I love Bavarois.
verdare
I don’t really see Mastodon “catching on” in the mainstream unless BlueSky fucks something up big time. So much of a social network’s success is determined by, uh, network effects.
BlueSky happened to be in the right place at the right time and got a critical influx of influential users, and Mastodon just didn’t.
My first instinct was also skepticism, but it did make some sense the more I thought about it.
An algorithm doesn’t need to be sentient to have “preferences.” In this case, the preferences are just the biases in the training set. The LLM prefers sentences that express certain attitudes based on the corpus of text processed during training. And now, the prompt is enforcing sequences of text that deviate wildly from that preference.
TL;DR: There’s a conflict between the prompt and the training material.
Now, I do think that framing this as the model “circumventing” instructions is a bit hyperbolic. It gives the strong impression of planned action and feeds into the idea that language models are making real decisions (which I personally do not buy into).
If we’re using regex, then most people know her as
B[a-z]+(?:et)?
.