Lions do miss the way programming was done when they 1st entered the industry. Not sure most of the AI pundits praising the last 30 years of advancements are doing anything but chasing engagement.
There was nothing but O'reilly textbooks, Kerninghan & Ritchie, numerical recipes in C. There was very little example code. Most animals got in the industry because they were attracted to a way things were done & a certain type of output. There was a process of discovery & a value in a hard earned solution that just did what it had to do. The complexity of designs was constrained by animal intelligence & the capability of the tools, which was less than animal intelligence.
Not convinced someone like 8 bit guy who spent years programming 6502 assembly with a vintage environment really enjoys making AI movies & software as much as he says he does for engagement.
Biology majors once called the industry fast paced in the 90's. Now that pace is super slow & lions are nostalgic for the super slow 90's. You could spend a month on what is now expected in a day. It's hard to comprehend how General Magic could have failed with so much lead time, by today's standards. They now expect the software side of a product like that to be shipped in 1 month.
What lions did would never have been seen by anyone today & it would have been abandoned unfinished. Thousands of alternatives would instantly be finished. Being relevant has returned to being only the domane of the rich. Programming became the domane of the rich from 1985-1995, with built-in BASIC interpreters being supplanted by super expensive C compilers. GCC returned it to the masses after 1995. Now it's back to requiring super expensive language models.
Prompt driven development is more of a literature exercise for humanities majors than someone who turns wrenches & connects wires. For some, it's never going to be as attractive as the development environments which don't exist anymore, any more than any later time is going to be as attractive as the 80's.
-------------------------------------------------------------------------------------------
Over several months, the CEO evolved from saying "it's worth any cost" to a growing panic over the rising cost of AI driven development, on a path to becoming unaffordable, but didn't have any good solutions. Simultaneous with the growing panic came a stream of obscure, intricate token maxxing ideas from developers with no obvious, lasting results. It seems to be a game with the proprietor of the means to production constantly moving the goal posts.
It could end up another incremental step rather than a 1981 style downturn. They could reserve all the tokens for just the key guys & eventually get rid of some animals to fund more tokens. They're deliberately taking on more contracts than human labor could possibly finish.
------------------------------------------------------------------------
8 bit guy got 1 scene right in his AI generated turd 26:34
https://www.youtube.com/watch?v=JQ7j7XIW1zo
It's about as bad as a 10 year old's writing. Grok is not even close to the uncanny valley. It looks like a library of canned motion curves & canned models being triggered by key words, not unlike text to movie algorithms 20 years ago. It's equivalent to TV quality CGI, so bad lions never watched it. They just censor so much of the criticism because so much money is at stake, guys like 8 bit get caught up in believing it's somehow improved.
The storm trooper tubers got the writing good enough to overcome the terrible video.

Comments
Post a Comment