Ever since the dawn of agentic AI I transitioned more from writing code to reading what AI is doing and monitoring it. Naturally going from 1 session to at times 3 or 4 has increased the amount of AI produced words I consume.

Recently I've started to notice more and more how I have unconciously adapted the subtle quirks of AI generated responses and have begun to sometimes have a double take if what I've just written is AI. I've also started to incoorperate language and other words that I have never used before in my life and I've actually seen this specific phenomenon on hackernews where the response is clearly written by a human, but there's 'AI language' in it.

Another thing I have noticed is that my writing tends to degrade due to the lack of care due to the fact that AI is often more than smart enough to figure out what you want even if it is filled with typos and is just raw thoughts written down as text.

What has been your experience, have you noticed similar results, but manifesting in different ways or am I alone in this?