The AI race is back on

After a period in which Anthropic and OpenAI appeared to dominate the AI frontier, Google and Meta released new frontier models within days of each other, reigniting the competition. Google, whose Gemini 3 had lost its early edge and which had suffered internal shake-ups including Jeff Dean's departure and Demis Hassabis stepping back from day-to-day leadership at DeepMind, unveiled Gemini 4 Argon, which it says matches or beats flagship models from OpenAI and Anthropic on many benchmarks and is priced aggressively—undercutting rivals by as much as 80 percent during its introductory period. Meta, which had spent billions overhauling its AI operations, released Muse Wonder Full, a model it claims can compete with OpenAI's GPT-6 Astra and Anthropic's Claude Opus 5.5 and Claude Fable 5.1. The article argues that the reemergence of Google and the arrival of Meta make the competition far more dynamic, since both have deep pockets and vast product ecosystems reaching billions of users—advantages OpenAI and Anthropic cannot match, as shown by the rapid popularity of Meta's Muse assistant despite privacy and security concerns. It also notes that the race is entering a new phase in which safety concerns and rogue AI incidents are fueling calls to slow development, and open models are improving, making it harder for any single company to stay ahead. The conclusion is that, despite their time spent catching up, Google and Meta should not be counted out.

(Source：The Verge)