OpenAI has introduced the “OpenAI Podcast,” a new series of conversations about artificial intelligence. The first episode, released on 18 June 2025, features Chief Executive Officer Sam Altman in discussion with host Andrew Mayne about the company’s roadmap and the broader state of AI research. During the conversation, Altman said the next flagship model, GPT-5, will “probably” be released this summer, with an end-June window under consideration. He framed the system as a unified successor to GPT-4o that will offer stronger reasoning, greater autonomy and better scientific insight capabilities. Altman asserted that definitions of artificial general intelligence set only five years ago have already been exceeded and that OpenAI is now aiming for “superintelligence”—machines capable of independently discovering new scientific knowledge. He predicted that AI-enabled scientific breakthroughs will eclipse all other impacts of the technology within five to ten years and said OpenAI has “cracked” the problem of reasoning. The wide-ranging discussion also touched on Project Stargate, updated research workflows, privacy safeguards, AI-assisted parenting tools and the possibility of advertising in ChatGPT. Altman additionally alleged that Elon Musk had attempted to use political influence to block an OpenAI partnership, a move he described as detrimental to the country.