I'm not sure if learning the term 'Brandolini's Law' makes me feel better or worse about all the time I've spent refuting bad arguments.

The important distinction is that 'lying' implies an intention to deceive, making an assumption about the internal process. If I were asked to list everything I ate last Thursday, I'd likely unintentionally get some things wrong - I wouldn't be lying though, I'd just be incorrect.

Really like this way of thinking about LLMs!

Framing them as cultural technologies is an excellent way of acknowledging the power, impact, and risk of the technology without anthropomorphizing it.

Definitely agree there is a risk/safety issue here! I think we should avoid using terms like 'lying' and other anthropic language about AI though, since it can be misleading. It's fine to just call this a performance/accuracy issue

Oh dear, this is a very confused AI! We should definitely be careful about how we use AI in education! What product are you using here @beckykeene.com?

That totally makes sense - the concern would then shift to how agents are programmed to behave and who has access to the data they collect. Definitely worrying considering how powerful the big tech companies are already!

Really enjoyed your essay! Somehow it made me feel slightly better about the prospect of having an AI agent take over my day to day decisions. I'm curious what your perspective is? Do you see the potential loss of personalization as a good thing?

I love this perspective. Beyond critical thinking along, I've found mindfulness to be a valuable tool in managing the overwhelm that comes with all the hype around AI as well.