I feel it's a shame Python won the day, like JS. And then there are languages like Elixir which are a joy, from my little experience, which don't receive the attention they should. Perhaps it will be less relevant in the future, with AI writing much of the code. But it still feels like a loss to me.

It would be interesting to study what is really occurring in the AI to produce its last response, in the vein of www.anthropic.com/research/tra...

It took 45 minutes but I eventually was able to reason the AI to conclude it was using a faulty line of reasoning.

On one hand it was a waste of my time, on the other hand it gave me understanding on how it was reasoning, giving me insights on how to prompt AIs.

Maybe their generous free tier is too much. Which then makes me wonder why. People trying it out? Software tools have been free, and so we expect the same? It's a challenge to get the org you work at to pay for it?

Knowing how to breakdown tasks and define them for the agent seems critical. This is the skill I've been learning.

Because this change seemed subtle and simple but yet had far reaching consequential implications is a good example of a type of AI issue to be concerned about, but they don't seem to receive much attention. It is this lack of attention which is what raises the level of these types of issues.