xianc78 @xianc78@gameliberty.club

Despite OpenAI's claims ChatGPT-4 is getting dumber.

tomshardware.com/news/chatgpt-response-quality-decline

GPT-4's ability to figure out if 17077 is a prime (it is) has fallen from 97.6% in March to 2.4% in June.

This highlights a couple of problems with the whole notion of GPT-4 and other Large Language Models:

1. They literally know nothing. It's all word games. If you understand what a prime number is it might take you a few minutes to run through the possible factors on a calculator and get the right answer, or just program a computer to do it for you. GPT-4 is incapable of doing that.

2. They learn nothing. If someone keeps asking you if 17077 is prime (it is) you might remember the answer, maybe even write it down. GPT-4 is incapable of doing that as well.

cartridge repros are actually awesome. it's not great for collectors trying to get the actual thing, but for people who just want to buy a game and play it on their actual console, it's not like you can buy it from nintendo anymore. they're a lot cheaper than buying the legit cartridge. as long as you know what you're buying is actually a repro, i don't see any harm in it.

ChatGPT translated version of my effortposting :

"No, it is incorrect for several reasons. The needs of consumption do not determine production (Ricardo's Objective Theory of Value), but rather the other way around, production determines the variables of consumption (Menger's Subjective Theory of Value). In other words, one cannot consume what has not been produced first. The drunkest person in the world will effectively be an abstainer if nobody invests in cultivating grapes or barley to later process them and create the wine or beer that the drunkard so "needs." Likewise, nobody will invest in cultivating or producing alcohol if they happen to live in a community of abstainers. So, from the very beginning, Marx's hypothesis is incorrect.

Secondly, the skills necessary for performing activities in rural and urban populations are different because the requirements of goods and services producers are different in each case. The type of skills needed for agricultural or heavy industry development (typical in rural areas) is very different from the type of skills needed for banking or commercial activities (more typical in cities). Therefore, trying to compare one with the other is falling into a false equivalence since the parameters that determine the economic growth of one are not applicable to the other. In other words, it is nonsensical to try to compare the "concentration of capital" of people in rural areas with those in urban areas."

Maximum Linux - May/June 2000
https://archive.org/details/maximum-linux-magazine-2000-06