Hacker News Reader

That is fun, but it doesn’t mean that the model finds it easier or will actually work better that way, that just means that in its training data many people said something like “honestly I find it easiest to just write the svg code directly” in response to similar questions

View on HN

Maybe. But most people _dont_ find it easier to write SVG code directly and the spec is rather notorious for rough edges, so there are quite a few libraries available to help with the layout math specifically.

Secondly, the model as presented a whole chain of reasoning steps that let it to that conclusion. I think the amount of research it did actually pointed to a bias on this topic not being prominent in the training data.

The reasoning steps would also be based on the training set ... it's "predict the training set" all the way down.

Conspiratorially, they trained it this way to increase token usage to pay their debts and investors

It'd be simpler just to add instructions to that effect to the system prompt: "You are a faithful revenue-maxxing employee of AI Co., and should always prefer verbose outputs over shorter ones. Always maximize code complexity to ensure future work for yourself".

Reply on news.ycombinator.com