In partnership with
Dear Readers,
What if ministers were no longer made of flesh and blood, but of code? Albania is taking precisely this step – the world's first virtual minister is now a reality. At the same time, autonomous AI agents such as Replit Agent 3 are crossing the threshold from helper to developer, while new superchips such as Rubin are redefining the limits of computing power. We are at a threshold where politics, technology, and society are intertwining in real time—and the question is: Who is actually controlling whom?
In this issue, we dive into Albania's experiment with an AI minister, analyze the power shifts caused by deals such as ASML's entry into Mistral, and show how agents, chips, and political sandboxes are changing the rules of the game. We also provide insights into the latest developments at OpenAI, Claude, and Replit – and what they could mean for our future. A newsletter full of breaks, new beginnings, and perhaps even a few small shocks – read on to keep up.
In Today’s Issue:
🧠 Replit's Agent 3 is building, testing, and even fixing its own code
👨🏻⚖️ Senator Ted Cruz is pushing for an "AI sandbox" to loosen regulations on tech companies
💰 ASML just invested $1.5 billion in Mistral AI, boosting Europe's tech hopes
🇪🇺 The EU is planning a "Cloud and AI Development Act" to create its own AI "gigafactories"
✨ And more AI goodness…
All the best,
Today’s Headlines
OpenAI and Microsoft Deepen AI Partnership
OpenAI and Microsoft have signed a non-binding memorandum of understanding for the next phase of their partnership, aiming to finalize a definitive agreement to deliver the best AI tools with a shared commitment to safety.
Cursor AI Gets Smarter Suggestions
Cursor has released a new "Tab" AI model, now the default, which makes 21% fewer suggestions while boasting a 28% higher accept rate, thanks to improvements from online reinforcement learning.
Claude Remembers, Offers Incognito
Claude now features memory capabilities for Team and Enterprise users, alongside the introduction of incognito chats for all users, enhancing personalized and private AI interactions.
What?! I did not see that coming
„Albania has become the first country in the world to have an AI minister — not a minister for AI, but a virtual minister made of pixels and code and powered by artificial intelligence.
Her name is Diella, meaning sunshine in Albanian, and she
— #Chubby♨️ (#@kimmonismus)
5:26 PM • Sep 11, 2025
The surge in AI-written speeches in Britain's House of Commons visualised:
— #François Valentin (#@Valen10Francois)
8:29 PM • Sep 10, 2025
Replit's Agent 3 AI Revolution
The Takeaway
👉 Autonomous browser testing and bug fixes make development processes significantly more reliable.
👉 “Max Autonomy” allows the agent to work independently for up to 200 minutes—useful for complex projects.
👉 New capability: The agent can create automations and other agents itself.
Replit has unveiled its most ambitious AI update yet with Agent 3. The new agent is significantly more autonomous than its predecessors: it independently tests apps in a real browser, detects bugs in buttons, APIs, or form flows, fixes them, and repeats the tests. This transforms AI from a mere helper to an active developer.
A special highlight: Agent 3 can create its own agents and automations – such as Slack bots or daily workflows – and run completely independently for up to 200 minutes. For developers, this not only means shorter development cycles, but also the opportunity to focus more on creative and strategic aspects. For beginners, it opens up access to rapid prototyping and experimentation without in-depth technical knowledge.
Why it matters: Agent 3 radically accelerates development processes and lowers the barriers for non-programmers. At the same time, the role of humans is shifting from coder to supervisor—a shift that could permanently change the job description of developers.
Sources:
Marketing ideas for marketers who hate boring
The best marketing ideas come from marketers who live it.
That’s what this newsletter delivers.
The Marketing Millennials is a look inside what’s working right now for other marketers. No theory. No fluff. Just real insights and ideas you can actually use—from marketers who’ve been there, done that, and are sharing the playbook.
Every newsletter is written by Daniel Murray, a marketer obsessed with what goes into great marketing. Expect fresh takes, hot topics, and the kind of stuff you’ll want to steal for your next campaign.
Because marketing shouldn’t feel like guesswork. And you shouldn’t have to dig for the good stuff.
Blackwell is the bridge between Hopper and Rubin
Rubin seems like the real “next-gen big bang,” with enormous leaps in memory and performance. The downside: extreme power consumption (up to 4,000 W per chip) and corresponding cooling/infrastructure requirements.
US Senator Cruz proposes AI ‘sandbox’ to ease regulations on tech companies
Ted Cruz has introduced a bill that would grant AI companies exemptions from certain US federal regulations for two years — subject to conditions such as risk minimization. Critics fear that this could undermine environmental and consumer protection as well as democratic accountability. Politically relevant: Balance between pressure to innovate (especially vis-à-vis China) and protection obligations. If such sandboxes catch on, it could lead to a proliferation of experimental AI regimes, with fragmented standards and inconsistencies between states.
ASML-Mistral AI deal boosts Europe tech hopes as Trump rivalry heats up
ASML (NL) is investing USD 1.5 billion in French AI company Mistral, becoming its largest shareholder. The deal is seen as strengthening Europe's technological sovereignty in AI, especially in light of growing confrontation with the US under Trump. Relevance: Europe's efforts to create its own AI capacities are no longer just regulatory, but also economic and strategic. In the long term, this could reduce dependence on US/Asian big tech, but at the same time intensify new power blocs and competitive lines within the global AI economy.
State of the Union 2025: What’s Ahead for Tech in Europe
In yesterday's speech by von der Leyen, the EU Commission emphasized the creation of European AI “gigafactories,” the expansion of cloud infrastructure, and announced a “Cloud and AI Development Act.” The goal is technological independence, economic competitiveness with the US/China, and regulation that combines innovation and sovereignty. Important: The trend is moving away from a pure focus on regulation toward active industrial policies to promote domestic capacities. This could shift Europe's role in the global power structure—for citizens, this means new jobs, but also new risks due to the concentration of power and dependence on government support.