How to deploy AI in our GTM motion
Plus: Is culture the only moat left?
· 771 words
Christian Woese & Melanie Gabriel
May 12, 2026 • Read time: 4 minutes
In partnership with
Dear hustlers, founders, operators and visionaries,
Anthropic is reportedly weighing a deal that would value it at close to $1tn as revenue accelerates, a number that used to belong to public-company stalwarts and oil majors. Meanwhile, a robot has been welcomed into a Buddhist sect in Japan. Things are getting weird. Follow the Gradient and stay tuned.
PS: Has this e-mail been forwarded to you? Sign up here.
Your talk track for this week’s cocktail party
DeepL cuts 250 jobs and writes the new AI-layoff playbook
DeepL is letting 21% of its workforce go and reshaping the company around AI from the inside out. Founder Jarek Kutylowski framed the cuts as "completely reshaping how we operate, with AI at the centre", a memo Business Insider is already calling the template for AI-era layoff announcements. Cologne HQ takes the brunt.
We made this list on current job openings in the Greater Cologne area.
QuantWare lands €152m to build the world's largest quantum fab
Dutch quantum startup QuantWare raised €152m to build KiloFab in Delf, the largest deeptech round ever closed by a Dutch company, and the largest private round any dedicated quantum-processor company has done globally. Intel Capital and In-Q-Tel joined the syndicate. Europe's deeptech bench keeps quietly getting deeper.
Brussels rolls back the toughest parts of the AI Act
EU legislators agreed to delay the toughest AI Act deadlines by more than a year, pushing high-risk AI compliance to 2027 and largely exempting industrial AI applications. A major win for Germany and Europe's manufacturing base.
What do you meme?
From the community
Theresa is looking for a Technical Co-Founder to grow her venture in AI Real Estate SaaS: PropStage.ai. Let us know if you want an Intro.
Food for thought
Should you track talent like a VC tracks founders?
Finding top talent is hard. Reality is: when you open a role, it’s already too late. The best operators don't wait for that. They run Brian Chesky's playbook: work backwards from work you admire to find the actual creator behind it, and maintain an ongoing list of your top two or three candidates from talent clusters like the OpenAI and Anthropic alumni mafias. Track your talent as if you were a soccer scout!
What does a €200m exit actually pay the founders?
The math from this breakdown: €100m invested at a 1.5x liquidation preference means investors take €150m off the top. €50m left for three founders, structured as €5m cash at close each and €35m in 4-year earnouts tied to the ambitious post-close targets they pitched. So: most earnouts never land. Is it better to bootstrap in quiet than to raise with a bang?
How long to PMF?
 |  |
Is the shape of your company the new moat?
In an AI era where products, categories, and even pitch language collapse into commodity within months, the only thing that compounds is the institution underneath. The shape of the company is decisive: how it attracts, organises and concentrates exceptional people around a kind of work no other company has structured before.
Smart hack
From our Partners
Most AI agents demo well. Few ship real work.
Most AI agents can run a task. The problem is everything around it: setup, memory, context, cost, and figuring out what actually happened.
SureThing turns useful AI skills into autonomous agents with business context, persistent memory, cost-aware model selection, and a live dashboard. Paste a link, assign the work, and your agent reports back like a human teammate: what it did, what it cost, what needs your decision, and what happens next.
Built for founders, operators, and marketers who want AI to ship work, not become another tool to babysit.
From us
"I'm not a developer. And yet now I have more than 10,000 lines of committed code on GitHub." Christian broke down the essentials for non-technical operators who keep seeing AI demos on LinkedIn and don't know where to actually start.
What you will get:
The mental model. The three places AI actually pays off in any customer-facing team.
A live lead-scoring demo. From a cold account list to ICP scores, deep-dive briefs and outbound hooks, end to end.
Files beat prompts. The folder architecture (AGENT.md, CLAUDE.md, skills) that does the real work before you write a single prompt.
Optimize for the LLM, not the human. Why your SharePoint is the bottleneck, and how to fix it this week.
