New Calling
New Calling – A cartoon captures how the runaway success of Anthropic and OpenAI is leaving many other tech companies struggling to keep pace with their own models.
Only six months ago, the AI race looked tightly packed, as Google, xAI, Meta, Anthropic and OpenAI released models in succession, with each advance in reasoning, coding or tool use answered by another competitor. That balance has shifted because Anthropic and OpenAI are turning model capability into products such as Claude Code and Codex, embedding them in workflows that strengthen both product and model.
Meta offers the clearest case study of this shift because it entered with capital, compute, talent and distribution, yet its AI push has failed to establish model leadership or a breakout product. Arguably, Llama 4 underwhelmed, Muse Spark ranked fourth in a performance index while lagging in coding and abstract reasoning and Vibes became a feed for AI video rather than the indispensable assistant Meta needed.
Meta responded by investing $14BN in Scale AI, recruiting Alexandr Wang, restructuring teams and budgeting up to $145BN for AI infrastructure in 2026, yet Mark Zuckerberg conceded that agent development “hasn’t really accelerated in the way that we expected.”
The latest twist is Meta’s plan to sell excess computing capacity. This doesn’t suggest its investments failed, but if Meta’s products were absorbing demand as expected, spare compute would be less available to sell to others.
Sources:
Peter Cohan (Jul 09, 2026) – OpenAI And Anthropic Are Preparing For IPOs—Here’s How They Stack Up – Forbes
Jon Markman (Jun 19, 2026) – Why Meta Paid $14.3B For Scale AI And Alexandr Wang’s Data Empire – Forbes
Parmy Olson (Jul 07, 2026) – Zuckerberg and Musk’s AI Failure Club Has Its Perks – Bloomberg
