23 09, 2026

Jev is a new type of AI to make decisions. But how accurate are they?

By |2026-10-08T12:34:28+00:00September 23, 2026|English, Practical AI usage|0 Comments

Jev is a new kind of AI that makes decisions instead of writing text. I tested it against four Claude models on 1,261 business decisions in German and English: how accurate, fast and cheap is it, and where does it break?

18 08, 2026

How generative text AI works: how LLMs choose what to say next

By |2026-10-08T12:23:24+00:00August 18, 2026|English, Practical AI usage|0 Comments

Every generative text AI, from ChatGPT to Claude to Copilot, does exactly one thing: it predicts the next token. Five interactive animations show how words become vectors, how text becomes tokens, how the model picks the next word, and why every chat message resends the whole conversation.

16 06, 2026

Model addiction: why chasing the best AI model is a trap for your business

By |2026-07-07T08:47:10+00:00June 16, 2026|Future of Work, Practical AI usage|0 Comments

A new “best” AI model arrives every few weeks, and chasing each one quietly raises your costs, deepens your dependence and leaves you fragile. Here is why most work does not need the frontier model, and what to do instead.

28 01, 2026

Clawdbot AI is going viral, but it’s like giving your keys and credit card to a hyperactive 6 year old

By |2026-01-28T11:34:40+00:00January 28, 2026|English|0 Comments

Everyone is talking about Moltbot (formerly Clawdbot before Anthropic sent [...]