AI Eng

Jacob Coxon, who worked at OpenAI and Anthropic, said both companies are moving toward self-improving superintelligence too quickly. He said researchers within the industry have serious concerns that such AI could pose a threat to humanity.

When people talk about AI in sales, they usually imagine an IT company: CRM, thousands of leads, automatic email sequences, transcription of Zoom calls, and an AI assistant that writes follow-ups. But I became more interested in another question: what can AI give to a sales manager in a traditional B2B business?

Ukrainian deeptech company AEROSTRATEG, founded in Kyiv in January 2026, is developing Ares Core — an AI-native platform for aerospace and defense engineers. According to the company’s director Volodymyr Mitasov, the product is already being tested by 12 Ukrainian defense-industrial complex enterprises, and the company has two signed commercial SaaS licenses and is preparing two more installations of local hardware and software complexes for industrial customers.

Jakub Pakhotski, Chief Scientist at OpenAI, has published a lengthy essay titled «An Alien Mind,» summarizing the company’s three-year journey from early results in training reasoning models to today’s Astra-class systems. The main message is that the pace of progress can be sustained until recursive self-improvement (RSI)—the point at which AI increasingly directs its own development—and no one is truly ready for that.

OpenAI’s internal system has created an analytical proof that the company claims solves one of the seven millennium problems—the existence and smoothness of solutions to the Navier-Stokes equations. OpenAI has also presented a formalization of the proof in the Lean language.

Bill Gates recently published a column, “ The turbulent AI era is here. The choices we make now are critical, ” about the transition to an economy where artificial intelligence takes over a significant portion of human work, and I read it twice: the first time I nodded, the second time I started arguing with myself out loud. Gates honestly describes the risks of junior professionals being displaced faster than the market can create new jobs, cyber threats growing faster than protection, and the impact of AI companions on children. He has one proposal that really worries me: taxing AI and robot tokens in the same way that payroll is currently taxed, to slow down the replacement of humans by machines and fund retraining.