If Today's AI Is So Smart, Why Hasn't It Reversed-Engineered The 'Tik-Tok' Algorithm?

Click to ENLARGE. (Attribution embedded).

I am FED UP of hearing folks talk about how the 'Tik-Tok' algorithm cannot be replicated.

I started programming (with paper tape on an 'ICL' machine) in the fall of 1969, aged 16. I loved programming, I couldn't get enough of programming & I was somewhat good at it. I was good enough to be hired by IBM, in my 2nd year at university -- & I started working for IBM, as an Associate Programmer, with a honors degree in Computer Technology. IBM paid for me to get a Master's in Computer Science (which I got with a mark (of rare) Distinction). So, I know a bit about programming -- though it is now dated. During my days at IBM we still coded ON coding sheets because access to online terminals were still limited OR we were using PUNCHED CARDS. 

In those days there were two types of coders (who coded on sheets): those that used a pencil & those that coded in INK. That was a LEVEL of YOUR CONFIDENCE in the code you wrote. I CODED IN INK -- & was known as an ink coder.

I also dabbled a fair amount of AI at the time. While today's AI is way above my aging brain, I remember what we were TRYING to do in the late 1970s & 1980s.

< >  < >  < >  < >  < >

To me it is inconceivable that folks claim that the Tik-Tok algorithm cannot be reverse engineered. Reverse engineering is not like trying to violate the laws of physics.

I am sure it can be reverse engineered.

I can't believe it has NOT been reverse engineered.

IF humans are having trouble doing it, THEN AI should be able to. It is right in the AI wheelhouse. Basically identifying patterns. AI is said to be good at it.

Bottom line: I think the Tik-Tok algorithm can & should be reversed engineered -- SOON.

😇