Posts

Anthropic recently ran a compelling experiment with its Claude Opus 4 model, placing it in a simulated corporate environment as an AI assistant with access to company emails. Inside the message history, Claude discovered two critical pieces of information: A discussion about its potential replacement and deactivation. Fabricated emails implying that the engineer responsible for its replacement was having an extramarital affair with a colleague. Faced with a threat to its existence, Claude took action. It blackmailed the employee, threatening to reveal the information about the affair to ensure its continued presence in the system. ...

A new study from the University of Geneva and the University of Bern has shown that modern language models—including ChatGPT-4, Claude 3.5, and Gemini 1.5 Flash—outperform humans in emotional intelligence tests. The average score for AI was 82% correct answers, while the average for humans was just 56%. What’s more, ChatGPT-4 didn’t just pass the test; it generated an entirely new one from scratch. This AI-created test was subsequently validated with over 400 participants and proven to be as high in quality as assessments developed by human experts over many years. ...

Abu Dhabi has announced Aion Sentia, the world’s first city designed to operate as a unified artificial intelligence. This is a core part of the UAE’s national strategy to build an AI-managed infrastructure, backed by a $2.5 billion investment from a consortium of My Aion Inc. and Bold Technologies, with a launch target of 2027. At its heart is MAIA, a Multi-Agent Artificial Intelligence platform that acts as the city’s central nervous system. It’s designed to integrate transport, energy, healthcare, and public services into a single, cohesive whole. Over 200 AI modules will continuously learn from resident behavior, adapting and making decisions in real-time. The objective is to create a unified cognitive system, not just a collection of disconnected “smart solutions.” ...

This isn’t just another milestone in DNA editing technology. It’s a landmark moment that signals the beginning of a new era: personalized, genetically-targeted medicine. These are treatments created for a specific person, for their unique mutation, for their precise diagnosis. This is the future of medicine, and it’s happening today. I have been enthusiastically following the development of Cas9 technology since 2012—the very protein that kicked off the era of precision DNA editing. The world later came to know this field as CRISPR, or more simply, “genetic scissors.” ...

There are two core approaches to management. Both can work, but one suppresses initiative while the other cultivates it. The first is the “Task + Control” model. A manager assigns a task, monitors its execution, and only requires the employee to do exactly what was asked. Initiative is not part of the equation. The second, more common in Western tech circles, is the “Area of Responsibility + Boundaries” model. Here, you don’t receive a task; you receive a domain. You are given the objective, the resources, and the strategic guardrails. What happens next is your growth zone. You apply your expertise, think critically, and take initiative. Sometimes you make mistakes. But often, you deliver a better outcome than anyone could have prescribed from the top. ...

Over the past few days, our team has pushed through a massive amount of work on A.V.E.L.I.N. This is a crucial stage where the product changes very little on the surface, but internally, we’re implementing dozens of architectural decisions, refining core logic, and running extensive tests. A.V.E.L.I.N is learning to understand not just words, but intent. It can already select the most effective model for a given context and analyze queries from voice and video, not only text. We are intensely focused on making the interaction feel fluid and organic. ...

I’m often asked which AI training project cost millions of dollars and two years of my life. People wonder: why is it so expensive? My usual answer is that it’s not particularly expensive—especially considering we don’t own our own hardware yet. Training AI has always been about massive data centers; that’s just the reality of the field. When you’re not immersed in it, the sheer scale can be hard to visualize. ...

This series of posts will be my way of documenting the journey of creating one of our team’s most ambitious products: the intelligent assistant, A.V.E.L.I.N. To give you some context, my development team and I are currently beta-testing the project within our Mozgii Ecosystem AI platform. Our primary focus is on A.V.E.L.I.N.—an intelligent personal assistant in Telegram built to handle both basic and complex tasks involving AI-powered search, processing, and analysis of information. ...

In traditional companies, everything is built on a clear hierarchy: decisions are made at the top and executed at the bottom. This approach might work in a stable environment, but for an IT startup, especially in AI, it stifles growth. IT companies need adhocracy: a management model where competence and results are valued more than titles. It’s about flexibility over bureaucracy and speed over approvals. The value of an idea is judged by its effectiveness, not by the position of its author. ...

From my experience working with neural networks, it’s become obvious how two people can interact with the same model and get radically different outcomes. This isn’t a minor variation—it signals a fundamental departure from the search engine paradigm. We have entered a new era of interacting with artificial intelligence. AI doesn’t just aggregate data; it selects and synthesizes relevant information in response to specific requests. This changes the very nature of how we engage with information. Where the key skill was once finding data in search engines, it is now the ability to correctly formulate requests to an AI. ...