Does the AI business model have a fatal flaw?
I promise, unlike my last post this morning, this is not an April’s Fool. My latest opinion piece for Reuters has been published this morning, I address some of the evidence that shows that LLMs from Anthropic and co. are being squeezed from cheap/free AI from the bottom while the sophisticated use cases are probably not going to work because of the inherent problems with the way LLMs are being built.
Does the AI business model have a fatal flaw?
LONDON, April 1 (Reuters) - Hundreds of billions of dollars are riding on the assumption that artificial intelligence will be reliable enough for high-stakes work. New research suggests it may never be.
The AI tools that power ChatGPT and its rivals - known as large language models, or LLMs - are a genuine productivity-enhancing innovation. But they have serious shortcomings, most notably, their tendency to hallucinate, or make things up.
Read the full article here: https://www.reuters.com/technology/does-ai-business-model-have-fatal-flaw-2026-04-01/
My work introduced the Chinese Moonshot Kimi model to use in our Claude Code setup instead of Anthropics models. Promises to be similar in quality but a lot cheaper which my boss likes.
And played around a bit with the free Chinese Minimax and free opencode (claude code alternative) at home. It's no Claude but it's pretty damn impressive too.
It also makes me wonder about the moat that Anthropic or OpenAI have...
We were also very close to achieving nuclear fusion...
In any case, we’re focusing on white collars just as the world is about to be flooded with robots that will replace millions of workers in logistics and factories.