Ah, FORTRAN 77.... Fond memories.....
Man, I thought I was old.
I started out with Basic on a TRS80 back in the early '80's.
Ended up really getting into Mac's in the mid '80's, so I had to learn Pascal, as all of the OS documentation was in Pascal for the Mac back then. Transitioned to C and C++ in the '90's, but only as a hobby. Started out with Dave Mark's "Learn C on the Macintosh" then on to his other book "Learn C++ on the Macintosh", and of course, backed this up with Kernigan and Ritchie's "The C Programming Language," as well as other really nerdy reference books.
But it turns out I was much better at flying airplanes than I was at writing computer programs; coding has always been just a hobby for me.
AI/NN stuff is fascinating, but I just don't have a big enough brain pan to ever be able to get into it in any meaningful way.
Back on topic...
After noticing some pretty good progress in 10.4 with smoothness of deceleration and a few other things, I'm also now experiencing the other problems guys have been reporting in 10.4, especially its propensity to run red lights, attempts to go around traffic stopped at stop signs or traffic signals, phantom FCW's, etc.
I've read a few posts about guys wondering if FSD takes the preceding car's tail lights into its decision making process. Next time you take FSD for a test spin, try this... every time the car in front of you shows brake lights, take note of if your car immediately applies some amount of deceleration, even if the preceding car shows no real deceleration.
Where I'm at (we have a lot of "snow birds") this condition happens a lot... the preceding car will show brake lights, with no to only very slight deceleration. Yet every time, my car will immediately start to slow as soon as the brake lights on the preceding car appear. The preceding car needs to be "fairly close" (which is admittedly subjective) for this to happen... if the preceding car is 1/2 mile ahead, it won't do it.
Maybe it's only a coincidence, which is why I'd like to hear other people's experiences with this.