In 2021 to 2025, most companies had begun to behave as if AI was the guaranteed future.

1. We saw large established technology companies such as Amazon, Google, Microsoft, Meta/Facebook start sending insane amounts on AI infrastructure
2. This has increased (shockingly) with each of Amazon, Google, Microsoft, Meta/Facebook spending $150 to $200 billion a year on AI infrastructure

Basically, these companies were behaving as if it is a certainty that AI will eat the world. Not just the way the Internet ate the world, but in a much more comprehensive way.

Now, in mid 2026 Everyone seems to believe Anthropic and Open AI are set to become the two biggest and only winners of AI

Please Note: I mean Core AI. I’m not saying no one else will make money from AI. Nor am I saying everyone else will die. Just that it seems the core AI race seems to be won already by Anthropic and Open AI

There are some signs of this

1. The US Government has stepped in to block Anthropic from selling Mythic and Fable outside the US, effectively saying – you’re too far ahead of everyone else
   • Amazon, which owns between 10% to 20% of Anthropic, stepped in to warn the US Government
   • There were talks for many days between Amazon CEO Andy Jassy and the US Government
2. Satya Nadella of Microsoft is talking about AI needing to be in the hands of users and companies. That everything being in the hands of Anthropic and OpenAI is dangerous
3. Microsoft owns 27% of Open AI
4. It’s interesting and scary that Amazon are so scared of Anthropic that despite Amazon owning 10% to 20% of Anthropic they are willing to sabotage it
5. It’s even more interesting that Microsoft are so scared of OpenAI that despite owning 27% of it, Microsoft is warning about dangers of an Anthropic + Open AI duopoly in AI

Large technology companies only run to the Government if they feel there is the danger of formation of a monopoly other than their own.

Basically, Microsoft and Amazon feel very strongly that Anthropic and OpenAI might get so far ahead that no one else can catch up. Now these large technology monopolies are crying about fairness and national security.

What does this tell us about the AI race and Anthropic and OpenAI’s place in it?

1. Firstly, as far as Microsoft and Amazon are concerned, Anthropic and OpenAI are too far ahead to compete with on fair grounds
2. Secondly, pulling in US Government is the ultimate trump card
   • You do not use it unless there is no other path forward

The large technology companies are very clearly signaling – AI race is over/almost over. Anthropic and OpenAI are so far ahead that we cannot catch them.

The AI Experts are voting with their feet

Whereas Microsoft have become also-rans in the AI LLM Wars. Whereas Meta/Facebook and Amazon do not even register in the AI LLM Wars any more.

Google was supposed to still have a good shot.

That hypothesis seems toast as arguably 2 of the top 5 AI minds in Google departed on the same day

1. John Jumper, co-winner of the Nobel Prize with Demis Hassabis, left Google DeepMind for Anthropic
2. Noam Shazeer, one of the lead authors of ‘All You Need is Attention’ left Google Deepmind for Open AI
   • That’s the paper that introduced the transformer architecture and made today’s AI boom possible
   • One of the 3 most cited papers of this century

We are talking about the very best of the best. Leaving the company they spent most of their careers at.

Noam Shazeer had already sold Character.AI to Google for $$2.7 billion and made between $750 million to $1 Billion.

These top of the top researchers would only leave Google if they felt Google had very little chance of overtaking Anthropic and Open AI.