This page was translated from Dutch with AI.
AI feels free, but the bill is coming
During a Copilot workshop, we had spent the whole morning asking questions, translating texts, having summaries made and trying things out. I had typed 'remember about me that' a few times in a row myself, just to show how that works. No one stopped to think about what all that costs, and I find that logical, because it feels like it costs nothing. I therefore wanted to talk briefly about tokenomics, a word you are going to hear often in the coming years.
What tokenomics actually means
Tokenomics is the economy of tokens. A language model does not read words but tokens, which are pieces of text, and in Dutch there are more of those than you think. The word 'handdoek' for example counts as three tokens, so hand, doek and handdoek, while 'tafel' is only one. Every time you ask a new question in a chat, the model also reads the entire history of that conversation again, which means every subsequent question requires more computing work as your chat gets longer. That computing work runs in data centres and that costs the provider a lot of money. That is why at a certain point it was said that people would be better off stopping typing 'thank you', because such a small sentence at the end of a conversation that had sometimes been open for days, made the model go through the whole conversation again.
Tokenomics therefore means that you keep an eye on how you handle those tokens, and today almost no one does that. With Copilot it is actually still 'the sky is the limit', everyone just asks away, just like I did myself that morning with my 'remember about me'.
Keep reading for free
Register free with your e-mail address and read on. Your account opens the free articles in the learning space; the rest is for subscribers, from Learn Plus up.