Today’s episode brings major news from the world of artificial intelligence. OpenAI has taken an unexpected step by bringing old models back to ChatGPT — just a week after removing them. Then we’ll cover the latest round of conflict between Sam Altman and Elon Musk — from mutual accusations to heated exchanges on social media. Next, we’ll talk about a new OpenAI startup developing brain chips — a direct challenge to Neuralink. After that, we’ll look at a major update to Deep Agent from Abacus AI, which is changing how we think about automation.

GPT-5 is one of the most controversial releases in the history of OpenAI. Some call the model a breakthrough, others - a disappointment. Why did the new version of ChatGPT seem cold and soulless, what caused a storm of criticism, and how technical errors spoiled the effect of the launch? We will analyze what has actually changed, why OpenAI could sacrifice “character” for the sake of a billion-strong audience, and how competitors like Google Gemini and Grok 4 are getting stronger against this background.

OpenAI has released GPT-5 - and it's not just a language model. The new AI can program apps, co-author scripts, help scientists conduct experiments, make business forecasts, and even support you in the most difficult moments of life. In this podcast, we analyze real-world use cases of GPT-5.

OpenAI has released GPT-20B and 120B as open source — a breakthrough that no one expected. Now the most powerful language models are available for free and work locally — without the Internet, without API, and without restrictions. We analyze the main benchmarks, weak points, opportunities for additional training, and the impact on the market.