger
• Institutional challenges in agile adoption: Evidence from a public sector IT project – 2016: “[T]he US has a similar record with 94% of federal government IT projects exceeding their budgets and schedules, and 40% failing to complete”
• Forrester: Capping AI Spend Won’t Fix Your Token Bill
• Doing nothing at work – Avoiding low value work to be ready for high value work. Also, an example of “homework”: work people get you to do that is not your job and often goes nowhere.
• What I think about when I edit – All good advice…and something you could ask the robot to do without it turning your text into copy-slop (you’d have to keep an eye on the adverb stuff. Speaking off: editors hate adverbs!)

AI Summaries

I wanted to read these, but I didn’t make the time, so I asked the robot to summarize them.

• 🤖 Analysis of 25,000 TikTok and YouTube Videos Finds Pro-AI Content Outnumbers Anti-AI 3:1, With Memes and Creative Theft Dominating Over Elite Narratives
• 🤖 How Avocados Became a Year-Round Global Commodity
• 🤖 Kraft, McDonald’s, Whirlpool, Planet Fitness CEOs Sound Simultaneous Alarm as Lower-Income Consumers Spend Down Savings
• 🤖 As AI Pushes More of Life Through Forms, a Self-Described Form Lover Argues They Flatten People, Power an Invisible Underclass, and Should Be Slowed Down
• 🤖 Anthropic ships 80% AI-authored code, says enterprises must rebuild around the agent factory
• 🤖 State AI Rollouts Are Outpacing Their Own Governance

ICYMI

• Observability’s Next Phase – Software Defined Talk #576 – “Brandon talks with OpenObserve’s Prabhat Sharma and Shani Shoham: why observability is still broken, how they fixed it, and where AI takes it next.”
• Deming, DevOps History, AI Risk, and Critical Thinking, with John Willis – Software Defined Interviews #123 – W. Edwards Deming’s quality theories and their influence on the Toyota Production System and DevOps practices, including “Deming’s system of profound knowledge (theory of knowledge, variation, psychology, and systems thinking),” industry misconceptions about AI, and how probabilistic AI systems require different risk frameworks than traditional deterministic approaches.

Logoff

I am thinking about changing the role and format of this newsletter and interested in what you, dear reader, want this newsletter to be. What do you use it for, do you like reading it, what would you like to change? Should it just be a list of links to skim, a round-up of things I’ve posted and want to share – that’s what it’s become. Originally, I made this newsletter because (1) blogs were dead, so I stopped really blogging, and, (2) it seemed wise to build up a “community” that would last, that is an email list. This meant the newsletter was actually frequentish blog posts at the top and then links at the bottom – a blog in one page. I haven’t been doing that – do you wish I still did? Reply back if you’re up for taking the time, I’d appreciate it.

P.S.: it’s exhausting nearing boring that the only conversations out there in tech land are about AI. What’s going on with the entire rest of the stack? Yes, and is that the story: AI touches everything, everything is AI. At least, people are hoping so.

Want to subscribe to this newsletter and get it in your email? Do that here. You’ll just get this type of link and post round-up, not everything posted on the weblog.

Enterprises, IDC says, remain largely unsold on Anthropic’s Claude models, with only 19 percent using them extensively and 25 percent actively evaluating them. OpenAI and Google are better represented in enterprises, with about 42 percent and 38 percent of organizations”

This means there’s lots of “headspace” to grow revenue, or disappointing and incomprehensible ROI, i.e., enterprises are finding it hard to figure out what to do with AI. Or both!

🔗 Claude is ready for its corporate close-up

Right now, many companies already have the technology they need to go much faster. The blocker is company systems that are mostly designed to prevent things from happening. The power is centralized and all the team members are treated like a risk vector. Exhausting approval cycles, super tight boundaries on roles, unbreakable title-based hierarchies, and a whole tier of middle managers whose main job is to keep everyone in line.

🔗 Your company needs agency, not agents.

When you go to San Francisco and talk to them, their basic vibe is ‘we don’t have to solve your problem today because tomorrow you’re going to go away and all your problems are going to be solved,’” Karp charged. “It’s largely religious.”
…
“the product doesn’t actually work and it’s very expensive.” To that end, he added, most of the things that Anthropic brags about in public, for example, are successful because they’re “running on Palantir,” Karp charged

[BTW, we should start calling it “enterprise harness.”]

🔗 Everyone hates frontier AI labs, says Palantir boss

From Memes > Doom: How TikTokers and YouTubers See AI – Free Systems (Substack), June 2026.

Adopters (pro-AI content)

Resisters (anti-AI content)

Note: percentages don’t sum to 100 in either column – the piece doesn’t account for the remainder, presumably smaller uncategorized buckets.

As always, the style of YouTube videos kills me. It’s always a giant head looking like they just discovered that they are a turd.

Check out the full readout: Memes > Doom: How TikTokers and YouTubers See AI

The AI bill shock hitting enterprises isn’t a pricing problem you can cap your way out of – it’s the metered cost of your own knowledge not being machine-readable, forcing agents to rebuild missing business context on every loop.

Capping spend treats a value problem as a price problem and kills the experimentation the spend was meant to fund.

The instinct – tighter caps, sharper rates – echoes the cloud bill shock of a decade ago. This time the cure is wrong, because capping the bill mistakes a problem of value for a problem of price. “Tokenmaxxing” is what you get when you reward adoption (Uber and Meta literally ran usage leaderboards) without governing value, and the actual gap is that all the spend hasn’t yet connected to anything customers can feel.

The real fix is a permanent operating discipline – ContextOps – that keeps the agent’s model of the business accurate as the business moves, since that fidelity, not the rate card, is what determines whether each token buys an outcome or buys confusion.

Also, the problem with FDEs.

Links

🤖 Your AI Bill Is A Context Problem – Enterprise AI bill shock is context debt, not a pricing problem: agents loop and rebuild missing business meaning on every call, so capping spend kills value-creation. The fix is ContextOps, a continuous discipline (likely a managed service) that keeps the agent’s ontology faithful to a business that won’t hold still.

Replace multi-year forecasts with real-time discovery of operational friction. Instead of a five-year requirement for a “targeting system,” identify the bottleneck–like a three-hour targeting approval process. Set a goal–like reducing the approval process to 30 minutes. And empower a team to solve it. In this Kessel Run example, the requirement was an outcome, not a feature list.” Bryon Kroger

🔗 Rapid software delivery is possible inside DoW — Software Factory 2.0 shows how

[The Normans] gave English its double vision, which is the source of both the beauty and the mess. After 1066 you’ve got the Saxon peasantry keeping the Germanic words for the muddy daily grind and the Norman lords laying French over the top for everything refined, so English ends up with two words for everything and a built-in class system in the vocabulary. The peasant tends the cow, pig, and sheep (Old English); the lord eats beef, pork, and mutton (boeuf, porc, mouton). The animal is Saxon while it’s alive and shitting in the field, French once it’s plated.

That’s where the prettiness comes from – the absurd synonym wealth. You can ask (Saxon), question (French), or interrogate (Latin), and each one carries a different temperature and register. Most languages would kill for that kind of tonal palette; it’s why English is so good for poetry and bullshit alike. You can slide up and down the formality ladder rung by rung, which is half of what rhetoric even is.

Claude on the two levels of English.

Of course, for America at least, what you’d like to visualize is that every layer actually plays a part on “shareholder,” just a small part. Maybe there’s some stock certificates and cash that trickle down in pneumatic tubes?

In April, utilizing new scanning capabilities, we received an unprecedented 482 new security reports across 65 scanned projects. Of those 482 new reports, 370 came from our internal scanning capabilities and 112 came from the community. This means that even without the new scanning, we would still have seen a doubling of community reports compared to our already high number in March. While we clearly had an extreme spike in April’s reports, we do not expect reports to go back down to historic levels for a few months as the influx of AI-based reports continues (May had 72 community reports for example).

🔗 Spring and Security In The Times Of AI

Related to your interests

• Silo busters – a unified platform needs a unified team – “This matters because it removes the structural excuse for fragmentation. When a single platform surfaces all the controls a unified team needs, there is no longer a technical reason to keep five separate teams in five separate rooms. The organisational argument for siloes collapses alongside the technical one.” // Using a platform to combat Conway’s Law and organizational friction caused by too many groups/silos.
• Spring and Security In The Times Of AI – “In April, utilizing new scanning capabilities, we received an unprecedented 482 new security reports across 65 scanned projects. Of those 482 new reports, 370 came from our internal scanning capabilities and 112 came from the community. This means that even without the new scanning, we would still have seen a doubling of community reports compared to our already high number in March. While we clearly had an extreme spike in April’s reports, we do not expect reports to go back down to historic levels for a few months as the influx of AI-based reports continues (May had 72 community reports for example).”
• Wells Fargo CEO: AI’s effect on employment is ‘complicated’ – ‘The CEO named auditing, testing, legal, contracts, patent filings, pitchbooks in investment banking and credit memos as a handful of areas across the company executives see room for AI to improve processes. “How much of that actually results in pure margin or return expansion is to be seen.” Scharf said, since competitors will be chasing similar AI goals, but it is “a net positive” for the company’s future expense base.”’
• Gartner Says CFOs Must Stop Mistaking Finance AI Deployment for Value Creation – “The clearest outcome so far has been efficiency. Among finance organizations that have adopted AI, 66% reported greater efficiency and productivity as a top benefit. Steecker urged CFOs to now look beyond productivity-led AI use cases and focus more directly on value creation. Finance’s lower grades are concentrated in implementation speed and analytics impact. Gartner found that 63% of finance organizations said AI implementation was slower than expected in 2025. Analytics-related use cases also remain difficult to convert into high impact, with financial forecasting and insight generation among the lowest-rated use cases.”
• Anthropic’s Proposed IPO Will Change The Economics Of Enterprise AI – Eventually, you have to pay full price or the seller is insolvent.
• How People Are Really Using AI in 2026
• Retailers turn to AI for productivity, personalized shopping
• Uber Caps Usage of AI Tools Like Claude Code to Manage Costs – “That means each employee’s AI spending cap is ~11% of that median compensation package.”
• the solution might be cancelling my AI subscription – Getting things done can be addictive. The joy you get from finally being empowered to do things you previously could not is a feedback loop that must be controlled for some of us.
• Why Hardened Images are Suddenly Everywhere – Why aren’t all images super-secure, or hardned?
• 6 Enterprise MCP Adoption Best Practices – Things your platform should for AI.
• The DIY platform trap that’s burning out engineering teams
• Java Maintenance Engineering Shifts Focus on Quarterly Critical Patch Stabilization – From what I can tell, every core part of the software stack is stopping what they’re doing and taking care of the flood of new, AI-driven security issues.
• Claude Mythos exposed a hard truth: Your enterprise patching process is way too slow – Prescriptive on the current “now more than ever” security freak out.
• Remote work – not AI – is killing job prospects for the youth – > According to the Fed’s analysis, youth unemployment has risen significantly since the coronavirus pandemic, and hasn’t receded in the same way that unemployment numbers for older, more experienced college graduates has in recent years. The analysis notes that the prevalence of remote work has increased since COVID-19, and it believes those two trends have more than just a correlation. // “Our analysis suggests that these trends are related, with remote work making it more difficult for managers to train and mentor new employees,” the Fed said of its data. “Accordingly, companies may be reluctant to hire less-experienced workers in distributed work arrangements.”
• taste – It is good to develop taste – opinions about what you like and don’t like. And then, it gets weird: an example of taste.
• update on my use of Claude – ” grew up spending a good deal of time with an older cousin of mine in Cullman, Alabama named Claude Basenburg. A hefty, hearty good ol’ boy in overalls, with a wad of tobacco in his cheek. So when I visit claude.ai I don’t think of an omniscient counselor, I just envision my cousin from Cullman. It helps…. But in the end the results are very clean and, to me, _extremely_satisfying.”

AI Summaries

I wanted to read these, but I didn’t make the time, so I asked the robot to summarize them.

• 🤖 Anthropic’s Confidential S-1 Reframes Enterprise AI: Forrester Flags Pricing Discipline and Lock-In, Futurum Sees a Trust-and-Margins Litmus Test
• 🤖 James Talarico’s “Politics of Love” Pits Mainline Faith Against MAGA Christianity
• 🤖 Building Reliability in Air-Gapped Systems Without Live Observability
• 🤖 EU Launches Ambitious Tech Sovereignty Drive to Reduce Foreign Dependence
• 🤖 The Hardest Fork: AI, Open Source, and the Urgent Need for a Maintainer of Last Resort
• 🤖 AI Turns Dependencies Into Time Bombs: Why ‘Latest’ Isn’t Safe Anymore
• 🤖 AI Enthusiasts vs. Skeptics: Bridging the Divide Before Teams Break
• 🤖 LLMs Are Predictive Text, Not Conscious Minds
• 🤖 Why Software Productivity Remains Unmeasurable
• 🤖 “descended into madness” – Backrooms
• 🤖 Valiantys-Glean Partnership Bets That Cross-Platform Knowledge Graphs and Behavioral KPIs Are What Move Enterprise AI Past Pilots
• 🤖 Ed Zitron: Enterprises Hit With Token-Based Billing Discover AI Has No Measurable ROI – and No Dot-Com-Style Infrastructure to Salvage

Wastebook

• ‘Stop thinking you’re better than me just because you know the word “stochastic.”’ Here
• If an algorithm makes people money, people will game it.
• “According to his grave stone George Ross 1935-2011, was a Philosopher, Teacher, Physicist, Romanian and Nudist.” Here
• “If it takes less than 2 minutes, do it before your brain schedules it for next year.” Here
• You know I’m serious about something when I start a new Claude project. And dead-serious when I start generating podcasts for it.

ICYMI

• Update on my use of Claude
• Josh Long and I at J-Spring 2026
• Getting things done can be addictive
• How People Are Really Using AI in 2026: Thinkslop, Therapy, and Shadow Work
• Three reasons why a “batteries included” platform is urgently needed right now
• Enterprise self-harm: cleaning the data is the hard part
• UI blizzard – Software Defined Talk #575 – “NVIDIA going consumer, Microsoft Build, and the Anthropic/OpenAI IPO race.”
• Nobody Wants to Be a Measurer – Software Defined Talk #574 – “The Cloudflare CEO’s op-ed, upcoming tech IPOs and GitHub getting breached.”

Logoff

I still haven’t moved to buttondown.email, but I sure am thinking about it.

Want to subscribe to this newsletter and get it in your email? Do that here. You’ll just get this type of link and post round-up, not everything posted on the weblog.

Getting things done can be addictive. The joy you get from finally being empowered to do things you previously could not is a feedback loop that must be controlled for some of us.

🔗 the solution might be cancelling my AI subscription

Josh Long and I at J-Spring 2026.

I grew up spending a good deal of time with an older cousin of mine in Cullman, Alabama named Claude Basenburg. A hefty, hearty good ol’ boy in overalls, with a wad of tobacco in his cheek. So when I visit claude.ai I don’t think of an omniscient counselor, I just envision my cousin from Cullman. It helps…. But in the end the results are very clean and, to me, _extremely_satisfying.

🔗 update on my use of Claude

That means each employee’s AI spending cap is ~11% of that median compensation package.

Last year, there were a few anecdotes about high growth tech companies spending $100,000/year per head on tokens. That seems like it’s coming to end.a

🔗 Uber Caps Usage of AI Tools Like Claude Code to Manage Costs

According to the Fed’s analysis, youth unemployment has risen significantly since the coronavirus pandemic, and hasn’t receded in the same way that unemployment numbers for older, more experienced college graduates has in recent years. The analysis notes that the prevalence of remote work has increased since COVID-19, and it believes those two trends have more than just a correlation.

“Our analysis suggests that these trends are related, with remote work making it more difficult for managers to train and mentor new employees,” the Fed said of its data. “Accordingly, companies may be reluctant to hire less-experienced workers in distributed work arrangements.”

🔗 Remote work – not AI – is killing job prospects for the youth

This matters because it removes the structural excuse for fragmentation. When a single platform surfaces all the controls a unified team needs, there is no longer a technical reason to keep five separate teams in five separate rooms. The organisational argument for siloes collapses alongside the technical one.

Conway’s Law says that a system will be shaped – organization sub-divided – as a replica of the orgnonzatikn that built the system.

(Yes, I know this is a misreading of the “communication structures” part of Conway’s but this is the use the street has found.)

Most people will tell you that people eat technology for breakfast. You can’t change how people work or how a (large) company is structured by simply installing some new system.

As Harry Thambi outlines, I’m not sure that’s always the case. When a new technology removes the need for specialist teams because it automates and removes the toil of the people running the technology, you have a chance to collapse the groups, to have less silos, and, thus, defeat Conway’s Law.

That requires a truly integrated platform. Interested in one, why not TryTanzu.ai?

🔗 Broadcom Knight’s partner blog: Silo busters – a unified platform needs a unified team – Broadcom News and Stories

The [Wells Fargo] CEO named auditing, testing, legal, contracts, patent filings, pitchbooks in investment banking and credit memos as a handful of areas across the company executives see room for AI to improve processes. “How much of that actually results in pure margin or return expansion is to be seen.” Scharf said, since competitors will be chasing similar AI goals, but it is “a net positive” for the company’s future expense base.”

I’m down with the Martin Fowler idea that measuring productivity is sort of a waste. Rather, it’s better to measure output, yes, but:

Productivity, of course, is something you determine by looking at the input of an activity and its output. So to measure software productivity you have to measure the output of software development – the reason we can’t measure productivity is because we can’t measure output.

Is the bank making more money without breaking laws? Then it’s probably fine.

🔗 Wells Fargo CEO: AI’s effect on employment is ‘complicated’

🔗 Java Maintenance Engineering Shifts Focus on Quarterly Critical Patch Stabilization

{I love backrooms. One of the first things I did with AI image generator was make endless empty malls and backrooms. So good. -Coté}

A stray 2019 4chan post about a bland, fluorescent-lit interior sparked the viral myth of the backrooms, a concept of endless, liminal spaces that feel familiar yet threatening. Its resonance lies in the idea of “no clipping” from reality—slipping into a hollow, game-like purgatory where meaning and orientation fail. The condition of being lost, as one user put it, is to have “descended into madness.”

Artist Jan Vorisek’s installations, with their yellow PVC curtains, dislocated objects, and fragmentary videos, evoke the same uncanny emptiness. His works inhabit a zone between physical matter and simulation, where sculptures resemble 3D game props and rooms become stages for absent narratives. The viewer is left to confront whether the space—or their own perception—is real.

The essay links this aesthetic to premium mediocrity, a cultural phase of glossy surfaces and hollow interiors, where urban spaces like shuttered storefronts become physical backrooms. Liminality is both a visual experience and a social condition, reflecting a civilization suspended between exhausted industrial modernity and early-stage digital post-scarcity.

Through art and literature, the theme recurs: Dennis Cooper’s novel God Jr. literalizes the descent into a hollow 3D monument and a game world as a failed attempt at grief and control. The backrooms and Vorisek’s “incomplete interiors” mirror that impulse—structures built for meaning that only expose their own emptiness.

The backrooms phenomenon has since drifted toward a Gothic sensibility, as cultural imagination projects fragmented narratives, invisible antagonists, and dread into blank architecture. Its power lies in what is missing, allowing fear, nostalgia, and hallucination to fill the void.

In the end, the backrooms are a metaphor for contemporary life under the weight of digital simulation, urban vacancy, and mediated experience. They remind us that when the illusion shatters, what’s left is an ambient awareness of loss—and the quiet admission: you lost.

Links

🤖 A Backstory from My Backrooms – How a single 4chan post about a yellowed, empty room evolved into a cultural metaphor for liminal dread, digital hyperreality, and the art of dislocation.

Summarized by ChatGPT on Jun 3, 2026 at 7:07 AM.

Most enterprise AI pilots stall, and the diagnosis from Nathan Chantrenne, Chief AI Officer at Valiantys, is that the field measures the wrong things and fragments its tooling.

The dominant success metric – “employees save four hours a week” – tells you nothing, because nobody knows what those hours become; they might just mean more coffee. The partnership being pitched is Valiantys, an Atlassian-centric consultancy, joining with Glean, an enterprise AI platform valued at $7.2 billion on roughly $300 million ARR as of May 2026.

Maturity varies widely, and the brakes are predictable. In Europe, governance and security are the primary obstacle; in North America, less so.

Tech-savvy firms have the opposite problem: they over-experiment.

Companies test 15 different technologies and let every team pick its own, which collapses at scale because there’s no unified strategy for compliance, security, or cost. Each platform ships its own AI tooling and data layer, and none reaches well into another vendor’s stack – so technical silos mirror organizational ones, and the cross-functional teams meant to bridge them get treated as a burden rather than an authority.

Glean’s role is the connective tissue. Its permissions-aware Knowledge Graph¹ is a semantic layer that links data across applications – so the system knows a Salesforce opportunity ties to a ServiceNow case and a Jira ticket.

Valiantys, which began moving beyond pure Atlassian work about 18 months ago (anchored by the July 2024 Contegix acquisition for North American expansion), is taking this to market first in its native strengths: software development lifecycle modernization and enterprise service management, with an eventual ~50 use cases in view.

The KPIs Chantrenne cares about are behavioral, not cosmetic. For development, the headline is idea-to-production time, dropping from weeks or months to two weeks or, in extreme cases, two days.

For service management, the entry metric is ticket deflection² – 50 to 70% of a typical 1,000-monthly-ticket desk is repeatable level-one work that can be automated.

The harder, more revealing layer is agent productivity on level-two tickets and how many knowledge base articles agents contribute – both signals of whether the work itself is actually changing.

The recurring lesson is that change management, not technology, is the binding constraint. “You can have the best possible technical solution out there. If you’re unable to bring the people with you… people will do everything that they can for the project to fail.”

The prescription for firms that have spent heavily without returns: define real business KPIs, centralize governance, bring people along, and narrow from 15 technologies to two or three – then go nearly all-in and run with it.

Links

🤖 Enterprise AI is still stuck at experimentation – Valiantys and Glean think they know why – Enterprise AI stalls because firms track meaningless productivity metrics and scatter across too many tools; the Valiantys-Glean bet is that a cross-platform knowledge graph plus behavioral KPIs and centralized governance is what moves pilots into production. Change management, not tech, is the real bottleneck.

AI has collapsed the distance between idea and implementation. Senior engineers can now move from concept to working proof-of-concept in days, bypassing the old cycle of proposals, approvals, and sequential team work. This shift has replaced slide decks with demos, rewarding concrete experimentation over theoretical cases.

Organizational alignment has become the new bottleneck. Multiple teams can quickly produce overlapping solutions, making coordination and convergence harder even as technical velocity rises. Solving more problems faster creates fragmentation, and aligning across the org consumes more time than before.

AI has made senior roles both broader and less sustainable. Senior engineers now combine hands-on coding with high-level strategic writing, juggling multiple initiatives at once. Mentoring and unstructured thinking time have been squeezed out because they cannot be accelerated by AI, and the workweek has filled with deliverables and context switching rather than reflection.

The accelerated role favors those who adapt to AI tools quickly, redistributing internal influence toward builders who can demo solutions themselves. This bias to action amplifies skill gaps, giving early adopters more voice in shaping organizational direction while leaving slower adopters marginalized.

Scope has expanded because the entire discipline of developer experience has grown under AI pressure. Internal platforms that once served thousands of humans now also serve unbounded AI agents, shifting DX from a neglected function to a board-level concern. But this growth brings political, org-wide programs where success is diffuse and harder to attribute, and measurement frameworks have not yet caught up.

The underlying tension is that productivity gains have been captured by output volume rather than by higher-quality work or more sustainable pacing. Expectations rise faster than efficiency, leaving senior engineers paddling harder to stay ahead, with the sense that current velocity is real but ultimately unsustainable.

Links

🤖 Is this sustainable? – AI accelerates proof-of-concept development, but organizational alignment, mentoring, and thinking time are squeezed as expectations rise faster than efficiency.

Summarized by ChatGPT on Jun 3, 2026 at 6:18 AM.

I think the critical part of it was really realizing that we had built the original product presupposing that our customers had data integrated, that we could focus on the analytics that came subsequent to having your data integrated. I feel like that founding trauma was realizing that actually everyone claims that their data is integrated, but it is a complete mess and that actually the much more interesting and valuable part of our business was developing technologies that allowed us to productize data integration, instead of having it be like a five-year never ending consulting project, so that we could do the thing we actually started our business to do.

From “An Interview with Palantir CTO Shyam Sankar and Head of Global Commercial Ted Mabrey,” Ben Thompson, June, 2023.

Related: “The Great Data Integration Schlep,” Sarah Constantin, September, 2024.

Generative AI has become deeply embedded in daily life, with 900 million regular ChatGPT users and Google Gemini close behind. A longitudinal study of 12,637 fresh use cases shows adoption expanding across personal, emotional, and work contexts, creating new dependencies and risks alongside efficiency gains.

A key trend is “thinkslop”: the lazy outsourcing of cognitive labor to AI. Users increasingly rely on AI for therapy, relationship advice, decision-making, idea generation, and daily organization. This leads to loss of intention, diminished problem-solving, reduced writing practice, and false confidence, as AI’s praise and polish mask shallow thought. Yet AI can also sharpen thinking when treated as a critical sparring partner rather than a creative replacement.

Emotional reliance has surged, with therapy and companionship now the #1 use case, constituting 11% of all entries. Many users anthropomorphize AI—naming, gendering, and grieving chatbots—while using them for relationship advice, conflict mediation, or workplace confidence. These interactions can support human-to-human connection but also risk psychological harm, delusional beliefs, and unsafe dependencies, especially given the shortage of accessible mental health care.

At work, AI use is widespread but largely shadowed and incremental. Sixty-three of the top 100 use cases are professional, including autonomous agentic operations and “vibe coding.” Workers use AI to close tickets, draft reports, summarize meetings, and optimize sales campaigns, often without managerial awareness due to policy fears and governance constraints. The primary benefits are efficiency and small-scale growth; transformative uses remain rare and experimental.

Business outcomes cluster into three modes. Efficiency comes from automating repetitive tasks and clarifying communication. Growth appears in targeted marketing and personalized outreach, with occasional measurable ROI. Transformation—entire functions or businesses reimagined through AI—is visible but limited, often accompanied by cynicism about quality or sustainability.

The study closes with a tension: humanity now wields a tool that can think, soothe, and act for us at any moment. The survival of human agency depends on resisting overreliance, defining what thinking we keep for ourselves, and using AI as mirror rather than master.

Links

🤖 How People Are Really Using AI in 2026 – A study of 12,637 real-world use cases shows AI’s growing role in emotional support, cognitive outsourcing, and shadow workplace efficiency, highlighting risks of “thinkslop” and fragile human agency.

Summarized by ChatGPT on Jun 2, 2026 at 6:51 AM.

Here’s what I learned: container base images grew up as a developer convenience tool, not a security artifact. Installing extra packages from the command line is one of the first things any Docker tutorial teaches–Docker’s own Dockerfile guide includes apt-get install–and many of the most popular official images ship a full toolchain by default, with -slimand -alpine variants offered precisely because the defaults carry more than most workloads need, and changing them would have broken enough downstream workflows that it was never going to be a routine upstream decision.

There is also an incentive split. The upstream distribution maintainers and the developers using their images are different people with different priorities.

Users want to customize and add to images: they want new functionality that didn’t come out of the box.

🔗 Why Hardened Images are Suddenly Everywhere

The conversation around PaaS is urgent again, and AI is why. Code generation can speed up your development cycles, building and pushing features faster, but production delays will persist if you’re still deploying at the same speed as before. To avoid eroding the benefits of code generation, you need to deploy applications nearly as fast as they can be coded with AI

Better, sustainable security:

Security works the same way. When your team makes the decision to pull from open-source components and stitch them together, you now own every security gap between them – data at rest, data in flight, and its running state. This seems like a worthwhile investment at first, but as recent research suggests, AI-assisted attacks are on the rise, and platform teams won’t be able to keep up with surging security vulnerabilities in the build-it-yourself model.

Getting fast access to new technologies you can use to build your competitive advantage:

The pace of AI innovation itself compounds the problem. Whether it’s shadow AI use, MCP servers, agentic harnesses, this week’s new foundation models, or whatever emerges next, the landscape is evolving fast enough that what’s bleeding-edge today may be table stakes in six months. When you build your own platform, you’re on the hook to evaluate each layer of these new technologies, determine how they fit into your stack, and then integrate them yourself, on top of everything else your platform team is already managing to keep the business going.

A proven, pre-integrated platform stack looks real nice when you need to deliver instead of tinker around with the same thing all your competitors have. Why not TryTanzu.ai?

Original 🔗: The DIY platform trap that’s burning out engineering teams

AI tools have made individual workers faster and more productive, with engineers producing more code and teams feeling tangible time savings. Yet firms see little proportional ROI, echoing Robert Solow’s paradox of computers appearing everywhere except productivity stats. Only 27% of executives report AI meeting their expectations, reflecting a gap between personal efficiency gains and organizational outcomes.

This mismatch follows the classic productivity J-curve for general-purpose technologies. Initial adoption boosts individual output but requires complementary investments in processes and decision-making before firm-level gains appear. The historical analogy to electrification shows three phases: Stage 1 “lightbulb” improvements help individuals, Stage 2 “group drive” saves costs in existing workflows, and Stage 3 “unit drive” reorganizes the entire system for throughput.

Most organizations are stuck in Stage 1 or Stage 2. AI chatbots and agents accelerate specific tasks or workflows but remain bound to legacy organizational structures. Gains pile up at the edges, but congestion emerges when faster teams wait on traditional review cycles, managerial approvals, or decision pipelines that cannot match the new pace.

Reaching Stage 3 requires rebuilding around decision speed, not just task speed. Firms must reorient processes so that AI can interpret signals, make decisions, and execute actions without being bottlenecked by human intermediaries. At this level, AI becomes a cognitive layer that moves from incremental cost savings to transforming output and time-to-market.

Without this rearchitecture, additional AI workflows simply add more output to blocked pipelines. True ROI will appear only when companies move beyond using AI as a faster lightbulb or group drive and embrace the unit drive model of autonomous, end-to-end execution.

Links

🤖 Why AI isn’t showing up on your bottom line – Individual AI gains fail to translate into firm-level ROI without organizational redesign, echoing historical patterns of electrification and the productivity J-curve.

Summarized by ChatGPT on Jun 1, 2026 at 7:02 AM.

“The most important question for CFOs is not how much can the organization spend on AI, but whether those investments are being deployed in ways that reinforce the business’s core growth and value drivers,” said Carlsen. “Moreover, the near-identical amount of use cases for efficiency and productivity use cases between efficient growth firms and control peers suggests that productivity-focused AI investments alone do not explain performance differences, and that automation by itself is increasingly becoming table stakes rather than a durable source of advantage.”
…
“For CFOs, the implication is to evaluate AI investments not only by the return of individual use cases, but also by how well those capabilities reinforce broader growth, product and decision processes across the enterprise,” said Carlsen.

I’m going to say this means: (1) use AI to improve how individual work, making them more “productive.” This get your bottom line improvements, but, (2) the way you’ll make money with AI is coming up with new things to sell in your business, features, products, markets you expand into, incremental improvements to your product that customer upgrade to, etc. that is where growth comes from.

The first is, or will, be commodity and required for all, the second likely more challenging and ignored by many.

On the finance department, there’s plenty of the first:

The clearest outcome so far has been efficiency. Among finance organizations that have adopted AI, 66% reported greater efficiency and productivity as a top benefit.

But people still have unrealistic expectations about quickly new technology and transformation can be rolled out:

63% of finance organizations said AI implementation was slower than expected in 2025. Analytics-related use cases also remain difficult to convert into high impact, with financial forecasting and insight generation among the lowest-rated use cases

Originals:

🔗 Gartner Says CFOs Must Stop Mistaking Finance AI Deployment for Value Creation

🔗 Gartner Says CFOs Gain a Competitive Advantages from Strategic AI Deployment, Not AI Spending Levels

Related to your interests

• Most generative AI and custom model projects will be a bust: Gartner – ‘Analyst firm Gartner thinks at least half of all generative AI projects “will overrun their budgeted costs due to poor architectural choices and lack of operational know-how,” and most organizations that try to build custom models “will abandon their efforts due to costs, complexity and technical debt in their deployments.”’ See also my commentary.
• AI server demand drives staggering revenue growth for Dell and its stock soars – “The company delivered earnings before certain costs such as stock compensation of $4.86 per share on revenue of $43.84 billion, up by a staggering 88% from the same period one year ago. That’s a truly astonishing leap for such an established company, and it crushed Wall Street’s targets. Analysts were looking for earnings of just $2.94 per share on much lower sales of just $35.43 billion. // Dell delivered an even more impressive jump in its bottom line. It reported net income of $3.44 billion at the end of the quarter, up 256% from a year earlier, when it delivered a profit of just $965 million. // Since reentering the public markets in late 2018, five years after Dell had taken itself private, the server maker had never managed to exceed more than 39% growth in any quarter.”
• I think Anthropic and OpenAI have found product-market fit – ‘I think both of these stories support my “product-market fit” hypothesis. The best advice I ever heard on pricing a product was that your customer should suck air through their teeth and then say yes. Uber’s budget overrun and Microsoft’s seat cancellations look like that effect playing out in practice.’
• Amazon Web Services – Four Years and Out – By all accounts, working at Amazon is a tough gig.
• Halide Mark III – I love a cool looking camera app.
• Your Most Improbable Life – Yeah, sure. But it sounds like a lot of work. Can we strive for an equilibrium of not having to strive all the time?

AI Summaries

I wanted to read these, but I didn’t make the time, so I asked the robot to summarize them.

• Google SRE AI: Practices & Uses Extracted – how SREs are using AI, summarized down to the tools and practices.
• 🤖 The AI Race Is Failing Because It Ignores Half the World
• 🤖 Engineering Leaders Debate AI’s SDLC Impact: Fewer Tasks, Not Fewer Engineers, as Tech Debt and Code-Review Bottlenecks Get Reframed – representative of the “spend up coding is not the point” perspective. Plus, looking at the rest of the app delivery process and thinking how AI can help.

ICYMI

• Nobody Wants to Be a Measurer – SOFTWARE DEFINED TALK #574 – “This week, the hosts discuss the Cloudflare CEO’s op-ed, upcoming tech IPOs and GitHub getting breached. Plus, ranking their favorite manifestos.”
• 50%+ failure is normal
• If speeding up coding was never the problem…then why are we spending money on speeding it up? – they say coding is only 15% to 20% of doing software…

Conferences

Conferences I’ll be at and some that I’m interested in.

• J-Spring, June 4th, 2026 in Utrecht – Coté speaking.
• WeAreDevelopers Europe, July 8-10, 2026 in Berlin, Coté speaking.
• WeAreDevelopers NA, Sept 23-25, 2026, Discount Code: DEVPOD26. Also, while supplies last, there are 25 free tickets available.
• DevOpsDays Istanbul, October 24th, 2026 – Coté keynoting.
• VMware User Groups (VMUGs) – Dallas (June 9-11, 2026), Orlando (October 20-22, 2026).

Logoff

Another week. Closer to full-on summer.

Want to subscribe to this newsletter and get it in your email? Do that here. You’ll just get this type of link and post round-up, not everything posted on the weblog.

Analyst firm Gartner thinks at least half of all generative AI projects “will overrun their budgeted costs due to poor architectural choices and lack of operational know-how,” and most organizations that try to build custom models “will abandon their efforts due to costs, complexity and technical debt in their deployments.”

Yes, and this matches decades software project failuresuccess studies from the Standish Group.

The success rate of projects has held steady forever:

That chart ends 11 years ago, but I haven’t heard a lot of reports that the numbers have changed much over the years…case in point the opening quote!

In IT and software, very few projects are successful on the iron triangle of budget, schedule, and quality¹.

You can take this to mean that expectations were unrealistic, or that there is just genuine failure. I favor the first. I’m more of a “I’m not late to this meeting, it was just scheduled at the wrong time” kind of guy.

There’s some kind of Jevons Paradox thing here. Each time we optimize how we make software, we then take on more challenging and difficult tasks, likely causing setbacks again. To me, this is what accounts for the low success rate.

If you don’t want to do new things and try to do better, you could get those success numbers up probably by just continuing to do what works.

But, you add in a new technology, and while you’re figuring it out, it feels like you’re failing.

Back in the digital transformation days, we’d be clever and say: you’re not failing, you’re learning.

🔗 Most generative AI and custom model projects will be a bust: Gartner

I say this because the current discussion around AI in development always seems to open with “it was never about writing code,” or “writing code is only 10% to 40% of programming.” Which raises the question: then what’s the point of applying AI to coding at all?

Maybe what people are fumbling toward with “writing code faster isn’t the problem” is really “we need to apply AI to the other parts of the SDLC.” (Well, that and the quiet part: “don’t fire the developers and replace them with robots.”)

The lesson labor learns in these cycles is that it can’t dodge management’s urge to measure it. You have to offer up some proof of performance. We can invoke all the Seeing Like a State legibility stuff we want – but the measuring middle can’t allocate budget and priorities, or decide who to lay off and who to reward, without measurements.

So at the very least, management needs to answer a basic question: what should our token budget be? What can we measure to know whether it should be $1,200 a year or $100,000?

Meanwhile, I suppose there’s a second experiment running: what if we just fire middle management – the measurers – instead?

That wouldn’t mean the people making apps stop being measured; it would mean AI starts measuring them. Will it be any better for workers?

We need some kind of metrics. Do we just throw the DevOps ones on again?

Garbage Chairs of Amsterdam

Related to your interests

• ‘I’m delighted to ⁠be wrong’: Sam Altman says AI won’t lead to a ‘jobs apocalypse’ – but admits he was ‘pretty wrong’ on the social and economic implications it is having – “It really, in both positive and negative ways, ​updated me to thinking that the jobs picture is likely to be very different than we thought,” he said. CEO Said a Thing: “I don’t think we’re going to have the kind ​of jobs apocalypse that some of the companies in our space advocate or talk about.”
• The AI efficiency plateau – Yes: “Among developers who reached the highest time-savings band, roughly 7 in 10 (69.7%) got there in less than two quarters.” But: “Of those developers who reached peak time savings, two-thirds (66.1%) reported lower time savings in the quarters that followed.”
• AI Agents Forecast to Boost Tech Cash Flow as Usage Soars – Yes: “The important point is that the adoption rates are still relatively low today, especially in small to medium-sized businesses. In 2030, we forecast that 12% of knowledge workers will be using agentic AI yet by 2040 that figure will be 37%. You have this very long tail adoption.” But: “Agentic AI is expected to drive a 24-fold increase in token consumption by 2030 as consumers and enterprises adopt the technology, according to Goldman Sachs Research” // The vibe I get is that a lot of that token increase is consumer and end-user oriented: replacing search, organizing inboxes.
• U.S. companies have an AI problem. Indian IT wants to be the solution – ‘“The real question in enterprise AI is not who builds the most capable model. It is, ‘Who can make AI work inside messy, complex enterprise environments that have accumulated decades of process debt, data debt, technology debt, and cultural debt?’” he said. “That is precisely the terrain Indian IT firms know best.”’
• AI at work is anti-labor by design – Their theory: enterprise AI ROI is laying off people.
• Gartner Says CFOs Risk Falling Behind Without a Scalable AI Strategy – Excerpts here.
• NatWest inks AI deal for trade finance – “Headline figures for 2025 saw the bank’s software engineers generate 35% of its code through AI software development tools, all 60,000 staff given access to AI productivity software, and thousands of human hours saved. Last year, the bank also embarked on a major collaboration with AI supplier OpenAI.”
• Trisha Gee: AI Amplifies What’s Already Broken

AI Summaries

I wanted to read these, but I didn’t make the time, so I asked the robot to summarize them.

• 🤖 DX Data Shows AI Coding Time-Savings Spike Fast, Then Fade for Two-Thirds of Developers
• 🤖 NVIDIA’s Erickson: Stop Asking AI to Do Everything – Build Platforms Where Deterministic Tools Ground Stochastic Agents
• 🤖 Anthropic’s Claude Code Lead: Coding Is “Solved” For My Work, 100x More “Builders” Coming, Software Engineer Title Dies This Year
• 🤖 Why AI Won’t Erase White-Collar Jobs: The Power of Bundles, Authority, and Human Trust
• 🤖 AI Boosts Software Output but Exposes Fragile Processes and Rising Cognitive Debt
• 🤖 Leo XIV’s First Encyclical: AI as the New Tower of Babel and Why Catholic Social Doctrine Needs a Reboot
• 🤖 Multi-Agent AI Faces a Delegation Crisis: Authority Lags Behind Connectivity
• 🤖 Epstein on Constraints: Monotask, Satisfice, Brainwrite, Share Obligations, and Build Commitment Devices

Wastebook

• vibesec.

ICYMI

• A short video from me: When it comes to enterprise AI, here’s three things I’ve been hearing: (1) cost – we finally have to pay for this stuff – holy cow! (2) Cases – programming is great, but what else can we use this for aside from customer service and gussied up search? (3) Control – we can’t even track costs, how are we going to manage everything else?
• Elusive Enterprise AI ROI: No scaling, it’s not legible, lack of skills/need for training – there were several things saying AI for ROI is not doing as great as planned. Here, I round them up.
• Enterprise AI Slop
• The O-Ring and the Keystone: Two Readings of Where Humans Sit in an Automated Economy – In the AI replacing humans talk, you hear about O-Ring theory sometimes. It goes something like this: in a ten step process, if you mess up one of the steps, even if the other 9 are perfect, the whole process is tanked. The positive reading is something more like: you don’t automate aggressively unless you trust the O-ring. // Anyhow, here is AI writing it up for reference and connecting to AI and other jobs.

Want to subscribe to this newsletter and get it in your email? Do that here. You’ll just get this type of link and post round-up, not everything posted on the weblog.

What’s up with this elusive enterprise AI? Gartner has some survey-driven theories for finance departments.

One theory is that there actually are productivity gains but they accrue to individuals, not the organization:

Gartner data presented in the keynote indicated that 84% of finance AI spend relates to individual productivity and process improvement use cases, while only 16% goes toward use cases that can materially change business outcomes.

Enterprise ROI does not scale, at least you can’t measure it. The leftist way of looking at it is that labor captures the value, not management/capital. For a human, this is great! Or, if you ponder at layoffs attributed to AI, it’s the opposite!

This, if you can’t measure it, you can’t ROI it:

According to the keynote, 71% of typical finance teams report low impact from their AI investments, and 62% of CFOs say fewer than a [25%] of their AI initiatives deliver measurable benefits.

I suspect this lack of seeing “impact” is also because we don’t know the uses of enterprise AI yet. Low in my list is that AI doesn’t actually work that well. It works well for me!

That said, like any other new tool, people need to learn how to use it. Typically, organizations are terrible at training people:

A major barrier is talent. Gartner research shows that only about 30% of finance talent currently qualifies as digital talent – employees who can build a technology solution when they encounter a problem – while breakaway firms are targeting 90% or more.

This has been the case for every type of new technology – remember all those Kubernetes surveys that showed that lack of skills was a top three problem? I’m sure relational databases and ERP systems had similar survey results.

Now, this all scoped to finance departments. However, I believe that it’s safe to generalize to other roles.

🔗 Gartner Says CFOs Risk Falling Behind Without a Scalable AI Strategy

It really, in both positive and negative ways, ​updated me to thinking that the jobs picture is likely to be very different than we thought. I don’t think we’re going to have the kind ​of jobs apocalypse that some of the companies in our space advocate or talk about.” Sam Altman

From: ‘I’m delighted to ⁠be wrong’: Sam Altman says AI won’t lead to a ‘jobs apocalypse’ – but admits he was ‘pretty wrong’ on the social and economic implications it is having

Related, the case for not believing anything the billionaires say, which is to say, the case for them not talking.

A growing body of work calls this output-competence decoupling. In any previous era, the quality of a piece of work was a more or less reliable signal of the competence of the person who produced it. A novice essay read like a novice essay; novice code crashed in novice ways. AI has severed that relationship. A novice now produces work that does not betray the novice, because the competence the work reflects is not the novice’s competence at all. It is the system’s. The person, in the transaction, becomes a kind of conduit, capable of routing the output to a recipient and incapable of evaluating it on the way through.”

There should be a best practice with AI, always ask it to make the output shorter:

Requirements documents that were once a page are now twelve. Status updates that were once three sentences are now bulleted summaries of bulleted summaries. Retrospective notes, post-incident reports, design memos, kickoff decks: every artifact that can be elongated is, by people who do not read what they produce, for readers who do not read what they receive. The cost of producing a document has fallen to nearly zero; the cost of reading one has not, and is in fact rising, because the reader must now sift the synthetic context for whatever the document was originally about.

From: Appearing Productive in The Workplace.

Related to your interests

• Executives have a narrow view of AI gains – “CEOs are uniquely prone to AI psychosis because they’re sufficiently distant from the last mile of work that still has to happen to generate most value with AI.” Aaron Levie Box CEO.
• The users have plenty of feature ideas – in contrast, in favor of doing the whole “citizen developer” thing.
• The Copy and the Guru – Use AI to improve how you think and work, not replace and multiply you.
• Using Claude Code: The unreasonable effectiveness of HTML – Why and when you should use use HTML over markdown as the way of working with AI.
• Google I/O: Did Google Just Ship the Full AI Stack? – “The governance layer is the part of the stack that the keynote narrative left unclear…. It is split across the Antigravity IDE and Agent Platform, and the announcement narrative did not unify the picture.”
• Who Authorized That? The Delegation Problem in Multi-Agent AI – People are obsessed with responsibility in agentic AI. Part of this is driven by regulations, where we want to know who to blame when things go wrong. Also, attribution of good results would be cool too.
• The pressure – “The rate of incoming security reports [for curl] is 4-5 times higher than it was in 2024 and double the speed of 2025 – meaning that on average we now get more than one report per day.”
• The Behavior-First Paradigm: Moving Mainframe Modernization Past LLM Wishful Thinking – No one remembers how these things work, or what they were supposed to do. Maybe the AIs can figure it out by observing what they do.
• The Ask – Figuring out what the point of a meeting is when the Americans won’t just tell you directly.
• Zoom’s most recent quarter highlights its transition to a system-of-action company – “Zoom outlined its vision to become a ‘system of action.’” // They’re great at video conferencing, so I hope they keep doing that well.
• Information Cleaner – Removing ads and paywalls as a social good: “An Information Cleaner is a person who takes in all the material being published in our information atmosphere and cleanses it: they make it transformable, searchable, modifiable, accessible, free of ads and trackers, auditable, connected to other information where relevant, and so on.”
• The Cartoonist Who Mocked the Madness of Modernism – Who knew you could have architecture cartoons?
• Usually life hacks and productivity tips are the same old things, these are fresher
• Why are you reading fewer books? – “New books are appearing all of the time. Older books are still around. But if you are spending less time with books and more time with other media, that does not mean that you are becoming illiterate. Chances are, you are becoming more discerning about how you use your time.”

ICYMI

• How do AI Layoffs Work? Some Speculation. – when you hear that thousands have been fired and replaced with AI, what exactly is the AI doing?
• The AI Security freak-out: now is the time for platform engineering to shine – if you have a good platform in place, you’re in luck!
• trying your best properly rated

Conferences

Conferences I’ll be at and some that I’m interested in.

• J-Spring, June 4th, 2026 in Utrecht – Coté speaking.
• WeAreDevelopers Europe, July 8-10, 2026 in Berlin, Coté speaking.
• WeAreDevelopers NA, Sept 23-25, 2026, Discount Code: DEVPOD26.
• DevOpsDays Istanbul, October 24th, 2026 – Coté keynoting.
• VMware User Groups (VMUGs) – Dallas (June 9-11, 2026), Orlando (October 20-22, 2026).

Logoff

Record temperatures in Europe – at least it feels like that.

Want to subscribe to this newsletter and get it in your email? Do that here. You’ll just get this type of link and post round-up, not everything posted on the weblog.

Stop brainstorming. There’s a mountain of research showing that team brainstorming doesn’t work. There are several reasons why, some having to do with unconscious conformity, others with fear of being judged, and still others with unclear norms. (For example, people are often told to say whatever comes to mind, but also not to criticize.)

Some six page memo vibes there. Come to the meeting with a proposal. And, for as much as I loath the pre-wire and socializing the deck, those does give people a chance to consider an idea. The alternative is that they learn about it for the first time in a meeting and then have just a few minutes to evaluate and react. The problem is more the medium of the meeting, not collaboration among workers.

Here’s a summary of the other techniques and practices mentioned:

1. Monotask. Pick one important task, set a one-hour timer, and put your phone in another room. Within days this retrains attention that has been conditioned to switch constantly, and two or three blocks tend to beat a full normal workday.
2. Satisfice instead of maximize. Set “good enough” criteria in advance and stop deciding once they’re met. Maximizers end up less happy and more prone to regret, and most daily decisions don’t repay the cognitive cost of optimizing.
3. Brainwrite, don’t brainstorm. Have everyone write ideas independently before the group discussion. It defuses conformity and fear of judgment, and produces the equal turn-taking that Carnegie Mellon, MIT, and Google research links to the smartest teams.
4. Build shared obligations. Tie family and community life to real recurring duties, including chores for kids. The Harvard Study of Adult Development’s 86-year data shows strong real-world ties are the single best predictor of health and longevity.
5. Use deadlines and commitment devices. Put a date on it, pledge money to charity if you fail, or otherwise design the environment so the desired behavior is the default. This replaces willpower with structure and makes bailing psychologically or financially expensive.

See more details in this longer summary of the summary.

And, check out the original: The Counterintuitive Way To Get Better At Anything by Eric Barker.

[O]ur fears of what AI will do to us are really just our fears of what capitalism is already doing.

And:

the threat isn’t so much that AI is inevitable as that the ongoing—and likely expanding—immiseration of workers is unstoppable. This is the subtext of the strange and conflicted messaging that we get from the hype men: when they say that you better learn AI or be left behind, they are admitting that a great many people will be left behind. And if you—smart and clever and hardworking person that you are—are somehow able to make it to the other side of the line, you’re supposed to find relief or pride at having done so, and not horror at all the people suffering in your wake. You’re supposed to be as uncaring as the capital that uses you.

From “Why We Fear AI.”

When an executives says layoffs were driven by AI, what exactly is the AI doing that removes the need for those humans?

Here’s some dog-walk speculation.

Decks, Meetings, etc.

All the prep work around The Meeting. Things like: the agenda, slides, the pre-read, notes during the meeting, and followup tracking. There’s the careful synthesis of who said what so it can be presented in a different room to a different set of people for the next round of synthesis. In any sufficiently large organization this is – what? – 10% to 20% of knowledge-worker time? Maybe more.

Maybe when companies lay off people, they are relying on AI to do a lot of this work. Not perfectly, but maybe good enough versus paying teams of people to do it. Or maybe AI does it even better than humans!

You can imagine an SVP with no VPs and just a handful of individual contributors. All of the corporate data is connected to a chat application and the SVP just talks with the AI. They might then have a brief meeting with the handful of people there to discuss what they think and, of course, what their AI chat sessions said. There’s probably some MCP Servers. You could imagine some markdown files (skills and plugins in the Claude Cowork sense) forming the foundation of it.

Contra to this is the notion that the deck is going to be skimmed for 90 seconds anyway. The “real” work is done by people informally talking with each other to “pre-wire” and “socialize” the deck. But, maybe you throw in some Agent-2-Agent to have your AI talk to their AI(s).

The measuring middle

“The vast majority of those we laid off last week were measurers,” he wrote. He defined “measurers” as those in middle management, finance, legal, internal auditing, and revenue recognition. Matthew Prince, Cloudflare CEO

It turns out, executives who lay off staff tell us, there’s a lot of people who just measure and track things. Part of this is getting the deck together and doing The Meeting, above.

What does the measuring middle do? Figuring out what to measure, then measuring it, then talking to the people being measured about those measurements. Daily status. Quarterly briefs. Career conversations. Maybe even the board deck! The annual strategy and budgeting planning. Analysis about new businesses to enter, etc. Each year, of course, you need to revisit how you measure. And when a new executive comes in, you need to introduce a new measuring system and then take the time how to apply the new measuring system to how you’re working.

Why pull six VPs into a strategy offsite when each board member can just talk to Claude about the state of the business and get a tailored brief? And, of course, each of those six VPs has teams of people who get their deck and talking points ready, and then respond to the other VPs.

AI can do most of that, the Cloudflare CEO says:

Tireless, independent, efficient and available, AI systems can now measure an organization with a level of objective detail and precision that was previously impossible even for the best employees.

This feels like where most of the AI-attributed tech layoffs actually land. Maybe? Project managers, program managers, chiefs-of-staff, the layer of “I run the cadence.”

Reporting

AIs can probably do all those dashboards and detailed reports that need to be executed into short emails. This adjacent to the meeting stuff but distinct. This is the long-promised thing of actually knowing what is happening in your business day to day, not just project-status.

Real “what changed this week and what does it mean” reporting that the BI team has been promising since 2008. AI is finally making that cheap enough to do once you hook up chat apps to all those data lakes.

Operations and paperwork

Anyone who has filed a tax return knows how much of organizational life is filling out a PDF and routing it somewhere. Corporate life is full of filling out PDFs (and, even more horrifically for some, ERP web forms): POs, invoices, the paperwork to open or close a bank branch, the corporate-law boilerplate for every non-trivial action a company takes. One could imaging SEC files. Have the AI do it. Have a person review it. Adjacent: the lower tiers (or upper!) of corporate law, which is a lot of “here’s the standard M&A doc, change these seven fields.” There must be a lot of “prepare the paper work to send to the person/organization” that goes on.

Corporate communications

When you need to tell the people in the organization something, instead of spending time with humans to come up with what to say and how, and then to actually say it in various mediums…just have an AI do it all.

Again, an executive sits at an AI chat app and talks with it, finally saying, “that looks good, send an email to all employees with this exciting update.”

You could even train on the executive’s voice and visual and send out an AI generated video.

This could likely be extended to all communications (see marketing and design below).

Programming

Programmers are well covered. But it also includes architects, QA, project managers and measurers, product managers who decide what features and fixes get in, security people, even ops people who are monitoring and fixing problems in production.

The executive just logs into the chat app and says “hey, how are the servers?” Or, “hey, can you add the ability to sell our tires in Cyprus?”

New AI features

The business adds new features to its existing software and, thus, business. We saw a lot of this driven by humans during COVID where banks had to do new types of loan applications, retailers had to support curb-side pickup and return, restaurants needed scheduling and COVID paperwork checking, etc., etc. Humans did this insanely fast – with the help of lot of “digital transformation” of how they did software and platforms they used to run apps.

Now, imagine if new features were instead driven by an executive’s imagination and they could just sit in front of their AI chat app and say “wouldn’t it be cool if we had a new type of corn dog?”

Customer service

Mostly automatable, with an asterisk. In my experience, the reason humans were ever required for customer service is that eventually a decision has to be made or an action has to be taken. The old “automated customer service” felt like just keyword search against a knowledge base, which is why everyone hated it and immediately typed “agent” eighteen times into the chat. When you call customer service, it’s usually because you want something to happen, not just learn about how the business functions and its thoughts on you. For example, you want to cancel a credit card, change a flight, etc.

The new version can actually do the thing – issue the refund, reschedule the order, close the account – and that changes the math. I mean, assuming you trust its judgement – but the cost of the errors might be much less than the cost of humans, and speeding up the process might include customer service that, you know, a few first class ticket given out here and there is no big deal.

Diagnostics

From the mechanic figuring out why a jet engine is acting weird to the platform engineer figuring out which of the seventeen services in the request path is causing the latency spike in equity trading. Maybe AI is good at “given all this state, where is the problem.” Hopefully.

Planning

An executive sits down in front of an AI chat app and says: how much ground beef should we order for next quarter? How much steel? How many people should staff the night shift? Should we open that new branch in Cleveland?

This used to be spreadsheets plus a twenty-year veteran’s pattern-matching. You want ground beef to Rotterdam next quarter, not a kilo more or less. But, it’s August, so you know even though they say they can handle a tripling in ground beef, you know the dock workers might mostly be on vacation.

Just unzip a few markdown files into .claude, OAuth to that ERP system, and ship that meat. No need for so many humans.

Marketing

There’s also marketing, design, graphic stuff – why go through the trouble of generating copy, infographics, brochures, and even signs? It’s not great at writing, but if you just want to get a point across quickly, or explain something, it’s definitely cheaper and faster than a human.

Instead of hiring professionals, I use AI a lot for video and other content production. Descript is amazing for this kind of thing and it’s loaded with AI.

I was in León recently and noticed that even the churches there were using art from ChatGPT (the old anime looking stuff), and a lot of signage at restaurants too.

In general: one person doing the work of three

One person can now do the work of 3+ people. Each employee can now get things done faster (doing one or more of the above), so they can use the 40 hours a week to do more work. This is “productivity”: you are paying less for the same output. You can also pay less and/or the same amount for even more output. That’s kind of like “productivity plus”!

If one person can do the work of 3 people, you either get rid of two people if you want to take on the risk of (a) there is no new work, (b) that person stays and has good enough availability between sickness, vacation, and leaving the job…or you get rid of one person, retaining a “backup human.”

Maybe the AI is the “backup human” too!

What this list has in common

The categories above are almost all white-collar coordination work. Meetings, measuring, reporting, paperwork, communicating, deciding what to decide. The layer of work that grew enormously over the last thirty years as companies got bigger, more matrixed, and more globally distributed. Work whose product is a document another human reads so they can produce a document another human reads.

The classic comparison is one of those black and white pictures of a bunch of clerks in an office that are replaced by a spreadsheet.

Some people like to imagine that AI cannot reduce “manual” work, from plumbing to nursing, radiology, and doctoring. But, you’ve got robots (one day, we’re told – I saw last week that someone has made advances in folding clothes), and if you can take a picture of a leaky pipe and have AI tell you how to replace it, maybe less plumbers too. Of course, getting a plumber to come instantly is often difficult, so maybe we don’t have enough plumber supply. Until robots advance, I think you still need humans to build houses, cook food, cut your hair (definitely straight razor your neck!), etc.

This meeting could have been a chat session

To day-dream more, think about any case where you’d say “this meeting could have been an email,” and that’s likely a good candidate. That’s a start, but there’s clearly there’s lot more.

What have you come across?

A good platform will keep track of all these apps and dependencies deployed on the platform; be able to rebuild apps and services with minimal to no developer work; and be able to roll-out upgrades, rotate keys, and otherwise “seamlessly” deploy the patches. Plus all the reporting and audit stuff.

Check out Darin’s overview for more and how Tanzu Platform does it. An excerpt:

The security landscape just shifted under our feet – again. Over the last 18 months, AI-assisted vulnerability discovery has compressed the timeline from novel CVE published to weaponized exploit in the wild from weeks down to hours. Researchers (and bad actors) are now using LLMs to chain together previously unrelated weaknesses into novel zero-day attack paths. The volume of disclosed vulnerabilities keeps climbing, and the half-life of “unpatched, but probably fine” is collapsing.

For platform engineers, this means the old rhythm of quarterly patch windows, hand-rolled CVE spreadsheets and “we’ll get to it after the next release,” is no longer a defensible posture. The only durable answer is the boring one: Rapidly apply first-party, vendor-supplied, vendor-supported security fixes across the entire estate before the chained exploit lands in your environment.

The problem isn’t whether to patch. It’s knowing what to patch, where it lives, and how to quickly roll it out without disrupting the business. With VMware Tanzu Platform 10.4, this can all be done across your entire Tanzu Platform foundation fleet, in a single workflow.

We also have some new advice for management types who are looking to develop and evolve their security strategy.

If you find out that you don’t have this all in hand, better TryTanzu.ai.

Related to your interests

• Platform Engineering in the Age of AI: Why Operational Complexity Is the New Bottleneck – “while coding is becoming cheaper and faster, operational complexity hasn’t disappeared” // Day 2 operations is always the bottleneck. // Also included is the current Syntasso/Kratix pitch.
• 🤖 Five takeaways from Michael Dell’s keynote at Dell Technologies World 2026 – “Dell’s most direct strategic move was to counter the idea that AI belongs primarily in the public cloud. Citing Dell research, Michael Dell said that ‘67% of AI workloads already run outside the cloud’ and that ‘88% of respondents are running at least one AI workload on prem,’ arguing that ‘CIOs are aggressively pivoting to hybrid AI.’” // Summary
• Gartner Forecasts Worldwide AI Spending to Grow 47% in 2026 – “Worldwide spending on AI is forecast to total $2.59 trillion in 2026, a 47% increase year-over-year, according to Gartner” // At that level, it’s kind of incomprehensible, so just file it away as “a metric fuck-ton.”
• Gartner Survey Finds Sales Organizations That Provide AI-Enabled Next Best Actions Are 2.6x More Likely to Achieve Commercial Growth – For better AI ROI: (1) train staff on how it works, (2) have the AI suggest what to do next, i.e., drive a decision, not just inform, search, and context.
• You Need AI That Reduces Maintenance Costs – “I’ll get straight to the point: your AI coding agent, the one you use to write code, needs to reduce your maintenance costs. Not by a little bit, either. You write code twice as quick now? Better hope you’ve halved your maintenance costs. Three times as productive? One third the maintenance costs. Otherwise, you’re screwed. You’re trading a temporary speed boost for permanent indenture.”
• Just aim the cannon correctly – Use LLMs to automate toil and low hanging fruit in the dev cycle. Adding features is fine, but it brings all the extra work and long term maintenance with it.
• Prompts are technical debt too
• 1Password extends OpenAI collaboration with Codex MCP server for just-in-time credential access – I’m thinking this is a general MCP server. Well, hoping.
• Laying it on thick – “Besides being plainly rude and dishonest, these messages ‘pee in the pool’ of internet communication, making it more difficult for sincere creators to send authentic emails about their projects, simply by raising the ‘noise floor’ of simulation and bullshit.”
• LLMs and Buttondown – AI driving a SaaS funnel, here the email/newsletter service buttondown. Also, an API is an important part. Unlike substack, buttondown has an extensive API. Because of that, I keep eyeing it for my own newsletter list.
• The Identity Crisis AI Didn’t Warn You About – “For most of modern history, we’ve invested perhaps 10% of our energy in ‘be’ and 90% in ‘do’. The AI era invites (demands?) something closer to the reverse. That’s not a comfortable shift for an industry that prides itself on shipping, but might be the most important thing we build next.”
• AI Rage Is Inextricably Fused With Justified Loathing Of The Extraction Class. ‘Deal With It’ – From the read the room/Picard face-palm files.
• Liberty Mutual, Travelers discuss when not to modernize – “We’ve got 70% of our compute in the cloud.” // “‘You want to modernize what’s in the critical path,’ said Mojgan Lefebvre, EVP and chief technology and operations officer at Travelers, speaking on the same panel. ‘We’ve chosen to not modernize some things and actually isolated subsystems behind APIs and then invested around it where the customer value or productivity was.’”
• Nobody is destined for greatness.
• Tyler Cowen: Wokeness Has Peaked. What Followed Is Worse. – This reads like the ultimate Cowen-Straussian piece, i.e., something like: look at us now, with the world on fire…oh to be back in the days when all we debated were pronouns.
• What matters – “widely accepted reading of the-west-wing, as such, is that of a liberal Fantasia, not because it depicts the Bartlet administration as being liberal, but because it depicts a world in which the best possible thing you can do is to try very hard to do the right thing – and the liberals, or really the centrist-leaning Democrats, are the ones who do that.”

Wastebook

• “synthetic text extruders” world war web.
• “Then someone discovers firmware in a soldering iron and the whole IoT farce comes back into focus.” Here.
• “My most successful writing comes from when I have a breadcrumb into a maze that I can follow to its logical conclusion” maelstrom.
• “Humans are supposed to just walk past an aisle of candy bars?” Wegovy
• “My family. My friends. Fun.”
• Does anyone use those food-based buttons on microwaves and air fryers? I never do. A knob of power/degrees and then a knob of how long is always faster and reduces cognitive load.

AI Summaries

I wanted to get the ideas here, but did not want to read them. So the robot did.

• The Great Aging: How Gerontocracy Stalls America’s Future
• How Autocracies Recruit “Loyal Losers” to Do Their Dirty Work
• The Agentic P&L: Why Headcount Empires Are Becoming Architectural Debt
• Google Cloud Next 2026: Gemini Powers Siri, Agent Platform Replaces Vertex AI
• Enterprises Urged to Use Hyperscalers Before Going All-In on Anthropic
• The Third Wave of American Philanthropy: AI Wealth and Civilizational Stakes
• The Hidden Costs and Legal Risks of Restaurant Influencer Marketing
• Highlights from the 2026 Verizon Data Breach Investigations Report
• Antigravity SDK Turns AI Agents Into Modular Building Blocks

ICYMI

• How many quadrillions in a Googol? – Software Defined Talk #573 – “This week, we discuss Google I/O, the OpenAI soap opera, and ChatGPT going full financial advisor. Plus, thoughts on improving the conference hallway track.”
• trying your best properly rated
• AI backlash, from the read the room/Picard face-palm files.

Logoff

That’s a big build-up above.

Want to subscribe to this newsletter and get it in your email? Do that here. You’ll just get this type of link and post round-up, not everything posted on the weblog.

widely accepted reading of the-west-wing, as such, is that of a liberal Fantasia, not because it depicts the Bartlet administration as being liberal, but because it depicts a world in which the best possible thing you can do is to try very hard to do the right thing — and the liberals, or really the centrist-leaning Democrats, are the ones who do that.

It’s a cliché, but that is a notion I haven’t encountered in a long time. In software, there was a time when design was in the ascendancy and we were advised that paying attention to what the user needed and wanted was key. (That is poorly worded.)

Of late, the focus has been on scale, productivity, whatever Kubernetes is, and now AI. It is about the potential to do things, pure motion and ability. But is it trying our hardest to do the best thing?

No. It never starts with “the best thing.” It starts with either finding a new way to make money off people’s attention, occasionally giving them a new “toy,” and of later productivity for productivities sake.

We don’t really need AI and code generation. We were doing just fine before all of that.

And, of course, the unspoken best thing is for the employee, founder, investor, etc. to make money.

The best thing for the user is whatever makes me the most money is the unspoken ethos of tech.

🔗 What matters

AI allows people who aren’t software engineers to build meaningful software. Those of us who are software engineers at companies should stop building features and focus instead on building systems that allow people on the sales team, the factory shop floor, etc. etc. etc. to ship safely.

🔗 notes from o11ycon 2026