Replace with the AOL algorithm from circa 1997

[TolstoyDotCom]

Is your feature request related to a problem? Please describe. The "Probable Spam" section is full of normal tweets indistinguishable from the tweets that Elon Musk thinks aren't spam. Vijaya, Roth, and others created a horrible system that even the Stasi would have rejected as inaccurate. Elon Musk has put his own grossly incompetent spin on their system.

Describe the solution you'd like Since the goal is to ID spam tweets (that is the goal, right?) then turn to the experts: email providers who figured this out decades ago. In fact, the AOL spam algorithm from the 90s is about 100,000 times more accurate than the Vijaya/Roth/Musk setup. I'm sure Steve Case would be willing to make a deal.

Describe alternatives you've considered If Steve Case won't come to the bargaining table, approach whoever took over from Compuserve.

Additional context Elon Musk is not only very pro-censorship, he's technically incompetent.