ChatGPT 5.5 vs Claude Mythos: The Best AI in the World Is Here

Summary

Ejaaz and Josh discuss the launch of OpenAI's ChatGPT 5.5, its benchmark superiority over Claude Mythos, and new autonomous capabilities. They also cover SpaceX's partnership with Cursor, Amazon's $5B investment in Anthropic, and the rapid improvement of Chinese open-source models. The hosts highlight the competitive dynamics in AI and hardware, and note market reactions to Anthropic's design tool launch.

• OpenAI launches GPT 5.5, claiming it beats Claude Mythos on most benchmarks.
• GPT 5.5 can handle 20-hour autonomous tasks and is used internally at OpenAI.
• SpaceX partners with Cursor (with an option to buy) to accelerate its AI coding capabilities.
• Amazon invests another $5B in Anthropic, increasing its stake to ~20%.
• Google releases new TPU chips specialized for training and inference.
• Chinese open-source models Kimi K2.6 and Quen 3.6 narrow the gap with frontier models.
• Anthropic launches Claude Design, a design tool that competes with Figma.
• President Trump comments on Anthropic, hinting at government access to Mythos.