In this edition we dive into the latest implication that AI usage is causing in real life. What this means in practice is that, the Human is in the loop but Humanity is not. Practically, it takes AI less than 1 second to make a decision if you have the right to a claim or not, while looking only at numbers and not at all at the human aspect. Might be good for the business, but for the human side, it’s not!

Additionally we dive into a free dictation app from Google, which runs offline on your device, how Slovenia is setting an example in a collaboration with Khazakistan, and how AI giants like Anthropic have already reached 30B in revenue.

In the race for speed, we’ve traded judgment for buttons. New data reveals a chilling trend in how AI is being used: In the military, strikes are approved in 20 seconds. In healthcare, insurance claims are denied in 1.2 seconds. Technically, there is a "human in the loop." In reality, humanity has left the building.

When a physician "reviews" 60,000 claims a month, they aren't practicing medicine, they are acting as a digital notary for an algorithm. This isn't just an ethics problem; it’s a massive business risk.

When humans stop engaging, they stop learning. Your team loses the muscle memory needed to handle a crisis when the AI fails. Customers don't mind AI efficiency, but they demand human responsibility. "The computer said so" is no longer an acceptable defense.

The "weight" of a difficult decision is the only thing that keeps an organization honest. If your AI implementation removes the friction that makes people pause and question, you aren't becoming more efficient, you’re becoming numb.

Last month, 48% of visitors to documentation sites across Mintlify were AI agents—not humans.

Claude Code, Cursor, and other coding agents are becoming the actual customers reading your docs. And they read everything.

This changes what good documentation means. Humans skim and forgive gaps. Agents methodically check every endpoint, read every guide, and compare you against alternatives with zero fatigue.

Your docs aren't just helping users anymore—they're your product's first interview with the machines deciding whether to recommend you.

That means:
→ Clear schema markup so agents can parse your content
→ Real benchmarks, not marketing fluff
→ Open endpoints agents can actually test
→ Honest comparisons that emphasize strengths without hype

In the agentic world, documentation becomes 10x more important. Companies that make their products machine-understandable will win distribution through AI.

Make Your Docs Agent-Ready

In a landmark financial update released this morning, Anthropic revealed that its annualized revenue has surged to $30 billion, up from $9 billion just four months ago.

The number of enterprise clients spending over $1 million annually has doubled since February, now exceeding 1,000 customers.

To sustain this growth, Anthropic has signed a massive infrastructure deal with Google and Broadcom. Starting in 2027, Anthropic will consume 3.5 gigawatts of next-generation TPU-based compute. For context, that is enough power to run approximately 2.6 million homes, marking the largest single-cluster compute commitment in AI history.

Southeast Europe is formalizing its role as a "Sovereign Digital Hub" through strategic bilateral partnerships and industrial automation.

Slovenia and Kazakhstan finalized a Memorandum of Cooperation this week. Slovenia, now a top-25 global digital economy, will provide the technical "know-how" for Kazakhstan’s "Year of AI" (2026), focusing on e-government and banking automation.

In a massive "Applied AI" win for the region, GrayMatter Robotics (GMR) and HII (America's largest shipbuilder) signed an MOU today to integrate "Physical AI" into shipbuilding. This technology, which automates grinding, coating, and welding, is the same stack being piloted in Croatian and Serbian industrial zones to accelerate throughput by a projected 15% in 2026.

Google just did something very un-Google-like: they released a powerful new app on iOS without a single press release or blog post.

The app is called Google AI Edge Eloquent, and it’s a direct shot across the bow for paid dictation services like Wispr Flow and SuperWhisper. While competitors often charge $15/month or more for AI-powered transcription, Google is giving it away for free and it works entirely offline.

Using on-device "Gemma" AI models, the app transcribes your voice in real-time without sending data to the cloud. It doesn't just transcribe; it edits. The app automatically strips out filler words, stumbles, and mid-sentence corrections, turning raw speech into polished prose.

With one tap, you can transform your dictation into bulleted "Key Points," a "Formal" email, or a "Short" summary. Because it’s offline-first, your confidential meetings and brainstorms stay on your phone.

Google is leveraging its "Edge" AI to move processing from massive data centers to your pocket. For business professionals who "think out loud," the barrier to entry for high-end AI dictation just hit zero.

The era of "chatting" with AI is over. The era of the Autonomous Agent has arrived, and it’s already breaking the stock market. Last week, Anthropic’s release of Claude Cowork, a specialized agent for legal and financial automation, triggered what analysts are calling the "SaaSpocalypse," a sharp sell-off in SaaS stocks as investors realized that specialized AI agents might soon replace entire software categories.

But while the big labs fight for the enterprise, a wild card has emerged from the open-source world: OpenClaw.

Unlike a chatbot, OpenClaw is like a digital "maid" with the keys to your house. It lives in your messaging apps (Slack, WhatsApp) and has deep system access to run shell commands, manage files, and triage your inbox autonomously.

VentureBeat warns that 180,000 developers just made their security models obsolete. Because these agents operate within your authorized permissions, traditional firewalls can’t see them. If an agent is tricked via "semantic injection," it can exfiltrate data without triggering a single alert.

To make agents useful, you have to give them power. But one "hallucinated" wire-cross, like an agent adding an illegal write-off to your taxes or frying a server via a bad script, can be catastrophic.

The Takeaway: We are moving from "Human-in-the-loop" to "Human-as-the-guardrail." The companies that win won't just be the ones with the most powerful agents; they’ll be the ones with the best ontology, a shared "code of conduct" that ensures their digital workers don't go rogue.

The Question for Leaders: If your AI agent makes a $50k mistake while you’re asleep, who is accountable? If you don't have an answer, you aren't ready for the agentic age.

Thanks for making it to the end! I put my hard work and dedication into every email I send, I hope you are enjoying it.

Btw if you want to get your brand in front of a fast-growing audience of founders, investors, innovators, and tech professionals from South-East Europe all the way to Europe and the US, Signal connects the dots between local and global opportunities, and your message can be part of the story.

Send an email at [email protected].

See you on the next edition,

Çelik