Haha, I just said to my husband, so do you know what lossy means and he is giving me a lecture on compression algorithms (I should say that I come from the deepest darkest humanities and he comes from science.....)

I had a somewhat technical thread on Twitter about compression algorithms and underlying conceptual ideas right after Ted Chiang's "blurry JPEG" New Yorker article came out. I have deleted my Twitter account, but perhaps I should resurrect that thread from the downloaded archive.

That would be helpful. I wrote a blog post on AI metaphors last year where I make a somewhat dismissive remark on that metaphor. I should have asked my husband earlier! blogs.nottingham.ac.uk/makingscienc...

Wasn't sure if you really wanted an answer. An easy way to think of it is bones, structure. When we find a skeleton of a human who died 1000s of years ago we can learn some things about how they looked. Their height maybe, or how their teeth were. We can't learn everything about them though. We

likely can't tell what their haircolor was or whether they were right or left-handed or what their name was. That's what lossy means. Structure gives us some of the information, but we lose detail.