Zum Hauptinhalt springen

Steady funktioniert nicht richtig, wenn JavaScript ausgeschaltet ist. Bitte schalte JavaScript ein und lade die Seite neu. Hier findest du eine Anleitung wie du JavaScript in deinem Browser einschaltest (Öffnet in neuem Fenster).

Posts

When every agent is correct, and the system is not

Picture a repository on an ordinary Tuesday. Three coding agents are at work inside it, each competent, each pointed at a task that, taken on its own, it handles well.…

30.06.2026

When engineers can direct AI, and when they cannot

Engineers shape how generative AI enters their work, but only when their workflow, their autonomy, and their governance leave them room to do so. Last week I gave the…

23.06.2026

The capability you cannot fork: software's break with universal…

For four decades, an engineer could obtain and run the core tools of the field on the same terms as anyone else, whatever their nationality, employer, or location.…

16.06.2026

Europe is about to make human-first software non-optional

In May 2026, Pope Leo XIV released his first encyclical, "Magnifica humanitas," a text devoted to safeguarding the human person in the time of artificial intelligence…

09.06.2026

Ericsson's AI-native software programme: principles, deployment, and…

In May 2026, four Ericsson engineers (Britto, Palmgren, Saini, and Ohlin) released a paper titled The AI-Native Large-Scale Agile Software Development Manifesto. The…

02.06.2026

Disarm AI, Re-Arm Teams: Two Levers Against a Lost Cohort

The cohort graduating into software work in 2026 enters a labour market in which the entry-level rung has been quietly compressed by AI, and the empirical record on what…

26.05.2026

Three years of AI: what the evidence settles, contests, and leaves…

After three years of GenAI adoption in software engineering, the empirical literature has stabilised enough to do something the field could not do a year ago: separate…

19.05.2026

A Bad Benchmark Does More Damage Than No Benchmark at All

The central failure of AI evaluation today is not that organizations test too little. It is that they test with instruments they have never validated. Every engineer who…

12.05.2026

Cleaner Code, Foggier Teams: The Debt AI Moves Upstream

Generative AI is reducing one form of software debt while quietly accumulating two others, and most engineering organisations are still measuring only the one that is…

05.05.2026

The Composition Trap: Why Agent-Dense Teams Break in Ways Human Teams…

Multi-agent AI systems fail in ways that individual agent evaluations will never surface, and the software engineering field has not yet built the vocabulary to describe…

28.04.2026

Your Agents Pass Every Test. Your System Can Still Fail.

Organisations building multi-agent AI pipelines face a measurement problem that standard benchmarking cannot resolve: individual agents can pass every capability test in…

31.03.2026

Cutting the Junior Pipeline Before the Automation Arrives

The numbers from Harvard are unambiguous: junior employment at GenAI-adopting firms fell roughly 9 percent relative to comparable non-adopters in less than two years…

24.03.2026

AI Can Solve Your Benchmark. Can It Fix Your Codebase?

The most informative thing about AI’s current capabilities in software engineering is not what it can do on a benchmark. It’s who the benchmark is comparing it against.…

17.03.2026

56% of Organizations Get Nothing from AI Because Compatibility Was an…

The consulting industry has spent two years producing reports on why AI investments fail. The data is consistent and substantial. What the reports lack is a theoretical…

10.03.2026

The New Engineering Judgment: Five Skills That Survive the Agentic…

A software researcher at University College London built a production-grade, multi-module data analysis platform (118 Python files, 31 REST API endpoints, 200+ unit…

03.03.2026

When Code Becomes the Output: Rethinking What We Teach (and Hire For)

A synthesis of industry roadmaps and complexity science research suggests software engineering education is teaching the right skills—for the wrong paradigm. Every few…

24.02.2026

Speed at the Cost of Quality: The Hidden Tax of Agentic AI

The promise of agentic AI was straightforward: it would write the code, leaving us to architect the systems. The reality, as revealed by empirical data from late 2025…

30.01.2026

The "Country of Geniuses" Requires a Constitution, Not Just a Login

The 'country of geniuses' arrives 2027: 50M agents, Nobel-level IQ, 100x speed. They'll game your tests. You need a constitution, not just a login.

30.01.2026

Davos 2026: From “Chat” to “Work”

Davos 2026: 87% adopted AI. Only 31% see ROI. The problem isn't the model—it's the fit.

30.01.2026

Europe’s Tech Renaissance Beyond the “Kill Switch” Narrative

After my last week's article, A Compass for Europe’s Software Future, my inbox filled up with variations of the same question: “Do we have a service kill switch?” Not a…

30.01.2026

Technological Sovereignty in a New Era: A Compass for Europe’s…

Three years ago, most “cloud strategy” discussions in Europe sounded the same: Where are the best engineers? Which vendor gives the biggest discount? Can we move fast…

30.01.2026