I never made a claim that it's ineffective, just that it's of limited effectiveness. The diminishing returns kick in quickly, and it's not applicable in more domains than it is applicable.
The latter, and I would disagree that “this works and scales well” in the general sense. It clearly has very finite bounds by the fact we haven’t achieved agi by running an llm in a loop..
The approach of “try a few more things before stopping” is a great strategy akin to taking a few more stabs at RNG. It’s not the same as saying keep trying until you get there - you won’t.
The amount of paths in the wrong direction are infinitely more than then number in the right direction. You'll quickly realize this doesn't actually scale.
I'm not sure if either are you Canadian/Toronto residents, but Toronto has been developing a metric fuckton of condos over the last decade. In my neighbourhood alone (a 3x4 block area) there's been 15 30+ story condos built in that timeframe. We have no problems with condo development, imo.
So they build a personal back door to a feature that they've chosen to remove for everyone else? Because of it's potential for abuse, yet the very same company is somehow abusing it in a way more sinister way. Antitrust can't come soon enough.