Trail running is the most selfish thing I do. Even if I don’t get around to doing it as often as I wish. It is also, somehow, the most communal.

You go out alone into something hard, for reasons nobody else can fully access, and then the moment you finish the whole thing belongs to everyone who has ever done it. That contradiction is the engine of the sport, and it is also the reason the sport is so hard to write about, sell, or scale without losing the plot. The five articles below circle it from different directions: culture and class, one runner’s absurd origin story, an elite training block, and the open question of who gets to tell these stories at all.

Enjoy these Five Things Running!

Subscribe now

The Trail was Always the Teacher

Trail running is simultaneously the most selfish and most selfless pursuit I know. You go alone into something hard. Profoundly alone, at 3 a.m., in the cold, with a headlamp and a list of things your body would prefer you not do. That’s selfish in the most elemental sense — a private reckoning. But here is what the elders taught me: the moment you finish — or don’t — that story belongs to the community. It becomes part of the shared library. The veteran at the aid station who looks you in the eye and says I’ve been here. I know what this is — that is culture operating exactly as designed.

This is why no FKT, no podium finish, no race series acquisition can touch the center of this thing. The center is a conversation between a person and a trail and, eventually, between that person and everyone else who has had the same conversation. That exchange is ancient. It is not going anywhere.

On point.

(…continue reading.)

C is for Community. C is for Culture.

There were repeated comparisons to skateboarding, and I reject them wholeheartedly based on the simple fact that skateboarders are cool. Skating is an anti-establishment subculture. Skaters challenge societal norms simply by using the built environment (parking lots, stairs, railings) for a secondary, often-destructive recreational activity.

Trail running on the other hand is just running through public lands with thousands of dollars of equipment, occasionally adding another hefty bill (entry, flights, hotel, crew) to run a race. The time and money required dictates that trail runners come from a different socio-economic class than those who take a board to a parking lot to grind curbs.

I am glad we have people like who continue to point out what we need to improve to make our sport more inclusive.

(…continue reading.)

I Ran A Marathon. Here’s Why I’m Dumb Enough To Do It Again.

It was Mick Jagger who got me into running. This sounds comical, not least because of all the hobbies one might plausibly inherit from the Rolling Stones frontman — from the music to his signature struts, to cocaine addiction — the one I picked up was much worse: long-distance running. But there he was, in an interview, explaining how he manages to stay in such preposterous shape well into his 70s while continuing to prance across stadium stages like a possessed aerobics instructor. His secret, apparently, is running.

What a great way to start an article about running a marathon. It got me hooked right away.

(…continue reading.)

How Vincent Bouillard Trained for Western States

As the coach of such an incredible athlete, my main job is just not to screw him up. Like many of us, Vincent enjoys smashing himself in training. Part of the reason he wanted to work together in the first place was to make sure he was doing that productively so that he could bounce back and do it again a few days later. The other part was that in addition to working part time on the innovation team at Hoka, he was going to be a dad soon and wanted to make the most of his training time so that he could still be a good employee, partner, and parent. Our relationship is very much a collaboration. I wrote out the training two weeks at a time, and his input and feedback were essential to making it work.

I honestly don’t know how people do this: running, job, family - I don’t have the discipline for that, let alone the urge to have the discipline for that. It’s impressive.

(…continue reading.)

The Return of Trail Running Journalism?

In all seriousness, the more interesting question sits above all of that: in a landscape where trail news is no longer chased by journalists but fed straight to hungry Instagram aggregation accounts by PR teams, and where the sharpest independent analysis lives on Substack but is narrow (one athlete’s training log, one gear obsessive’s spreadsheet, more opinions on the story of the sport as a whole than are necessary) what’s the actual lane TrailRunner is claiming? Its own framing pushes toward the terrain the aggregators and Substackers already occupy: personality, lifestyle, vibe, speed - but with more resourcing and bigger pockets.

Methinks that the trailrunning space is still too small for more independent and objective journalism that moves beyond tales by runners and sponsored shoe reviews.

(…continue reading.)

If you missed last week’s edition, you can read it here:

Now, go running!

— Nico

🏃🏻‍♂️

Dang, it’s cold. The temperatures dropped from 39° Celsius to 19 or so and the rain doesn’t make it any better. Why can’t they just set the thermostat at 28°C and leave it there until October?

The USA is currently elebrating the 250th anniversary by largely ignoring the festivities Donald Trump and his croonies planned. This is well deserved and continues to give me hope that the MAGA idiocy will come to an end soon. Now would be a good time to figure out who the leader of the Democrats could be who unites and heals the people.

Speaking of well-deserved: the German national team had to pack their bags after just four games at the World Cup. For Germany, this is a major crisis and of course 84 million head coaches could have done a better job than Julian Nagelsmann, who spectaculary failed and has had to step down. He won’t be missed as he so overpromised and so underdelivered. Everybody is still baffled how a team packed with so many star players who get traded for absurd amounts of money could so underperform. This is not the summer fairy tale we hoped for.

Speaking of hope: last evening I had hoped to get this newsletter done before my best friend came over and I assumed he’d be late, but he was way too early, so I had to stop writing this newsletter and had to resume it in the morning. “Where’s Five Things?” was the first message I got from my wife this morning, which means that I have at least one reader, but I assume she just reads it as she fears I might write something snarky about her, which I’d never do, of course.

Anyhow, here’s Five Things, a little later than usual, but very much worth reading!

Subscribe now

But first, coffee: The drink that energized the American Revolution

The Boston Tea Party protest was targeted at the British government’s passing of the Tea Act in 1773, which granted the East India Company a monopoly over tea sales in the colonies. While the British had removed some unpopular taxes in the preceding years, they left tea taxes in place. Colonial merchants were especially upset that the act allowed the East India Company to undercut their tea business.

To build solidarity for their cause of sovereignty, some patriots called on colonialists to swear off tea in favor of coffee. It’s why many histories point to the Boston Tea Party as a turning point when Americans switched from mostly drinking tea to mostly coffee. The anti-tea sentiment was immortalized in a founding father’s now-famous letter.

I studied American history at university and while I am sitting here, sipping my coffee, I am trying to recall if coffee was ever mentioned in any of the books I read or lectures I attended. Fascinating. Also, bonus link: The Coffee Belt Runs Through Hamburg’s Roasters - Hamburg is still the coffee place to be…

(…continue reading.)

Germany were once World Cup kings. Now they’re simply not that good

When Germany beat Argentina in 2014 to claim their fourth World Cup title in 17 attempts, it capped a four-tournament run since 2002 that had seen them finish second, third, third and then first. It also came five days after they beat the hosts Brazil in arguably the World Cup’s most shocking final score: 7-1.

At that point, only the most shameless Brazilian would have argued with the idea that Germany were the World Cup’s top nation. OK, Brazil had one more star on their jersey than Germany but Germany had also been runner-up four times and had won four third-fourth play-off matches.

Add all that up and Germany have finished on FIFA’s podium 12 times — three more than any other nation. So, their highs were higher than almost every other nation’s and their lows were higher than everyone’s.

There are two truths we have to face: the rest of the world has gotten so much better at football and that Germany is no longer a “Turniermannschaft” - a tournament team, that excels under pressure and succeeds due to a rare combination of luck and talent. I guess now we just have to win the Euro 2028!

(…continue reading.)

On the origin of continents

In other ways, plate tectonics is a very unusual paradigm. It was discovered twice: first intentionally by geologists in Europe and its colonial sphere, and second, accidentally, by North American geophysicists. The two paths offer opposite lessons on the secret of scientific discovery. There is no lone genius who can claim the majority of the credit. The theory went from fringe to total acceptance over the course of ten years, with nary a funeral in sight. And it all came together much later than one would think: around the time of the first moon landing in 1969. This is the story of how two separate groups of scientists near-independently discovered plate tectonics with wildly different approaches, and what this can teach us about paradigm shifts.

Do you know how sometimes there are familiar names and you just don’t know (or bother to look up) who that person was and what he did? Of course I knew the Alfred Wegener Institute, but had no idea who Alfred Wegener was. Until recently, when I read the wonderful book Traversal by Maria Popova. Fascinating stuff.

(…continue reading.)

We Need A Way To Prove Personhood Online

Media coverage of the rise of political bots in the mid 2010s alerted the public to the downsides of automatons among us — they could be deployed en masse to harass people out of conversations, to manufacture consensus and to kick off fake “trends,” making the public believe that more people cared about a thing than actually did. They were interlopers in a human environment, but they tended to be primitive, easily detectable and largely manageable.

That is no longer true. Generative AI has transformed bots from mediocre copypasta accounts into chatty, Large Language Model (LLM)-powered avatars with visuals tailored to appeal to their targets’ aesthetic preferences. And now, we have agents like Rathbun: software that doesn’t just generate content but takes autonomous action — writing code, posting, negotiating, booking and sometimes … attacking.

This is a really interesting debate. I am writing this newsletter all by myself and I also read all the articles I write about all by myself. But how can you know for sure? And while humans have ID cards, I sure want people to be able to stay anonymous when they are online, just as we can stay anonymous when we are offline. It’s a topic that is so hard to wrap your head around, but we need to find solutions that are not total surveillance. At least that’s what my agent suggests…

(…continue reading.)

Are Crows Really Our Friends?

As we’ve moved to urban areas, crows followed, and like coyotes, they’ve found that city dwellers are less likely to trap or shoot them. Ever savvy, crows have learned new habits, such as nesting on utility poles, and as with many types of urban wildlife, they seem to be growing less wary around us. Such changes may be cultural, with birds passing down knowledge over generations. Behavioral shifts could also become baked into their DNA if birds with an inborn interest in, say, eating fries or nesting on a high-rise are more likely to survive to reproduce.

Over millennia of following us into new habitats, crows also learned that we are a fickle species. Humans provide food, but sometimes we harass or kill them. Our settlements can be alternately deadly or sheltering. To “avoid danger and take advantage of riches,” Marzluff says, “they can’t not be interested in people.” So they’ve evolved to watch closely what humans do. And their keen attention may be key to a friendship that’s singular.

We have crows in the tree in front of our house and my wife has turned into a bird-watcher. So basically I picked this article so she could tell me “see, I told you so!” all Sunday long, which is something that she normally would never do.

(…continue reading.)

That’s it. Have a great Sunday! If you missed last Sunday’s edition of Five Things, have a look here:

Subscribe now

— Nico

The internet did not build itself. Neither will whatever comes next. We spend our days talking about the shiny layer on top. The models, the apps, the demos. This week is about the layer underneath. The man who wrote the rules that let networks talk to each other is finally retiring. The data centers powering the boom are drinking more water than they admit. Quantum is speeding up but still years from mattering. The robots are learning to fold our laundry. And the rules for all of it remain unwritten. The foundation is never the headline. It should be.

Here are five things tech worth your attention.

The ‘Father of the Internet’ is finally retiring

Cerf, 83, and collaborator Robert Kahn are credited as being the architects of the networking protocols that became the internet we know today. His work developing and popularizing TCP/IP — the basic set of rules that lets different computer networks talk to each other — beginning in the 1970s has been recognized with numerous honorary degrees, the Presidential Medal of Freedom, and a Turing Award, among other honors.

Since 2005, Cerf has served as vice president and chief internet evangelist at Google. (At this point, we can safely say the internet is fully evangelized, for good or ill.)

I had the pleasure to meet Vint Cerf some ten years ago or so at a luncheon in Berlin where we discussed tech regulation. It was a delight to listen to him talk about the inner workings of the internet.

(…continue reading.)

AI Data Centers Use Far More Water Than Most Tech Giants Report

Recently, Nvidia said it had solved the data-center water issue, showing off a closed-loop cooling system that doesn’t require additional water once filled. This design appears to be a win on two levels, says Gillingham, the Yale professor: It zeroes out direct use of water while also reducing the total amount of energy required for cooling.

Microsoft has committed to similar closed-loop tech. In 2024, the company announced that all its new data centers would use it, starting in 2027.

Unfortunately, most existing data centers use evaporative cooling systems that are energy-efficient but water-hungry, according to data from the 2024 Lawrence Berkeley report. Experts say retrofitting those could be prohibitively expensive.

Data centers are creating huge problems. But data centers are also hugely important and they need to be run profitably. So clever engineers will find solution to all the issues we currently see: too much water usage, to much power usage, too much noise - the list goes on and will be addressed.

(…continue reading.)

Quantum computing is about to get a lot more real

Capital is only one accelerant. The generative AI boom is also helping quantum developers move faster. Artificial intelligence coding tools are helping scientists accelerate the development of materials and components used in quantum designs, says Heather West, an analyst at IDC. That, in turn, is shortening the timelines on product road maps.

Industry consolidation has also helped, often by bringing different parts of the quantum stack under one roof.

Even is Quantum development is speeding up, I do not think we’ll notice anything really happening in this space in the next 5 years. It is still very early.

(…continue reading.)

We Didn’t Build the Atomic Bomb This Way

Artificial intelligence, funded overwhelmingly by private capital, has careened forward despite immense concerns about the effects it will have on labor, education, science, defense and civic life. A.I. companies have outpaced public oversight and, at times, successfully lobbied against it. The central achievements of the industry, the proprietary “frontier models” developed by companies like Anthropic, OpenAI and Google, are guarded intellectual properties even as they are incorporated into schools, offices, hospitals, courts, commerce and our everyday devices. The public did not ask for these A.I. tools and now can hardly opt out of them.

There is no one-size-fits-all solution to addressing the impacts that may be coming. But the scale of the concerns requires ambitious responses that serve the public. The United States can start by building a national A.I. laboratory.

This debate is just a few years too late, but it is still necessary that we discuss this.

(…continue reading.)

This Soft Robot Wants to Fold Laundry Without Pretending to be Human

Isaac 1 is meant to handle its main tasks autonomously, though Weave says teleoperation support can step in when needed. That point may be one of the more realistic parts of the project. Homes are inconsistent environments, full of changing furniture positions, loose objects, and edge cases that are difficult to predict. A robot that can ask for help may be more plausible than one claiming to solve every room by itself.

Folding laundry and picking up stuff my kids drop everywhere - this could be awesome. A professional cleaner is still cheaper though…

(…continue reading.)

That’s all for now! Thanks for reading! If you missed last week’s Five Things Tech, you can find it here:

🤖

— Nico

Subscribe now

We spent two years being told AI would change everything. This week we get to read the receipts. Enterprises are throttling usage because the tokens are too pricey. Architectural bloat does not beat one strong agent once cost enters the math. Google quietly drops a boring file format that might matter more than any of it. And the economists measuring the whole shift cannot agree on the basics. The technology is maturing. So are the questions about what it is worth.

Read on, my dear! Or have your agents tell you what to read first!

Subscribe now

Companies Are Throttling Employees’ AI Use Because It’s Too Expensive

The news shows the looming fallout from companies adopting AI as quickly as possible, and AI providers’ moves to charge enterprises based on how much they use AI rather than a flat fee. Emails obtained by 404 Media even show some companies cutting off access to some AI models altogether in an attempt to stop burning through their AI tokens, and big tech companies like Adobe are ending unlimited access to Claude.

We are currently in a stage where we have to figure out what works and what doesn’t. As always, some companies moved faster than others and then had to recalibrate their efforts. It is a sign of the maturing of the industry, not a sign for concern.

(…continue reading.)

Open Knowledge Format (OKF): The Markdown Standard Built for AI Agents

The agent boom has created a strange bottleneck. Models are getting better. Tool-calling is getting more reliable. Protocols like MCP are making it easier for agents to connect to external systems. But the knowledge those agents need — what a metric actually means, which table joins to which, how an internal process works, why a dashboard number changed last quarter — is still scattered across wikis, data catalogs, shared drives, code comments, Slack archaeology, and the heads of senior employees.

Google’s bet with OKF is simple: before every company builds yet another proprietary knowledge layer for agents, maybe we should agree on a boring file format first.

Granted, this is still early with OKF 0.1 just being released, but when they did, I was kind of suprised that not too many people took note. Google has developed and released quite a few standards around interoperability in the Agentic AI space and hopefully they will hand this over to the Agentic AI Foundation as they did before with the A2A Protocol stack.
(…continue reading.)

Smooth AI criminal drives ‘first’ end-to-end agentic ransomware attack

JadePuffer’s “self-narrating” payloads “contained natural language reasoning, target prioritization, and the kind of detailed annotations that human operators don’t often write but LLM-generated code produces reflexively,” Clark added. “The operation also adapted in real time, retrying failed steps within refined parameters. In one sequence, it went from a failed login to a working fix in 31 seconds.”

After exploiting CVE-2025-3248, a missing authentication vulnerability in Langflow that allows remote, unauthenticated attackers to execute arbitrary Python on the host, the AI agent began scanning for and collecting secrets, including LLM provider API keys, cloud credentials “with explicit coverage of Chinese providers” including Alibaba, Aliyun, Tencent, and Huawei, while also scanning for AWS, Azure and Google Cloud Platform, cryptocurrency wallets, and database credentials.

What a time to be alive, we now have agentic criminals.

(…continue reading.)

Multi-agent AI keeps collapsing back into one agent. A fair test just proved it.

They ran the audit across a stratified set of backbone models (GPT-4o, GPT-5, GPT-OSS 120B, and Gemini-2.5-Pro) and a spread of hard benchmarks, including GPQA-Diamond, HLE-Math, SWE-Bench Lite, and a deep-research task called BrowseComp-Plus. The guiding question was simple: do these automatic systems deliver consistent, cost-effective gains over a strong single-agent baseline?

The answer was mostly no. Across the ecosystem, the automatically designed systems rarely outperformed a strong single agent once cost and baseline strength were accounted for. The authors describe this as evidence of “architectural bloat,” sophisticated agent graphs that do not translate into functional reasoning gains.

These findings are really interesting, yet counter-intuitive. I’d love to know how this would now look like with a bunch of agents running on cheap LLMs compare to Fable 5.

(…continue reading.)

A.I. Is Reshaping the Economy. Good Luck Measuring How.

Researchers can’t even agree on basic questions like how many companies are using A.I. or which workers are most vulnerable to the disruptions it could cause.

The conflicting signals partly reflect the challenge of detecting economic shifts in real time. Government statistics are inherently backward looking, and they are better at measuring broad trends than developments in specific sectors or regions. New technologies that might lead to the emergence of new products, jobs or entire industries can be particularly difficult to measure.

Plus: AI comes in so many different flavors and usecases, the impact on each and every company when “using AI” can be totally different as well.

(…continue reading.)

If you missed last week’s edition of Five Things AI, you can read it here:

That’s it for Five Things AI this week! 🤖

— Nico

I had the brilliant idea to go running during the most extreme heat wave we ever had in Northern Germany. I tried to go running early in the morning, but then life happened and I finally left the house around 11 and it was already getting hot outside. The heat and my general lack of fitness was not the best combination for a run, but I had a good time nonetheless, even though it felt as if I could drink buckets of water afterwards. Unfortunately I hadn’t read the really interesting article about hydration before my run, but you will be prepared if you read that article - and the other four.

I am still trying to get into the running flow again, but the late night soccer live broadcasts are not really helping with getting up early…

Subscribe now

2026 Western States 100 Results: Jenn Lichter and Vincent Bouillard Set Course Records in Cool Weather

It was safe to say that expectations were high for a new record or two this year as runners set out on the 100.2-mile journey across the Sierra Nevada over a course that challenges with its 18,000 feet of climbing and 22,000 feet of descent.

In a testament to the ideal race temperatures, and the exceptionally strong field, the second to fourth runners – Francesco Puppi of Italy, Ryan Montgomery, and Thomas Cardin of France, were also all under the previous men’s course record, with Zach Miller’s fifth-place time still the 10th fastest time in men’s race history.

I spent some time watching this and while I do not believe that these kind of races will ever become mainstream spectator events, the performances each year are nothing but extraordinary.

(…continue reading.)

‘Without Fear:’ Trail Running Is More Than a Sport for These Run Clubs

Between President Trump’s immigration crackdown and a rise of racially motivated hate crimes in recent years, it hasn’t been easy for people of color to feel safe, according to the runners interviewed for this story. But now trail running groups like Sin Miedo are popping up around the country, giving more Americans a chance to get outside.

“ICE was coming in May of last year and taking people away from their families,” Marrero said. “We made a conscious decision to say, ‘We’re going to show up.’ These things happening around us are awful and destroying our community, but we’re going to help people escape for an hour or two. And so many people have shown up in the group, providing help to those who need it. Places to stay, food, whatever… The trail running has been the catalyst for us to continue showing up for each other.”

It’s sad that this is necessary, but it is wonderful how run clubs can create community and make people feel better.

(…continue reading.)

Running Is Free Until You Become a Runner

The assumed cost of running is always physical. The repeated impact, the stress on your joints, multiplied across every stride. The average 5-mile run is 10,000 steps. Do the math on your own mileage and the strain we put ourselves through becomes hard to ignore. We keep going anyway, until we eventually need new knees - at least according to every non-runner with one podcast mic and zero medical credentials.

But lately I’ve been wondering about the other cost…

The financial one.

Running is supposed to be the simple sport. Basically free. A pair of shoes, a road, maybe a watch if you’re trying to flex online. But after five years of diving headfirst into the economy of running, I can say with confidence: this sport is not free.

I made a deliberate decision about a year ago to ignore all those in the runnings space who only talk about the latest products and want us to buy more and more. Turns out that I can run in last season’s shoes and in shorts I bought years ago.

(…continue reading.)

Is Hydration Just as Important as Your Carb Intake?

Hydration is not secondary to carbohydrates. They are equal partners in performance. Understanding your sweat rate and sweat sodium concentration gives you the tools to create a personalized hydration strategy that supports your training, keeps you safe, and helps you perform at your best.

Taking time to get hydration right pays off on every long run and every race.

Drink up!

(…continue reading.)

Unofficial Rules of the Trail

There are no official “Rules of the Trail.” No one hands you a rulebook when you take your first step onto the dirt, and there’s no pre-race briefing where someone explains that you should always pack an extra snack or jacket, that the best climbs are for hiking and soul-searching or that small clothing problems will, without exception, become very big clothing problems. These are things you learn the way most trail runners learn everything: slowly, stupidly and usually the hard way.

Spend enough time on trails – long runs, wrong turns, thunderstorms, the emotional rollercoaster of running for 10 or 20 or 30 hours – and you start to develop a set of rules. Not rules you have to follow, but rules that, if ignored, will eventually teach you why they exist.

Experience can be really helpful when you’re on the trail, so learn from others and havea look at these rules.

(…continue reading.)

If you missed last week’s edition, you can read it here:

Now, go running!

— Nico

🏃🏻‍♂️

Dang, it’s hot. We almost scratched 40 degrees Celsius today. I live in the North, I want snow in the winter and a really hot day should be just above 30 degrees.

Anyhow, to cool off, we went to the pool on Saturday and we drove to the pool my dad always took us in the 70s and 80s. Nothing much has changed at that pool and it still has the charm of an outdoor pool in a small village. It has a nice slide, some trees for shade and of course french fries. Obviously, nothing is as good as “Freibad-Pommes” - the fries they serve at an outdoor pool just hit different.

Earlier this week I went to Kiel to meet my brother and his family and stroll around Kiel Week, which is not just a huge sailing event, but also has plenty of things to do for people all ages, especially if they are interested in food and beverages… and I noticed that Corn Dogs are getting more and more popular!

Subscribe now

Heat waves can leave homes dangerously hot – even for young, healthy adults, as Texas shows

Within hours, indoor temperatures can climb well above what the thermometer shows outside, especially on upper floors and in rooms with south-facing windows. Over longer periods, especially if temperatures don’t cool off overnight, conditions can become lethal.

Most heat-related deaths occur indoors. When a heat dome sent temperatures soaring in the Pacific Northwest in 2021, 98% of the more than 600 deaths in British Columbia happened inside homes. Washington and Oregon also saw high numbers of deaths in homes that lacked air conditioning.

In Europe, where only 1 in 10 households have air conditioning, heat waves killed an estimated 60,000 people in 2022 and 47,000 in 2023, largely inside buildings never designed for these temperatures.

My wife and I have discussed buying an air conditioning device for the first time. This is how hot it got around here. The staircase of our building is nice and cold, but sleeping there is probably not an option. Rain has been promised and then we will complain about a summer that is too cold and too wet again.

(…continue reading.)

America’s Big Mistake in Iran

When the United States and Israel launched the war on Iran in February, their plan was simple: bomb Iran until either the Iranian public rose up and overthrew the government, or the existing government capitulated to American demands. It rapidly became apparent that neither was going to happen. The Iranian people didn’t revolt against their oppressors. The Iranian government hunkered down, closed the strait, and gambled that the U.S. would be unwilling to invade or strike at crucial infrastructure.

So it seems U.S. planners made an obvious, if common, mistake: They assumed that a war could be won via aerial bombing alone.

I still think that after decades of diplomacy and embargos, the war against the Mullah Regime was justified. I did however assume that there was more planned than just bombing Iran for a week or so., before realizing that Iran controls the the Straight of Hormuz.

(…continue reading.)

AI Warfare Is at the Point of No Return. What Now?

Today, AI-guided weapons can autonomously home in on objectives a controller picks. That selection generally happens when weapons approach a target, involving a few drones with limited firepower.

Soon, though, swarms of drones will independently cross great distances by air, water or land to hunt down and strike targets without human intervention. And targets won’t necessarily be on a battlefield.

Killing isn’t AI’s only military assignment. Its role is ballooning across all the less-visceral chores that militaries tackle, particularly in giving priority to intelligence for selecting targets. U.S. commanders say they are selecting targets at more than 10-fold the tempo in Iraq. In the Ukrainian National Guard’s Khartia Corps, automation has tripled the pace of missions, said its top drone engineer.

Now would be a good time to have reasonable people in charge to figure out how we can prevent this from totally escalating.

(…continue reading.)

Europe Is Fed Up and Wants Its Own AI

It sounds almost delusional for Europe to think that it can build the world’s second-best AI. More than 20 nations would need to work closely together, overcome their continental impulses to strangle innovation with red tape, and lure unprecedented sums of investment. Most of all, Europe must shift from a risk-adverse mindset to a moonshot mentality. But Macron has made some progress. His “Choose France” initiative has won pledges of over 100 billion euros in AI infrastructure, anchored by Softbank's 75 billion-euro commitment to build huge data centers in France—pending approvals, of course.

It just takes too long for European leaders to understand the impact of a new technology. It happens again and again, both in politics and business. And even now not all leaders fully understand what is happening.
(…continue reading.)

Heads in the game

For FIFA, soccer’s international governing body, SAOT is among the latest in a portfolio of innovations used at the World Cup. From goal line technology to video assistant referee (VAR) tools, officiating tech is now commonplace at the top level of the game.

But SAOT is part of a broader sports technology landscape that stretches far beyond soccer. And one of the major players in that landscape is the very team that collaborated with FIFA to bring SAOT to the pitch in the first place: the MIT Sports Lab. Founded in 2015, the lab focuses on using technology and data science to tackle real problems facing athletes, teams, and sports organizations and brands.

While I do acknowledge that the tech is both amazing and is getting better and better, I really don ‘t want all that. I want 90 minutes plus up to two minutes of extra time, I want the refs to make the decisions and I want the fans to debate the decisions. It’s part of the fun!

(…continue reading.)

That’s it. Have a great Sunday! If you missed last Sunday’s edition of Five Things, have a look here:

Subscribe now

— Nico

Here are five things tech worth your attention.

The state of the AI economy

These revenues are growing faster than previous IT-oriented waves, roughly three times more rapidly than the mobile or Internet waves.

While many companies have moved beyond occasional pilots, they are still in the early stages of scaling and deepening. In conversations Azeem has had with senior execs across a range of industries in Europe and the US (from industrials to insurance, from finance to pharma), the consistent message is that they intend to invest more heavily in AI in the coming years. Companies are also becoming more vocal about the impact of AI on earnings calls, with the caveat that half of the surveyed CEOs believe their jobs depend on getting AI right.

I don’t think this AI thing is going away anytime soon.

(…continue reading.)

The New Software Lifecycle

Static context is loaded on every turn, so it’s reliable and expensive. Dynamic context is loaded on demand, so you only pay for what a task needs.

Static context is loaded every turn: system instructions, rule files (AGENTS.md, CLAUDE.md, GEMINI.md), global memory, core guardrails. It’s reliable, and it’s expensive, because you pay for it on every single call. Dynamic context is loaded on demand: skills that fire when a task matches, tool results, documents pulled from RAG. You only pay for the bits a given task touches.

Get that balance wrong in one direction and you burn tokens and bury the signal. Wrong in the other and the agent forgets the rules that keep it safe. The paper’s advice, which I agree with, is to treat the boundary as a real architectural decision: reviewed in a pull request, versioned like code.

The trick that makes dynamic context scale is Agent Skills with progressive disclosure. The agent sees a little metadata at startup, loads the full instructions when a task matches, and only pulls in the heavy reference material when it actually needs it. That’s how one agent can carry dozens of skills and still only pay for the one it’s using.

We truly live in interesting times. I am constantly revising my setup and while it is getting more complex every week, the tools to manage the complexity are also getting so much better. Some I even write myself, or at least tell the agents to write them…

(…continue reading.)

The A.I.-Design Aesthetic That’s Taking Over the Internet

The advent of A.I. has caused Silicon Valley types to fixate on the concept of “taste”; when machines can spit out images and words instantaneously, the thinking goes, humans must be discerning enough to separate the quality output from the slop. Ström-Awn described Claude’s default design choices as a mark of complacency, “a filtering mechanism” that exposes “people who didn’t spend the time” creating something unique. Nguyen brought up a credo of the mid-century American designers Charles and Ray Eames that defined the ideals of modernism: “The best for the most for the least.” What Claude Design offers, Nguyen said, is “the pretty good for the most for the least effort, and pretty low cost.” She continued, “You’re just paying a twenty dollar a month Claude Pro subscription instead of hiring a designer. Is this the modernist world we wanted?”

I actually think that this phenomenon is nothing new - we have a new wave in design and soon we will see the next wave and then everything will be a bit more sophisticated. The really cool thing is: even people like myself (I called myself web-designer 30 years ago just because I could do very basic website design) can now create pretty complex websites with good UX/UI.

(…continue reading.)

Smartglasses Are Inevitable. But What—or Who—Are They For?

Since the start, the idea of a face-worn computer has garnered skepticism, even hostility, and it’s understandable.

When most of us are trying to reduce our time in front of screens, it seems absurd to mount them right in front of our eyeballs. And why should we be OK with everyone we meet pointing internet-connected cameras at us? It feels like an assault on what little privacy we have left.

Then there’s the built-in AI. Artificial intelligence lets these glasses identify objects in our field of view, like a ficus tree or a Southern African meerkat. But there’s a growing unease with companies sticking generative-AI features into any and all products, and having a little AI djinn always whispering in our ear seems like a cautionary tale—or a “Twilight Zone” episode.

And yet there will be smartglasses, and you and I might even choose to wear them. They’re a technological inevitability, a new consumer-tech arms race for the last and best real estate on our bodies.

I will buy smartglasses as soon as they get progressive lenses, automagic shading and, most importantly, will be self-cleaning.

(…continue reading.)

Why ATMs didn’t kill bank teller jobs, but the iPhone did

So what happened to bank tellers? Autor, Bessen, Vance, and the like are right to point out that ATMs did not reduce bank teller employment. But they miss the second half of the story, which is that another technology did. And that technology was the iPhone. The huge decline in bank teller employment that we’ve seen over the last 15-odd years is mainly a story about iPhones and what they made possible.

But why? Why did the ATM, literally called the automated teller machine, not automate the teller, while an entirely orthogonal technology—the iPhone—actually did?

Technology hits differently. But it brings constant change and disruption - yet we are still not getting used to it.

(…continue reading.)

That’s all for now! Thanks for reading! If you missed last week’s Five Things Tech, you can find it here:

🤖

— Nico

Subscribe now

This week the AI story splits in two directions at once. On one side, the money finally works: revenue is clearing the cost of the hardware, generative AI is scaling three times faster than the cloud ever did, and the buildout that looked like a leap of faith is starting to pencil out. On the other side, the mood is souring and the engineering is getting smarter. People are asking why they should love a technology that concentrates wealth and burns through power, while researchers quietly demonstrate that you don't need a trillion-parameter Nobel laureate to do most of the work. Coupled oscillators that compute with physics at a fraction of the energy, knowledge agents that beat frontier models with better structure, world models that make agents sharper without ever training them as agents. The pattern this week: the economics validate the giants, but the cleverness is migrating to whoever builds lean. Here are five things worth your attention.

Subscribe now

Why Does Everyone Hate AI?

Finally, AI is tightly linked in the public mind with the tech oligarchs who are pushing it. There is widespread awareness of the growing concentration of wealth and power at the top and how this is distorting our politics and harming our society. Aside from the MAGA faithful, Americans overwhelmingly favor government policies to reduce wealth inequality:

And AI is widely perceived, for good reason, as a technology that will increase the concentration of wealth at the top. Indeed, as I said, the AI companies themselves have already told us that the technology will have extremely negative effects on workers.

I am certain that AI will bring plenty of positive effects and will revolutionize the way we work and live. Founders like Altman and Amodei need to calm down the fears and focus more on the amazing potential AI has in medicine and in science in general. And while data centers are consuming plenty of energy, I believe that AI will lead to numerous breakthroughs that will reduce global warming.

(…continue reading.)

AI Demand Begins to Justify Massive Cost of Data-Center Buildout

Much of the AI boom has been measured from the supply side, through disclosures from public semiconductor companies like Nvidia Corp. and hyperscalers like Alphabet. Demand has been harder to quantify because many of the most important AI labs, including OpenAI and Anthropic, remain private.

Generative AI revenue, excluding China, reached $110 billion over the past 12 months and is scaling three times faster than any previous information technology wave including the internet, mobile applications and the cloud, according to the report.

When we look at the Gartner hype cycle, we can assume that AI will quickly reach the plateau of productivity, which we can also see by looking at the revenue numbers.

(…continue reading.)

Introducing Un-0: Generating Images with Coupled Oscillators

At Unconventional AI, we’re building a new kind of computer, one that harnesses the laws of physics to do the computing. Our goal is to run modern AI on a fraction of the energy today’s machines need, around 1,000x less. As a first step, we ask: can we train a physical dynamical system to generate images at scale?

The best AI models today are conventional deep networks with transformer backbones. However, there is also a long history of alternatives that seek energy efficiency by leveraging the dynamics of a physical system, such as the noisy, time-varying behavior of analog circuits that compute with analog voltage and current instead of conventional digitized numbers.

Wow. I’m intrigued. The power consumption of traditional AI models is extreme and even if they don’t get 1000s less, it would be huge.

(…continue reading.)

Knowledge Agents: Beat Frontier Models with Better Structure

First, a significant portion of frontier models’ huge footprint is “knowledge.” I might call it pseudo-knowledge, since it’s probabilistic and there’s no guarantee it’ll give you the right answer... but the biggest models have been trained on an enormously broad set of data. This is captured in numerical weights as “parametric” knowledge. While that’s very useful if you’re casually asking Claude Opus or GPT-5.5 about some random topic, it’s entirely irrelevant if you either already have the data you want to reference or the data isn’t publicly available anyway—so it could never train on it. The latter is quite common in fields that are specialist (areas of medical research), secretive (high finance), or proprietary (frontier, company-specific materials science). If I don’t need it... well, a lot of the massive size of the model to cover every random subject matter is a huge waste.

I liked this article a lot, basically because somebody finally pointed out that just using the best LLM out there is a bit lazy - and expensive.

(…continue reading.)

Alibaba’s model never trained as an agent — and improved agent performance across seven benchmarks

The research team trained agents inside the resulting simulator and found performance gains that exceeded what training against real environments alone produced. In a separate test, using world model training as a warm-up before agentic fine-tuning improved performance across seven benchmarks, including three the model had never seen during training.

The paper accompanying the release identified a gap in prior agent research. “We argue that world modeling is a crucial missing piece in the path to general agents.”

This means that we will see another boost in Agentic AI soonish. We live in interesting times.

(…continue reading.)

Read more

Sunday morning I saw a run club running by and I was appalled. They all ran around in the same white t-shirt with a sponsor logo on it. The person at the front was holding a flag with the same logo, a cosmetics company.

Seriously, this is what run clubs have turned into? Just some weird advertising campaign to reach a specific target audience? I don’t mind companies supporting and sponsoring run clubs, but I this takes it too far and I hope it will backfire. Running is such an individualistic sport, why would anyone try to make it all corporate?

Subscribe now

Midlife is the perfect time to start trail running – here’s how to get into it

Trail running takes place on dirt trails as opposed to paved roads or measured flat tracks. Races longer than marathon distance (26.2 miles) are called ultramarathons, or “ultras” for short. While some ultras are held on tracks or paved surfaces, most follow trail routes. Hence, “ultrarunning” typically refers to extra-long-distance trail running.

Trail running, especially at ultra distances, differs significantly in practice and culture from traditional road-and-track running. Trail races tend to be smaller and more intimate than metropolitan road races.

I actually think any time is a good time to start trail running and if you think about it, it is really not that strange an idea to run outside in nature.

(…continue reading.)

Running at Night

If you’ve ever struggled with finding the confidence and knowledge to embrace the nocturnal world, guided by nothing but the glow of your headlamp (and, hopefully, a well-planned GPS route), keep reading. A disproportionate number of women fears running on trails, let alone running on trails at night. The main threat to female runners in any environment is harassment from men, and that is a very valid fear. Although this threat is exponentially higher in a more populated area, it exists on roads and trails alike.

While various threats exist on the trail (including our fellow humans), most Colorado runners keep a nervous eye out for mountain lions. We share the woods with black bears too, though they tend to be more skittish than lions; in my experience, bears make plenty of noise crashing through the brush just to stay clear of us. Of course, a mama bear with cubs is always a concern, but just because you haven’t seen a cougar doesn’t mean it hasn’t seen you. To put these fears into perspective, let’s look at the stats.

I am not a big fun of running in the dark, but then again where I run, there are no mountain lions, which really puts it into a different perspective.

(…continue reading.)

Sweat, tears and camaraderie as 20,000 runners take on world’s largest ultramarathon

That first year, 34 runners, all white men, lined up for the race, conceived by the first world war veteran Vic Clapham as a way of honouring his fallen comrades. Sixteen of them finished. More than a century later, on 14 June, more than 20,000 people stood outside Durban city hall, hoping to make it to Pietermaritzburg before the 12-hour cutoff.

What started as an all-white, all-male test of physical endurance has become part of the fabric of South African life, something so ordinary that you would be hard-pressed to find someone here who does not know a Comrades finisher.

Running clubs bus in from all over the country. Security guards and shop workers line up alongside bankers and celebrities. And, for one day, every June, South Africa’s searing racial inequality seems to melt away.

What an event. Every year I read about it and every year I am amazed how big this race is and what the runners endure to finish it.

(…continue reading.)

What I learnt from running 100 miles

The mind is a very odd thing. Generally speaking I have a lot of mental resilience and have never wanted to give up in a race. My mind is usually far stronger than my body. So it was a very weird feeling when self-doubt arrived very early on.

The first 10 miles to be precise. I started the race feeling really tired as I’d been up since 4 am and hadn’t slept well the week before. Plus I had only just got over jetlag from my recent Easter Island trip.

I love these write-ups about amazing achievements as a runner.
Running 100 miles is truly a big deal and I am sure it changes a lot in a person’s life.

(…continue reading.)

O is for Offline

The easy solution to being constantly connected to an ecosystem of running apparel pyramid schemes is to do more things that are offline. You have a very distinct number of hours in your day. If you use them for something purposeful that makes you feel good, you simply won’t have time to be on your phone or computer.

It’s like if you eat a lot of vegetables and whole foods, you don’t have space in your belly to overeat highly-processed foods. Start with a positive act and you will organically live better. Leave a little space for a chicken nugget, of course.

Yes. Don’t buy so much apparel you don’t need, you’re just going to sweat in it anyhow and I remember when cotton t-shirts where all we had for sports and it didn’t kill us… also, buy This is Running - it is such a great book!

(…continue reading.)

If you missed last week’s edition, you can read it here:

Now, go running!

— Nico

🏃🏻‍♂️

Friday was a day to remember. It was my birthday. I turned 54. I woke up just after 6 in the morning when rain, thunder and lighting turned the morning into something dark, very wet and loud. So I told my kids that I’ll give them a ride to school. We went out to the car and you probably know this situation when somehow the car doors do not open and everyone is pulling the door handles? That’s what we did, only while standing outside in torrential rain. We went back inside, my daughter got the other car key from the apartment and I tried again. She finally managed to get the little key thingy out of the key fob so I could open the door. It worked. They came running to the car again. I couldn’t open their doors. I felt like being in a movie where some evil overlord is shutting down computer systems to create havoc. After about 15 minutes of trying to get the car going, I stopped the experiment, took the kids to the subway and went for a walk with the dog. When I came home I found out that the DWR coating of my jacket wasn’t as good as I remembered it. Then the rain stopped, the sun came out and my birthday turned into a really wonderful day with nice presents, great food and lots of time with my family. The car will now be towed back to the dealer, since it didn’t handle the latest software update as expected. Thankfully my wife managed to wrangle down a hotline again and we got a brand new Volvo XC90 in exchange for now. Of course I always remembered how my dad used to say that he doesn’t want any tech in his car, so we always had to manually roll the windows up and down as he didn’t trust electric windows… when I tell my kids about that, they probably think I was raised by barbarians, not to mention their astonishment when I tell them about black and white TV or landline phones and no mobile phone until I was in my mid-20s. Anyhow, I had a great birthday!

Subscribe now

The Art of the Non-Deal

It is clear that Trump is being driven to reopen the Strait of Hormuz at virtually any cost by the domestic pressure from rising oil prices and inflation. Being unwilling to send ground forces to Iran, he has had few cards to play over the past six weeks to get further Iranian concessions. So he has chosen to back down and accept a return to the status quo ante from before he began the war on February 28.

The world will indeed be better off if the Strait is re-opened. Perhaps Trump’s hardcore MAGA supporters can be persuaded that he has negotiated a consummate deal and achieved a great victory. But everyone else will understand that the world’s most powerful country is being run by a feckless and ignorant president who will impose immense costs on both other countries and his own people if he thinks it will benefit himself.

What I really do not understand about this war: the USA and Israel had established aerial supremacy, they could do endless sorties as they wanted and destroyed rocket launchers, ammunition depots, military compounds, infrastructure and what not. Why the fuck didn’t it occur to anyone that the Strait of Hormuz is so important for international trade and could therefore quickly be turned into the achilles’ heel? Especially since the Houthis did the same just a little over a year ago when tried to close access to the Red Sea. Unbelievable that the USA didn’t see that coming, and then again, I didn’t really expect Trump and Hegseth to be able to formulate a coherent strategy.

(…continue reading.)

Trump Put His Own Interests Above All in the Iran Deal

You don’t need to be a foreign policy expert to see what happened here. You need to be a domestic policy expert. Trump sold out America’s ally in the war, Israel, and the Arab Gulf states for the swing states of Pennsylvania, Georgia and Michigan. Trump knew that the food inflation and high gasoline prices triggered by this war were a prescription for a Republican wipeout in the midterms. He had to stop the war now to get prices down by November, because if the Democrats take the House and Senate, Trump will be looking at endless investigations into how he has used the presidency to enrich himself and his family — and possibly even impeachment.

So, Trump did what he always does: He abandoned all principle and all allies and put his personal interests above all other considerations.

I am sure Trump never had any other principle than making sure to use his office to increase his family’s personal wealth. What a disgrace.

(…continue reading.)

How will AI make moral decisions for you and me?

The short answer is we need a moral psychology of AI for the same reasons we need moral psychology of humans. We need to understand what drives humans to be more or less moral, in order to promote a more moral society and more moral behavior in our institutions. Now that AI machines are increasingly interfacing with the world, I think we need to do the same for them, since they will be making decisions with moral consequences.

We need to understand what moral principles they follow, if any, and to what extent these are due to their training, their programming, the datasets they were fed or the contexts in which they operate.

Oh, this is such an interesting topic and my brain starts to hurt when I think it through, knowing what I know about AI - and the tech business.

(…continue reading.)

In Praise of a Dumb House

Within the next year, Samsung will begin embedding Google Gemini directly into Bespoke AI refrigerators, microwaves, and ranges. Do I want my fridge cameras scanning my groceries (the images are called “shelfies”) and ordering more? LG’s Signature Oven Range has introduced Gourmet AI, which recognizes your dishes and automatically applies what it deems to be optimal settings. AI Browning monitors bread and sends notifications when it’s ready. But, like, I have eyes. A fridge that informs me if my milk is spoiling? I have a nose. Do I really need AI to tell me when fresh food is good or bad? What if I suddenly can’t turn off this allegedly smart oven and burn my house down?

I love tech and I love playing with new gadgets. But there is so little appeal for a smart home for me. I’m using Alexa as loudspeakers and to set a timer in the kitchen. I take a screen into the kitchen when I am cooking. Everything else is just tech for tech’s sake, without any real use case, at least for me. Also, I want to decide myself what to buy and what to cook, or rather carry out the decisions made by my wife.

(…continue reading.)

Where Has All the Cottage Cheese Gone?

Until recently, cottage cheese had been dismissed as a soulless diet food. Sales of the product peaked in the 1970s, when the average American ate about five pounds of it per year, according to the U.S. Department of Agriculture. But dairy manufacturers began to refocus their processing capacity on yogurt, which eclipsed cottage cheese sales by the mid-1980s.

A growing obsession with protein among American consumers has given the white curds a new life. A few years ago, online fans began posting about “protein-maxxing” with cottage cheese, adding it to ice cream, smoothies, flatbreads, bagels and pasta dishes. TikTok creators became cottage cheese converts, enticed by the product’s roughly 14 grams of protein per serving.

I have been eating cottage cheese for breakfast almost daily for many years. It helps me lose weight or at least not gain too much. And I really do like it, either pure or with fruit.

But, when in the USA, I have learned to avoid cottage cheese. All I ever found in the grocery store was gooey, over-priced and nasty-tasting stuff. Interesting to know that even this kind of cottage cheese is flying off the shelves there, it’s probably because of all the stupid “now with more protein than ever!”-labels.

(…continue reading.)

If you haven’t already, please drop everything and subscribe to Five Things now!

That’s it. Have a great Sunday! If you missed last Sunday’s edition of Five Things, have a look here:

Subscribe now

— Nico

disposable software: software is now just paper plates

a year ago the best AI coding tools could finish a function. today it can read an entire codebase, plan a change, edit dozens of files at once, run the tests, fix what breaks, and ship a working version of whatever was asked for. that did not exist twelve months ago.

the engineer typing the code used to be the bottleneck. that is gone. describe what you want in plain english and a working app shows up. one human plus one model ships on a saturday what used to take a team a quarter.

anything that takes a week to ship today will take a day to ship by december.

so what is the point of building it to last?

So the big question is: how will we keep everything running if it is easier to build new stuff than keeping the old stuff alive? We need to figure out consistency, otherwise we will play catchup all the time with an everychanging code stack.

(…continue reading.)

How to tame AI’s voracious appetite for energy

Though experts say the fastest energy savings will come from software tweaks, some are also taking aim at the energy-hungry processing chips that fuel AI computations. Engineers have made chips increasingly efficient over time by packing more computing capacity into individual processors — reducing the energy required to shuttle data between chips that are working together to perform AI computations. Engineers have done this by shrinking the size of transistors — microscopic electrical switches that process data — inside the chips.

But because engineers are reaching the physical limits of how small transistors can be, “we need to think of alternate ideas to improve the designs,” says computer architect Ajay Joshi of the Boston University Photonics Center.

The recent boom in data centers is really posing some interesting problems that need to get solved. I assume that this will happen quickly, otherwise the costs of running these data centers would just be too high. And these advancements will benefit our daily lives in tremendous ways, from the electric grid to more energy efficiency to small language models on personal devices.

(…continue reading.)

Solar Beat Coal in US Electricity Mix for the First Time in May

Despite the Trump administration’s attempts to drive a coal revival, it has been steadily losing ground to other energy sources in recent years, squeezed out by cheap natural gas and rapidly falling renewable energy prices. Solar power, in particular, has been on a tear as prices drop exponentially.

Last month, the two lines finally crossed. Solar supplied 12.8 percent of US electricity in May, edging past coal’s 12.2 percent share to become the country’s third-largest source of power behind natural gas and nuclear, according to recent data from energy think tank Ember.

Well, it just makes more sense to do wind and solar than digging for coal or pumping oil and then ship it around the globe.

(…continue reading.)

The Secret Revolution in Battery Technology: 3-D Printing

In 2025, researchers published around 25,000 papers about 3-D printing of batteries and their components, yet only a handful of startups have even proposed commercializing the technology.

Attempting to beat existing battery giants has doomed many a company to the graveyard. New battery tech must compete on lifespan, durability, energy density, safety and cost—which is dependent on massive scale.

Some are trying to use 3-D printing to create efficiencies in existing battery manufacturing systems. A brave handful of startups are pursuing radical new designs and approaches. They’re starting with defense applications, where cost and scale are less of an issue.

A few months ago I met a startup founder who was building really small batteries for drones and other small devices. I was blown away by what they were doing and how powerful really tiny batteries have gotten. So incredible.

(…continue reading.)

AI and brain-computer interface allow speechless ALS patient to work a full-time job

A team of scientists from the University of California, Davis, published a paper Monday detailing a years-long study of a brain computer interface (BCI) system implanted in a patient with amyotrophic lateral sclerosis (ALS, also known as Lou Gehrig’s disease), which destroys motor neurons and causes loss of motor control and eventual paralysis. According to the team, their patient, Casey Harrell, has been living with BCI implants since 2023 that are still working today, giving him the ability not only to control a computer cursor with his thoughts, but also to speak.

I have serious concerns about brain-computer interfaces, but this is really awesome. I mean, just imagine how this can change the lifes of so many people for the better!

(…continue reading.)

That’s all for now! Thanks for reading! If you missed last week’s Five Things Tech, you can find it here:

🤖

— Nico

Subscribe now

SpaceX is buying Cursor for $60 billion in stock it printed on Friday, which is less a strategy than a confession: Grok cannot win enterprise on merit, so SpaceX bought the editor enterprises already had open. Anthropic, meanwhile, turned safety into a moat and is quietly winning the talent it needs while OpenAI spends its energy fighting itself. The public has feelings about all of this too, and they are not the ones the industry was hoping for: Americans now narrowly back banning AI from software engineering, of all jobs, the one that spent a decade as the ladder out. Europe is still thinking its nineteen thoughts and committing to none of them. And the labs are moving everyone onto metered pricing, because somebody has to pay for the data centers, and it is going to be you.

Five Things AI. Let’s go.

Subscribe now

SpaceX’s $60 Billion Deal to Buy Cursor Gives It More AI Coding Power

Cursor makes a tool that allows developers to toggle between different AI models, from OpenAI and Anthropic to xAI and Google and others. The company competes with Anthropic’s Claude Code and OpenAI’s Codex, which can write code, debug software and automate tasks.

The Cursor deal could help SpaceX win over lucrative enterprise clients, who have largely shied away from using its in-house chat assistant Grok, which competes with Claude and ChatGPT. Elon Musk has positioned Grok as an anti-woke, truth-seeking chatbot, and it remains behind the leading models.

I didn’t see this coming, but it shows just how desperate Musk is to get some users for Grok. Now that SpaceX is publicly listed, people will soon find out just how much he overpromised.

(…continue reading.)

Anthropic’s Safety Superpower

Here’s the thing about these safety justifications: I think they work because, to Anthropic, they aren’t justifications. The company really believes that they are the only ones who believe in super intelligence, and thus are the only ones who are sufficiently concerned about the dangers. That excuses decision after decision, policy after policy, and confrontation after confrontation that, to people on the outside, look like a bizarre combination of cynicism and naiveté.

The contrast to OpenAI is massive: I think that one way to understand how and why OpenAI lost its lead is that, in the years following the release of ChatGPT, the company has been at war with itself internally as what used to be a research lab was suddenly seized with the burden of being the accidental consumer tech company; to the extent OpenAI solved that conflict, it was by bleeding huge amounts of talent to Anthropic in particular.

It’ll be interesting to see how the two IPO will look like.

(…continue reading.)

Americans want artisanal code

Overall, Americans supported the ban on AI in software engineering by a split of 44 to 41. It’s a small margin but a notable one given that bans for many professions were underwater by double digits.

One obvious explanation for this is that the public responds to a drumbeat of news stories about whatever industry is being automated next. The industry with the second-most support for a ban was manufacturing, which was sitting at break-even support (44 to 44). For a long time, manufacturing has been the stereotypical job facing automation, but now that software engineering is in that spotlight, Americans are willing to protect it.

That explanation is probably true, but it’s also a big step from hearing that an industry could be automated to supporting a ban to protect it.

The surprising support for a ban on the automation of software engineering likely also stems from the fact that it has spent over a decade as the field that represented social mobility, including for those who didn’t go to an elite school.

I doubt that this sympathy for the coders will last long.

(…continue reading.)

Nineteen thoughts on AI and Europe

6. It may seem implausible for a democracy to commit economic self-harm on that scale. But, through the energy transition, Europe has signaled that governments will tolerate destroying industries as important as the German car and chemical industries. It is not clear how you rebuild this trust.

7. If it does come down to a transatlantic stand-off for model access, the upside of combining forces in the European Union may be lower than the downside of increased coordination costs. There are many members of the EU with conflicted interests who might yield to pressure: Eastern Europe needs the American military presence in Europe to deter Russia. Having countries like these involved in decisions may diminish total European leverage.

These are some really thought-provoking points about Europe and AI, well worth a read. But Germany will not get back to nuclear power and why whould we? It’s too expensive, takes too long and we have plenty of wind and solar.

(…continue reading.)

How Meter Pricing Is Testing the Economics of AI

A growing number of tech firms have begun to introduce usage-based pricing options for their AI services rather than simply charging a flat subscription fee. As a result, the heaviest users are set to incur additional costs each time they ask an AI chatbot or agent to produce a slide deck, draft an email or debug a complex piece of code.

Leading AI labs have been spending tens if not hundreds of billions of dollars on chips, data centers and talent to develop and run their models. The shift to a pricing model more commonly associated with metered-power or pay-as-you-go phone contracts may lead to more selective use of a technology that consumes huge amounts of electricity and other resources.

This was bound to happen and it make total economic sense.

(…continue reading.)

Read more

Running looks like a solo sport. This week makes the case that it never really was. One million parkrun events. A champion who chose to crew another champion instead of racing. A 32-mile loop around an entire island, run for no medal and no prize. The five stories below have almost nothing to do with personal bests and everything to do with the people who show up: the ones who hand you a gel at mile 62, the tribe that turns a Saturday morning into a reason to drive across town, and the women rewriting whose race counts in the first place.

Read on, and then send this to the friend who keeps telling you they want to start.

Subscribe now

‘A huge spectrum of people coming together’: how parkrun made it to its millionth event

Initially a single timed run, the event now operates at 2,800 locations worldwide, with more than 12 million people registered to take part.

Every Saturday morning, tens of thousands of runners and walkers turn up to jog, push prams or simply complete the course at their own pace.

These events, which are entirely volunteer-run at a local level, have seen bridal showers in matching headdresses, milestone birthday parties, and even engagements at the finish line.

I have never attended one of these runs, but I think the whole concept is marvelous! People run for different reasons and if you like to run with a whole lot of people every Sunday, this is probably your tribe!

(…continue reading.)

The Best Race is the One You Don’t Run

I was admittedly a little nervous about screwing something up, especially considering Devon’s résumé. She’s one of the most accomplished athletes in our sport, and while I knew how it felt to be the one racing, I wasn’t nearly as confident in my ability to support someone else through their race.

Still, I had a few things going for me. I’d spent enough time in the sport to understand what runners need when things get hard, and I’d been fortunate to have some excellent crews of my own over the years (mostly at multi-days). I knew what had worked for me: the right encouragement at the right moment, practical help before it was requested, and the ability to read the difference between a runner who needed snacks, opinions, a kick in the pants, or simply a quiet presence.

The rest, I figured, I could learn as I went.

I think the aspect of crewing is still being overlooked when comparing trail and ultra running with other running events. Here we have one of the top athletes in the field supporting another top athlete, which I find really cool.

(…continue reading.)

Running and Feminism: What Does It Mean to Support Women?

The feminism of trail running—assuming that’s what we can even call it—consists of women using the playbook they inherited from (mostly white) men, but inserting (mostly white) women.

This same ole same ole playbook heavily prioritizes elite athletes. The focus is results, popularity, and profitability—over values. It focuses on the most famous events over championing the events that are best for women. It champions likes and followers and downloads over people and community. It doesn’t criticize or hold anyone accountable (like, say, a race director who features a confederate flag at their very famous race). And it leaves behind women of color—as well as women who don’t fit within the narrow swath of those who are competitive in the competitions it deems most important. Being fast and popular is worth more than—anything else.

It also means that that’s why Entrekin’s win is as much or more about her beating men as it is about an incredible athletic performance—this was a playbook written by men, for men, and men are the default standard bearers. Everything ultimately is compared to them.

I think this is a gradual change that is going on and unfortunately we need patience to witness the necessary change. I wish things would change faster.

(…continue reading.)

Their Idea of a Fun Run? A 32-Mile Loop Around Manhattan.

Running the perimeter is an old tradition: It was a regular route for Ted Corbitt, the Olympic marathoner and founding president of New York Road Runners. Sometimes, when he was training for an ultramarathon, he’d run the perimeter twice in a row, said his son, Gary Corbitt, who is 75 and lives in Jacksonville, Fla.

But in recent years, the perimeter run has become more popular, particularly among runners in the city seeking a fresh — and free — challenge. According to data collected by Strava, the fitness tracking app, the number of runs uploaded to the app with a map matching the perimeter of Manhattan increased in 2025 by 30 percent compared with 2024.

To me this sounds like the ultimate sightseeing run. While I don’t like running in the city as much as running through the woods, I enjoy exploring a city by taking runs into areas I haven’t been to. That’ what I do on my business trips, but I probably won’t do 32 miles casually in the morning before meetings…

(…continue reading.)

Running Cool at Western States

The canyons at Western States don’t care how fit you are or how much time you spend in the sauna. By the time you hit Foresthill at mile 62, your core temperature has been climbing for hours, the air is still hot and the trail offers no shade.

Heat acclimation prepares your physiology for heat stress. Cooling strategies manage it in real time. Both matter of these matter, and they’re not interchangeable. This piece is about the third part of that equation.

While I continue to think I could run 100 miles with enough training, I am not so sure about getting through the sun and the heat. I’d be in dehydrated lobster-looking mode real quick and it amazes me how runner battle the sun and the heat while running fast on a really long distance up and down the mountains.

(…continue reading.)

If you missed last week’s edition, you can read it here:

Now, go running!

— Nico

🏃🏻‍♂️

The World Cup has started. I am trying to get into the right mood for it, but I probably won’t stay up late to watch games at midnight or very early in the morning. I did buy two flags and put them on my car - one of the rare occasions when ordinary Germans display the national flag.

Obviously, I hope that Gary Lineker is right again: “Soccer is a simple game: twenty-two men chase a ball for 90 minutes and at the end, the Germans always win.” - and if we don’t, it will be akin to a national crisis, further deteriorating the collective mood our our nation.

Subscribe now

The U.S. Is the Future of Soccer, for Better or Worse

It is a moment to take stock of the altered landscape. For decades, soccer posed a riddle of American exceptionalism. What would it take for the world’s most powerful country to get on board with humanity’s favorite sport? The change, it turns out, came slowly, and then with shocking speed, transforming the United States not just into a bona fide soccer nation, but a novel and unusually influential one. It’s a story that holds a heady plot twist: For better and — more to the point, perhaps — for worse, Americans may not have joined soccer’s universe so much as soccer has embraced America’s.

I beg to differ and I’ll do it with this German sayings:
Geld alleine schießt keine Tore - money alone doesn’t score goals.

And while we are at it, I think it is hard to replicate a century of football tradition and folklore in just a few decades, this has to grow. And the best quotes have are already been coined by people like the legendary German National Coach Sepp Herberger, who famously said:
“Das Spiel dauert 90 Minuten.” - “The game lasts 90 minutes.”
“Nach dem Spiel ist vor dem Spiel.” - “After the game is before the game.”
“Der Ball ist rund.” - “The ball is round.”
“Das nächste Spiel ist immer das schwerste.” - “The next game is always the toughest.”
“Die Leute gehen ins Stadion, weil sie nicht wissen, wie es ausgeht.” - “The people are going into the stadium, because they don’t know how it is going to end.”

Try to beat that! Football philosophy is something Germans have excelled at for years!

(…continue reading.)

Amazon CEO’s Talks With U.S. Officials Triggered Crackdown on Anthropic Models

The government’s talks with Amazon—a big investor in Anthropic that supplies the AI company with chips for data centers, while deploying its best models to identify software vulnerabilities—are a sign of how America’s largest companies and governments are navigating the emerging technology’s novel capabilities. Friday’s rapid events show how quickly new discoveries and experimentation can affect government restrictions and, potentially, company fortunes.

The German saying “Eine Krähe hackt der anderen kein Auge aus” - “One crow doesn’t peck another’s eye” clearly does not apply here. Also, from a European perspective it is clear that we need to increase our efforts, otherwise we could be left with less powerful LLMs.

(…continue reading.)

At Delaney Hall, a V-Neck Sweater Can Change a Life

I first visited Delaney Hall with my husband and two daughters in August 2025. A parent from my local Facebook page was looking for someone to pick up and deliver pantry bags there. This was the perfect opportunity for me to visit a site I’d been eager to protest since its reopening in May 2025 but had been too intimidated to visit: What if I’m just in the way? What if there are a lot of cops? What if ICE is there? If Newark mayor Ras Baraka could be arrested, it must mean anyone can be arrested. It took me only a couple of visits to Delaney Hall to realize I needed to expand my definition of activism. In my mind, I had equated the term to loud protesters, megaphones, signs, gas masks, and of course phones for documentation purposes. This type of protest is necessary for political change; it brings visibility to a cause. In most cases, though, it doesn’t provide logistical steps to realize meaningful movement.

The Trump regime just disgusts me.

(…continue reading.)

Donald Trump’s Pallets of Cash

The joint U.S.-Israel campaign destroyed physical things that can’t be easily replaced, like nuclear facilities. Entering negotiations with Iran to secure physical things that yet remain—the remainder of the regime’s nuclear facilities and its stockpiled enriched uranium—is an acknowledgment that you are, at least at present, not capable of or willing to destroy or seize them. Shifting from war to talks tilts the balance of power in the other direction, away from the U.S., which at war made no concessions, and toward Iran, to which Trump must now make concessions to make a deal.

While I still think that we have to make sure that the Mullahs do not get nuclear weapons, I really do not understand how the USA and Israel could have missed the importance of the Strait of Hormuz. This gave Iran leverage and it was totally unnecessary.

(…continue reading.)

The Strange Defeat of Nuclear Deterrence

Nuclear weapons can appear impotent in the face of sustained conventional and hybrid attacks; in today’s warfare, an arsenal of nuclear-tipped intercontinental ballistic missiles, a flotilla of nuclear submarines, and squadrons of strategic bombers can do little to deter salvos of cheap drones—as long as those nuclear states remain unwilling to use their weapons. That should give pause to both existing nuclear powers and those that may want to acquire the weapons.

The Ukrainians have been showing us for years how to deal with the lingering threat of nuclear weapons. While the Europeans were still debating how to help without poking the Russian bear, Ukraine just fought back and continues to do what it can to show that Russia will never win this war.

(…continue reading.)

If you haven’t already, please drop everything and subscribe to Five Things now!

That’s it. Have a great Sunday! If you missed last Sunday’s edition of Five Things, have a look here:

Subscribe now

— Nico

All the Ways Europe Is Ditching American Technology

The moves are widespread—and growing. Last week, the European Commission launched its official long-term plans to rely less on US technology. The European Parliament has switched the default search engine on its devices from Google to the French alternative Qwant. Thousands of workers in the French government are using its own open-source office software —dubbed LaSuite—as officials aim to “break free” from dependence on American tech firms. An open-source documents offering from more than a dozen European tech companies, called Euro-Office, is due to launch imminently. Cities across the Netherlands, France, and Germany are all moving away from Microsoft Office and Google Docs

It’s not just productivity software, either. The Dutch government is moving its code away from Microsoft-owned Github to its own repository. In a series of decisions, Finland reportedly decided not to move its election data to Amazon’s cloud services, while the organization behind Belgium’s.be top-level domain has said it will move away from AWS. Meanwhile, Eurosky has been spun up as an interoperable alternative to Bluesky on the AT Protocol that underlies both social networks.

Don’t get me wrong, there a lots of really great US tech products. But we need to find our own path in Europe and reduce the dependency on US tech.

(…continue reading.)

AI can design and run thousands of lab experiments without human hands. Humanity isn’t ready for the new risks this brings to biology

Hands-on work in the lab has traditionally been a bottleneck to translating designs into results. Even a brilliant study plan still depends on skilled human hands to carry out. That may not last, as cloud laboratories and robotic automation become cheaper and more accessible, allowing researchers to send AI-generated experimental designs to remote facilities for execution.

AI systems are now able to run experiments autonomously and at scale, but existing regulations were not designed for this. Rules governing biological research do not account for AI-driven automation, and rules governing AI do not specifically address its use in biology.

We need so many new ways to define guardrails, otherwise we cannot keep up with AI.

(…continue reading.)

The Real Cost Of Cooling GPUs In Space Might Shock You

Proponents tout the many wonders of computing in space: abundant solar energy, free cooling, and freedom from Earth-based disturbances like earthquakes, floods, and protesters. But a sober look at the physics of space-based computing paints a much more nuanced picture.

Free cooling is perhaps the biggest misconception. Space is cold, but it also has no atmosphere. That means the best heat-removal mechanisms, conduction and convection, are off the table. The only option is radiation. To prevent a chip from overheating in space, a large, costly surface area is required to dissipate the energy and then radiate it.

Solar energy is abundant, but collecting it with functional solar panels that maintain perfect alignment toward the sun is a complex task requiring extensive attitude control systems. On top of that, ionizing radiation in space from cosmic rays and other sources poses a unique challenge, degrading the solar panels, the radiative coolers, and the chips themselves. Because regular maintenance in space is difficult, redundancy has to be built in at launch, and cost estimates have to account for efficiency degradation over time.

I really don’t see why people remain so optimistic about datacenters in space. So much hype, just for one IPO…

(…continue reading.)

Pokémon Go data trained AI that could assist military drones in war zones

Niantic Spatial – a spin-off company from Niantic – announced its partnership with Vantor, a company that specialises in spatial detection software for drones, including those used by some militaries, in December.

The agreement is designed to allow drones to navigate and coordinate precisely in areas where GPS is not available.

Pokemon generated so much data and this is now being used to train drones that could get deployed across our cities. What a time to be alive.

(…continue reading.)

Making a vintage LLM from scratch

In short, to build an LLM you need 4 things:

• the data -- an LLM has no discernment or understanding. It will learn from anything you tell it to, good or bad. This is the longest process.

• tokenization -- the Tokenizer is a little program that converts words or letters into numbers (tokens). LLMs don’t understand words, they only understand numbers.

• pre-training -- it’s a confusing expression and it means “base-training”, where the LLM learns to autocomplete text. If you’re going for a 300m+ params, this is the most expensive process.

• fine-tuning -- where the LLM learns how to chat in turns, question & answer.

This is so cool. Not really useful, but I guess if you really want to understand how an LLM works, you have to build it yourself.

(…continue reading.)

That’s all for now! Thanks for reading! If you missed last week’s Five Things Tech, you can find it here:

🤖

— Nico

Subscribe now

This week is a reality check. The agents write the code now, but the hard part just moved downstream to the humans who have to review it, hold the context, and catch the mistakes nobody saw coming. Meanwhile the studies keep landing: lots of output, plenty of slop, and surprisingly little ROI or GDP to show for it. Cohere drops a 30B coding agent that runs on a single H100, which is exactly the kind of thing that finally lets plenty of rather hesitant companies do agentic coding without sending its code to the cloud. Meta tries to sell subscriptions to a corporate market that has never wanted to buy from it. And Ramp reveals that the median firm spends 11 USD a month on AI while a tiny top tier spends 7,449 USD per employee. Five stories, one question underneath all of them: who is actually getting value here, and who is just spending?

Subscribe now

Agentic AI solved coding — and exposed every other problem in software engineering

As AI-generated code scales, human review is becoming a massive new bottleneck, and engineers are losing the context needed to catch agent mistakes. The companies that understand this will move forward deliberately and even create new roles because of AI. The ones that don’t will default to a simpler, far more destructive conclusion: Reduce headcount and increase AI spend.

I don’t think human reviews are the answer. Agentic peer-reviews will be the norm pretty soon.

(…continue reading.)

Slop, productivity, and why the AI-fueled world is going nowhere mighty fast

AI generates a lot of output (which fits many people’s informal notion of productivity), but it hasn’t yielded much in the way of RoI for many companies (per studies from MIT, McKinsey, Bain, and many others), and hasn’t materially changed GDP.

Actually, while I do see the issue of AI slop in a lot of places, including software development, I do see a trend for Personal Software on Demand. I can just build a tool for myself or my company and I do not expect any more than just a handful of people to use it. That is a true paradigm change.

(…continue reading.)

Cohere Open-Sources North Mini Code: A 30B Coding Agent That Runs on a Single H100

Cohere’s North Mini Code landed on June 9, 2026, with a straightforward pitch to developers: a 30 billion parameter agentic coding agent, Apache 2.0 licensed, that fits on a single NVIDIA H100. Independent testing by Artificial Analysis ranks it 8th of 127 comparable open-weight models on output speed at 210 tokens per second, with a time to first token of 0.25 seconds against a class median of 1.95 seconds. The catch — and it’s a real one — is verbosity: the model generated 75 million output tokens in Artificial Analysis benchmarking against a class median of 25 million. That’s three times the output of comparable models. For developers deciding whether to route agentic coding workloads through a managed API or a self-hosted open-source model, this is the clearest open-source contender to emerge in 2026.

I had a conversation a few weeks ago with the father of one of my youngest daughter’s friend and he told me that while he is really interested in agentic coding, he cannot do it in the office. He works at a large insurance company and code getting created with the help of an LLM outside the company is just out of the question. I am really interested how the development of the agentic coding LLMs will advance in the next few months. Public institutions will take note, hopefully.

(…continue reading.)

Meta’s Subscription Push Exposes Its Weak Hand in AI

Meta’s bid for corporate AI customers is even more far-fetched, with competitors including Google and Microsoft entrenched. Meta says it already has a million companies using its corporate AI agent, for which it eventually plans to charge a subscription fee.

But companies have never been core customers for Meta outside ad sales, and its commitment to them has wavered over the years. It discontinued VR headsets geared toward companies earlier this year as its priorities shifted.

Meta has used AI more effectively than pretty much anyone else to improve the performance of ads, and its revenue is expected to grow quickly because of that. Analysts expect around $60 billion of revenue in the second quarter, according to FactSet, a 27% increase from the prior-year period.

Here’s a thought: how about Meta would use AI to improve the user experience, finally moderate all those spam accounts everywhere and filter out all the AI slop?

(…continue reading.)

How much does it cost to be AI-pilled?

The top 1% of firms spend $7.45K per employee per month. The top 10% spend $611 per employee per month. The median firm spends just $11.38 – about the cost of a seat on an enterprise ChatGPT or Claude subscription.

And while several high-profile proclamations have said you should be spending as much on AI as you do on a software engineer’s salary…no one is actually doing that. For the top 1% of spenders, per employee AI spend is still less than half the typical monthly salary for an engineer. Not to say that won’t change – AI spend is still rising – but very few firms, if any, are actually doing that.

That is insane. I am spending a lot less, probably around 250€ and that is still more than plenty…

(…continue reading.)

Read more

There is a strange new category of travel where suffering is the souvenir. Ultrarunning has a name for it, the darecation, and the growth is wild. UTMB went from 50,000 participants in 2022 to 170,000, most of them not elites, but amateurs who turn themselves into engineers of their own bodies just to spend 30 hours moving through the mountains - and bring their support crews along. This week pulls at that contradiction from a few angles: the months of prep that quietly become a darecation of their own, the uncomfortable fact that the self-styled environmental sport actually runs on air travel and shoes built to die in a few hundred miles, who is really betting money on trail running and who is just dabbling, and a reminder from Frozen Head State Park that even the cruelest race on earth can somehow turn out to be just lovely. You just have to train a lot for it, I guess…

Subscribe now

Inside the ‘darecation’: Why travellers are choosing holidays that hurt

At UTMB World Series, one of the world’s leading organisers of trail running and mountain races, business is booming. Since launching with a single race around the Mont Blanc Massif in 2003, it expanded to 25 ultramarathon events in 2022 and now hosts more than 60 races globally. Annual participation has risen from 50,000 in 2022 to 170,000, most of whom aren’t elite athletes but highly committed amateurs.

“People become engineers of themselves for these races,” said Florian Lamblin, executive director of UTMB International and an ultrarunner himself. “They are trying to achieve something complex and, at the end, deliver something extraordinary, which is running up to 30 hours in nature.”

While I envy people who can take those vacations, I also ask myself: “don’t they have kids? or partners who want sightseeing or the beach? how do they do this? do they do a vacation afterwards to get some relaxation as well?”

(…continue reading.)

Becoming BFFs with the WSER course

We certainly trained; I ran 167 miles in seven days and felt better as camp progressed (since I’m Strava dark, this is your teaser for how training is going:P). We pushed hard and asked a lot of our bodies. But we also fell over laughing, put on baseball gloves and played catch, and met the guy who runs the Greengate aid station (who is also a butterfly dealer?!). Critically, the hard work was done in a context of fun, friends, and discovery. But an easy-to-miss part of training was strengthening our relationship with this 100-mile thread of trail through the California hills. I feel like I befriended the course, learning more of its quirks and paying it respect that is impossible to do while racing.

Wow, that preparation sounds like a darecation of its own, but it is just the buildup before the race.

(…continue reading.)

Trail Running Isn’t an Environmental Sport

That’s the argument in a single frame. Trail and ultra running thinks of itself as the environmental sport, and its community as nature’s stewards. It keeps Edward Abbey on the nightstand. It talks about stewardship and leaving no trace, about the privilege of crossing wild country on its own two feet. Then it runs on air travel to far-off start lines, on shoes built to die inside a few hundred miles [1], on races that bolt new distances and whole continents onto the calendar every year. The self-image is preservation. The machine underneath, however, is unbridled consumption. For two decades the sport has been building a sincere, expensive, well-staffed apparatus whose real job is to keep anyone from noticing the difference.

True, and everyone is aware of this, but doesn’t really want to change it. Just look at all the new products trail runners are supposed to buy to make the runs better.

(…continue reading.)

Who’s Actually Betting on Trail Running?

The framework reveals a bet that’s easy to miss. Concentrated brands, those that are fully run-focused or that separately report running/outdoor segments, can’t pull back on trail investments without it showing up in consolidated results. Diversified parents like Nike and adidas can dial trail spending up or down without materially affecting their financials.

The high-conviction brands are the canaries in the coal mine. They signal whether trail running continues to grow and remain a profitable category. Watch closely if Nike starts disclosing more on ACG, or breaks it out as a separate segment. That would be a clear sign the relaunch is yielding real financial results.

I am still rooting for the niche players and I love it that there are so many new brands building on this sport.

(…continue reading.)

Why the notorious Barkley Marathons is ‘nicer than you think’, according to a man who’s run it three times

It’s the world’s toughest foot race. The statistics speak for themselves: 20 finishers in 39 years. The leg-slashing sawbriars. The Blair Witch woods. The freezing fog and mocking lack of assistance (‘No help is coming’). But after my third visit, I need to tell you all that Frozen Head State Park, Tennessee, home of the Barkley Marathons, is simply one of the loveliest and most welcoming places I’ve ever been.

So, it’s just a fun darecation after all?

(…continue reading)

If you missed last week’s edition, you can read it here:

Now, go running!

— Nico

🏃🏻‍♂️

So this weekend something really weird happened. We only had one kid at home. Our son (k2) went on vacation, our younger daughter (k3) went to Berlin to visit our oldest daughter (k1), while our youngest daughter (k4) stayed at home. The house feels so empty. I miss them all dearly, but it is also really nice to have just one kid to focus on. Especially when she rather wanted to do her thing and meet a friend.
We also bought way too many groceries, because normally our fridge has been emptied right after filled it…

Grab a coffee and read these Five Things before you do what you do on a Sunday!

Subscribe now

Germany Has Lost What It Did Best

A year into Chancellor Friedrich Merz’s tenure, his approval ratings are low and the coalition government he leads is roundly disliked. Capitalizing on widespread discontent, the far-right Alternative for Germany is rising in the polls, inching closer to power. To make matters worse, America, in a Trumpian tantrum, has announced a withdrawal of troops from the country, endangering its bedrock geopolitical relationship. All the features that defined Germany’s recent past — political stability, social cohesion and Atlanticist foreign policy — are under threat.

Yet underlying all these problems is a more profound one. The German economy, once known for its efficiency, orderliness and stability, is in a terrible mess. It’s not just that the numbers are dire, though the country has basically been in a recession for three years. Or that every week or so, another famous old company announces thousands of layoffs.

Actually, I think Germany is too well off to realize it is in decline. We prefer to elect leaders who promise no change and that works until everyone is fed up and works for the other guys, the then actually do something, until everyone complains that they are doing too much, and then the other guys get into office again on the promise of not doing anything. And now the populists from the right and the left offer easy solutions, which also won’t help. Oh, and the business leaders are looking for their paycheque, nothing else. They all fumbled on innovation and digital transformation and now they look like they are stuck in the late 90s.
But still, we are the third biggest economy in the world, because Japan is not doing so great, either.

(…continue reading.)

Online Trolls Harassed Her Six-Year-Old. That Was Only the Beginning

In a shitposting world, Draper isn’t afraid to shitpost in response. He’s defended criticism of his business, saying people just need to “have a sense of humor” about it. And not all of his merchandise is provocative. Raygun had also made that “Smile like Audi” shirt — the one Kamdyn was sporting in her photo with Crooks.

So, it makes sense that when Emily was spiraling about online bullies, she would reach out to a brand that she saw as fighting for her side. She sent the shop a DM, asking for help. Mike Draper replied personally to set up a call.

What came next was a plan.

First of all, internet trolls are assholes. Secondly, Iowa is a wonderful state with friendly people and Raygun is just Raygun. What a story.

(…continue reading.)

Social Media Is Now Parasocial Media

In 2026, many major social media platforms feel icky because we are in the full throes of the third stage of enshittification. Today’s social media platforms are no longer centered around sociable activities. Instead, most platforms offer us a broadcast medium and invite us to learn how to game the algorithms so that we too can create assets for the major corporations (Cotter, 2019). Since scale is valorized in this platform economy, we are encouraged to curate ourselves in pursuit of fame and attention. We can still, in theory, create content for our 15 friends, but it’s not clear that they will see what we post. To actually be seen, we must work it.

Of course, for many people, it’s not clear whether working it for the algorithm is worth it. For many people, the benefits of joking around with friends on social media doesn’t feel worth the potential privacy risks, reputational risks, and social risks. Scrolling is easier. Sending funny videos to friends via text message feels safer than reposting.

Great paper, spot on. I am spending less and less time on social media and this explains why. I do miss the old Twitter, that was good fun while it lasted. And blogs, blogs where wonderful. Now we have Substack, which is okay, but feels different.

(…continue reading.)

Made in the USA

Extreme though Hegseth may be, he is a recognizable type: a jockish, puerile white man, a boy you knew in your public high school, if you went to one. He is the Jersey Shore as much as he is Kansas, Florida, Texas, and Oregon. You may recall him as the guy who shoved queer kids into trash cans in the cafeteria and said things about girls like “You’d need a crowbar to get her legs open.” As an adult, Hegseth is a man whom people have described leaving a bar, shit-faced, chanting “No means yes!” and “Kill all Muslims!” He is what the world thinks some Americans are, the bleakest caricature. But like the violence in the administration’s videos, Hegseth is real, and he is American, which means we have no choice but to ask what to do with him, and what to do with ourselves.

A cabinet full of sociopaths and evil clowns. No dystopian novel could have come up with this storyline and yet it is our reality and it is shaping the world.

(…continue reading.)

‘They surprise me every time’: bees can use tools to solve problems, study finds

The bees were given an adapted version of an experiment that, 100 years ago, first demonstrated chimpanzees could work out how to retrieve an out-of-reach banana by stacking boxes. Since then, various other primates, elephants and crows have joined an elite cohort of species known to be capable of this level of insight and spontaneous problem solving.

In the latest research, bees were shown to be able to roll a polystyrene ball to a specific location and climb on to it in order to access an artificial flower on a low ceiling. The findings challenge the longstanding assumption that insects operate purely on instinct and mindless trial-and-error learning.

Bees are amazing. And so is honey. What a combo.

(…continue reading.)

That’s it. Have a great Sunday! If you missed last Sunday’s edition of Five Things, have a look here:

Subscribe now

— Nico

That gap is the most interesting story in tech right now, and it runs through everything below: Europe waking up to its tech dependency, data centers quietly becoming the most powerful buyers in hard tech, startups stuffing servers into homes to spare the grid, and a billionaire whose track record on big promises is worth counting. The hype is loud. The infrastructure is where the real fight is.

Grab a coffee and read this now! And hug a data center near you!

EU plots long game against US digital supremacy

Instead, the tech sovereignty package — motivated in no small part by U.S. President Donald Trump’s weaponization of Europe’s dependence on American firms — takes a longer-term view: boost the continent’s players so they can eventually challenge their U.S. rivals.

“We cannot afford to depend on others for the technologies that keep our hospitals running, our energy grids stable and our services secure,” Commission President Ursula von der Leyen said in a statement. “This is about protecting our citizens, defending our interests and making our own choices.”

Ever since I started using Linux some 30 years ago, I had this eerie feeling that our dependency on US tech is not in the best interest of Europe. Now finally politicians have come to understand just how import tech is for our everyday lives, our economy and our society. And just think about the gazillions of dollars the EU spends on crappy products from Microsoft that just lead to frustration…

(…continue reading.)

Thank God For Data Centers

This is true for obvious things like GPUs, inference chips, and DRAM, but it’s also true for companies that you wouldn’t typically associate with AI data centers, like supersonic turbines, enhanced geothermal, modular construction, high-voltage direct current grids, solid-state transformers, silicon photonics, optical fiber, lasers, batteries, and nuclear.

Many of these technologies have the potential to be better and cheaper than the incumbent technologies they aim to replace, but they have been too expensive and unproven to compete. With backlogs in all of the traditional inputs to Data Centers, however, developers are willing to pay up for new technologies that can deliver fast, which gives them the opportunity to scale up and cost down.

For these technologies, Data Centers act as a third type of Buyer of Capabilities, a commercial analog operating on DoD-style procurement logic but commercial timescales.

Given the size of the budgets, the relative smallness of any one input’s cost relative to the overall project cost and revenue opportunity, and the speed with which Data Centers are making decisions and putting down deposits, Data Centers may meaningfully increase the odds of success of hard tech companies and Vertical Integrators more than the market realizes.

I have to agree here, even though I do not want data centers taking over the landscape everywhere. They are huge and operating them poses many interesting new challenges. Solving these challenges will have plenty of amazing side-effects. Cheaper energy will be one of them.

(…continue reading.)

Startups are installing tiny data centers in people’s homes to reduce strain on the beleaguered electrical grid

Span’s XFRA models are part of a wider distributed network of AI infrastructure, using a home’s underused electrical capacity to create something similar to a cloud of compute that can be given to service providers. Span can install nodes at six-times the speed of centralized 100-megawatt data centers and at about one-fifth of the cost of construction.

The company charges a flat monthly fee of about $150. In return, it essentially pays a host’s electricity and internet bills. The computing power generated from the nodes are distributed to customers like hyperscalers and AI companies. XFRA is not designed to replace commercial data centers, according to the company, but rather to reduce strain on the grid.

And if those small datacenters all have some networking and proxy elements to them, they could also reduce the bandwith needed on the data networks.

(…continue reading.)

There’s Never Been a Better Time to Study Computer Science

At the same time, new courses could offer students an introduction to software development without the theoretical baggage and proof-writing they might have otherwise had to wade through. Geoffrey Challen, a computer scientist at the University of Illinois at Urbana-Champaign, plans to offer a new course this fall in which he will teach students to develop software “without writing, reading, debugging, or viewing a single line of code,” he told me. Northwestern is also slated to offer an “entry-level creative coding” class for students without technical backgrounds. For all the talk of AI-literacy programs that teach students how to use chatbots, the real innovation might be in developing courses that train students in basic software-development skills. Most colleges require introductory writing courses because it’s understood that clear written communication is an important cross-disciplinary skill—even for students who plan to study physics or math. Classes that teach students how to use AI coding tools could become commonplace, providing students of all backgrounds with a baseline software-engineering skill set.

I think this is an amazing time to study computer science and of course it helps to understand how software works even if you will never have to write it yourself.

(…continue reading.)

Elon Musk Laid Out 602 Goals. We Counted How Many He Hit.

Strikingly, Mr. Musk’s annual rate of success declined over time, even as he made more promises. Of the 13 goals he declared in 2015, he later achieved nearly three-quarters of them, The Times found. But of the 27 claims he made in 2020, fewer than half have been accomplished on time. Some still have deadlines far in the future.

I think we should regulate and tax billionaires more. This is not healthy. Neither for them nor for our societies.

(…continue reading.)

That’s all for now! Thanks for reading! If you missed last week’s Five Things Tech, you can find it here:

🤖

— Nico

Subscribe now

Five stories on AI again this week. One thread: the tools are working. Small-business owners are building armies of AI workers, doing in days what used to take a full team. Pre-ChatGPT startups are getting eaten alive, weighed down by staffing models that made sense three years ago. Productivity is up across the board. Wages are not. The valuations of AI companies are hitting the stratosphere while the social safety net is getting cut at the same time. Developers are not complaining - they accept the shift from typing to thinking. Javier Milei is writing AI law and Peter Thiel is excited about it, which should tell you something. Every tool promises more, and most deliver more or less the same thing with a different coat of paint. The gains are real. The distribution of those gains is the open question nobody in a keynote wants to touch. The question nobody wants to answer is - working for whom?

Subscribe now

The Small-Business Owners Managing Whole Armies of A.I. Employees

These days, most people have realized that A.I. is creeping into the office. A recent Gallup poll found that 28 percent of the respondents use it anywhere from several times a week to daily. But these are mostly just the familiar chat sessions. The partisans of OpenClaw are doing something that is, depending on how you look at it, either more wildly ambitious or foolishly risky — or, as many will readily confess, both. They’re creating virtual employees, artificial staff that scurry around on their hard drives, doing work on their own say-so.

AI will help so many small companies to do so much more than was ever possible for them.

(…continue reading.)

As AI gets better, it reveals an empty promise

There’s a sinister tone lurking beneath some of these advancements in productivity, because the response to increased productivity has been one of the biggest scams of the past century. Well before consumer AI entered the scene, productivity exploded while wages failed to keep pace. Nobody is working less, they’re just earning less. And as more AI-related companies reap trillions in valuation, the current US regime is looting the social safety net — the kind that must exist if we’re all going to become out-of-work theater kids. You simply can’t look at these things separately. If the end result of private companies optimizing the workforce means nobody has to work, then we have to live in a society where people can still have a roof and a meal. Is anyone confident that will happen while leaders are cutting SNAP benefits while building taxpayer-funded ballrooms?

The author makes a strong point here: there are so many new productivity tools out there that promise so much. But if you actually play around with those tools, you find out that they all do more or less the same, just with different flavors.

(…continue reading.)

I Built an AI Company. Here’s Why AI Won’t Kill Coding Jobs.

This pattern is familiar. We replaced shovels with excavators, and then we built skyscrapers. We replaced manual arithmetic with calculators, and accountants did not vanish; they did more interesting work, if you can believe it. Each time a tool made a job easier, we didn’t run out of things to do. We found bigger things to do.

AI coding tools represent the same kind of leap. You still have to tell the machine what to build. You still have to understand what it’s doing well enough to check its work. And these coding tools still make mistakes when generating code, so you need an experienced software engineer to fix them. The job shifts from typing to thinking, which for most good engineers is the part we love anyway.

There is a huge change going on that transforms a lot of jobs, that’s for sure. The developers I talk to aren’t complaining, they accept the challenge and reinvent the way they work.

(…continue reading.)

Javier Milei: Argentina invites AI to free itself

At the beginning of the industrial revolution, Adam Smith illustrated the potential of technology and economies of scale in his celebrated recollection of the pin factory. And, as much as the industrial revolution freed us from the constraints of the human muscle, AI will free us from the constraints of the human brain, pushing productivity beyond our wildest dreams. It is for this reason that my government last week submitted legislation to Congress establishing a dedicated legal framework for the deployment of AI.

This is the libertarians’ wet dream. Peter Thiel is excited and moving down there. This is a terrible idea.

(…continue reading.)

‘Disrupted or dead’: AI is crushing a generation of startups built before ChatGPT

Companies built before generative AI are weighed down by bloated staffing models and software designed for a pre-AI world, according to Zhu, making it hard for them to transform themselves.

“Unless they make a stark, 180-degree pivot to rebuild the exact same thing from scratch, they’re going to slowly fail,” Zhu said. “What that means is that investors would rather just bet on new entrepreneurs at lower valuations rather than double down on older startups.”

The agents are eating the children of the digital revolution.

(…continue reading.)

Read more