This is exactly it. Also the reason that inputting an excerpt of a classical text (cfr. Dracula, Frankenstein, …) returns “likely AI” results - it’s literally what it was trained on.
But the fact that they say their system is “advanced” and “refined”, while they don’t even add some guardrails as, i.e., date checks, that are the minimum one would expect to be added to reduce the likeliness of both false positives and false negatives.
It’s just stunning.
This is exactly it. Also the reason that inputting an excerpt of a classical text (cfr. Dracula, Frankenstein, …) returns “likely AI” results - it’s literally what it was trained on.
For anyone doubting Kyrgizion’s claim:
But the fact that they say their system is “advanced” and “refined”, while they don’t even add some guardrails as, i.e., date checks, that are the minimum one would expect to be added to reduce the likeliness of both false positives and false negatives. It’s just stunning.
And that’s still not really how that works. Those checkers just aren’t very good.