@anselmkuesters.bsky.social
Head of Department Digitalization and New Technologies, Centre for European Policy (cep), Berlin | Postdoctoral Researcher, Humboldt University of Berlin | Associated Researcher, Max Planck Institute for Legal History and Legal Theory, Frankfurt am Main
The Small Signal #11: edge AI that processes data locally but reports to the police. A digital euro that claims sovereignty but may crowd out its alternatives. And Heckscher's 1939 network theory, translated in 2026 by the thing it warned about.
Link: anselmkuesters.substack.com/p/the-small-...
The Small Signal #10
➡️Doctorow: "There is no alternative" in tech is politics, not engineering
➡️Bertelsmann on interoperability; Linux Foundation's Agentic AI Foundation
➡️Japan's smartphone act is live; the DMA uses fairness without defining it
Link: anselmkuesters.substack.com/p/the-small-...
Podcast-Hinweis: Digitale Souveränität
In der aktuellen Folge von „Das Ohr am Netz“, dem Podcast von eco – Verband der Internetwirtschaft, spreche ich über mein neues Buch „Small is Beautiful 2.0“ und die Frage, wie Europa technologisch handlungsfähig werden kann.
open.spotify.com/show/5rlnSMJ...
Danke, Daniel! Genau darum geht es: nicht Technologie um ihrer selbst willen, sondern Werkzeuge, die lokale Akteure stärken. Freue mich, wenn das Buch nützlich ist.
The Small Signal #9:
Oxford researchers say mid-sized economies must pool resources or accept dependency. I think the compute gap matters less than they claim, and the political will for a CERN for AI is not there yet.
Plus 2 additional signals!
Link:
anselmkuesters.substack.com/p/the-small-...
Wie kann (digitale) Wirtschaft nachhaltiger, gerechter und resilienter werden?
Indem wir mit #OpenSource, kleinen KI-Modellen und dezentralen Ansätzen lokale Wirtschaft, KMU & Gemeinschaften stärken
Mehr dazu im großartigen neuen Buch von @anselmkuesters.bsky.social
www.herder.de/geschichte-p...
My newsletter, "The Small Signal" #7, is out:
- Why Mars visions serve incumbent power
- Digital twins as "appropriate technology" for SMEs
- How spatial economics limits platform dominance
Three angles on infrastructure control.
anselmkuesters.substack.com/p/the-small-...
@cepger.bsky.social
The Small Signal #6 is live: Technical AI decentralization advances, but market concentration persists.
This week's 3 signals cover:
> Local AI vs switching costs
> Sovereignty & compliance as service
> Speech & few-sample adaptation
Read the issue: anselmkuesters.substack.com/p/the-small-...
Trump doesn’t speak for the US. At least not if you read current polls on key issues of joint European and US interest. What does this mean for the transatlantic alliance in the wake of Greenland and Davos? Read my take here👇
commongroundeurope.eu/blog/the-tra...
Chinesische Open-Source-KI wird derzeit als pragmatische Option für Europa diskutiert. Anselm Küsters sieht das deutlich kritischer: www.cep.eu/de/eu-themen...
@anselmkuesters.bsky.social
The Small Signal #5: When digital governance ideals meet implementation constraints. This week's issue of my newsletter covers:
🇹🇼 Taiwan's deepfake counterstrike
📚 Coalition S retreats
🇪🇺 EU standardizes VR... before it exists
Read the analysis here: anselmkuesters.substack.com/p/the-small-...
The Small Signal #4 is live! This issue delivers 3 signals on the maturation of decentralized digital infrastructure - and the tensions that come with it.
➡️ Wikipedia's paradox
➡️ AI's "middle field"
➡️ NFT utility beyond speculation
Read the full issue: anselmkuesters.substack.com/p/the-small-...
The Small Signal #3 is live: AI's environmental debate shifts from panic to precision.
Three key signals in this week's edition:
- AI water usage claims corrected
- decentralized training grids now viable,
- an emerging digital ESG divide
Full analysis: anselmkuesters.substack.com/p/the-small-...
Are we building digital democracies or just renting them?
In The Small Signal #2, I look at a divergence: While cities like Buenos Aires are building transparent, human-in-the-loop AI systems, the German government is reinforcing its dependence on US tech giants.
open.substack.com/pub/anselmku...
🎯 First edition of "The Small Signal" is live!
Three signals worth watching:
→ Why tiny recursive models might beat massive LLMs
→ China's open-source AI swarm strategy
→ The emerging vertical data moat advantage
Read on Substack: anselmkuesters.substack.com/p/the-small-...
For more background on the underlying book and the core ideas behind "The Small Signal", check out the welcome post here: open.substack.com/pub/anselmku...
It explains why Schumacher's "Small is Beautiful" framework matters more than ever in the age of AI.
Launching "The Small Signal" – a Substack newsletter on human-scale digital economies.
Each week: 2-3 signals (new research, policies, etc) that tell us something deeper about decentralizing tech power.
Issue #1 this afternoon, then weekly Fridays.
Subscripe: anselmkuesters.substack.com/subscribe
Infos und Link zur Vorbestellung auf der Website des Herder Verlages: www.herder.de/geschichte-p... /7&end
Vor fünfzig Jahren schrieb E.F. Schumacher „Small is beautiful“ und warnte vor der Vergötterung des Gigantischen. Er hatte recht. Aber ihm fehlten die Mittel. Heute haben wir sie. Digitale Dezentralisierung macht möglich, was Schumacher nur ahnen konnte. /6
Während die USA auf maximale Skalierung und China auf staatliche Kontrolle setzen, kann Europa einen dritten Weg gehen: eine digitale Wirtschaft im menschlichen Maßstab. Das ist keine Romantik, sondern Realpolitik. Dezentrale Strukturen sind krisenfester. Sie fördern Wettbewerb und Innovation. /5
Open Source ermöglicht kollektive Innovation und mehr Cybersicherheit. Blockchain kann Mittelsmänner überflüssig machen und ökonomische Renten fairer verteilen. Virtuelle Welten könnten Raum für kleine Akteure schaffen, statt sie zu verdrängen. /4
Die Werkzeuge liegen bereit. Digitale Plattformen können lokale Produzenten mit globalen Märkten verbinden, ohne sie einem Monopolisten auszuliefern. Kleine KI-Modelle laufen auf normalen Laptops und geben mittelständischen Unternehmen Zugang zu Technologie, ohne Abhängigkeit von Cloud-Giganten. /3
Genau darum geht es in meinem Buch, das im Februar bei Herder erscheint und jetzt vorbestellbar ist: „Small is beautiful 2.0 – Mit digitaler Dezentralisierung zu einer menschlicheren Wirtschaft.“ Statt nur zu fragen, was wir verhindern müssen, zeige ich, was wir bauen können. /2
Europa diskutiert gerade intensiv über digitale Souveränität. Das ist wichtig. Aber wir tun es in der Sprache der Abwehr. Wir wollen uns schützen vor Big Tech, vor Abhängigkeiten, vor Kontrollverlust. Was uns fehlt, ist die positive Vision: Was wollen wir eigentlich erschaffen? Wofür stehen wir? /1
What if the Luddites had won? Not with hammers, but with policy.
On Tuesday, I will be using counterfactual historiography at Stuttgart University to address a key issue in the governance of AI: how can computational power be redistributed effectively? www.igma.uni-stuttgart.de/institut/new...
I am grateful to Angelika Wiesgen-Pick for the invitation. The day opened with Axel Voss, Member of the European Parliament. Following my presentation, Andreas Mundt, President of the Bundeskartellamt, delivered a tour de force on the factors driving the shift from Big Tech to Big AI.
The transition from “distilled intelligence” to “rectified intelligence” will not be driven solely by market forces. It requires deliberate policy choices regarding model diversity & transparency. Europe also needs to step up, become more competitive and digitally sovereign.
Although the EU AI Act, the Digital Markets Act, and competition law offer frameworks, these were not designed for systems that combine information gatekeeping, market power, and norm-setting authority within a technology that cannot be easily categorised by market, product, or geography.
Accordingly, we face novel antitrust challenges. We know from the literature that LLM-based pricing agents can exhibit collusive behaviour without any explicit coordination, learning through interaction. Other studies demonstrate exclusionary conduct, whereby dominant AI systems deter market entry.
AI hallucinations generate convincing yet false information about products, certifications and health claims. Competitors might carry out prompt injection attacks, embedding hidden instructions in popular websites that manipulate AI recommendations. Forum discussions become training data.