Should GPT exist?.
I still remember the 90s, when philosophical conversation about AI went around in endless circles—the Turing Test, Chinese Room, syntax versus semantics, connectionism versus symbolic logic&#…
Read in full here:
This thread was posted by one of our members via one of our news source trackers.
I’m just not buying that chatGPT will be anything more than “demo ware” for many years to decades to come. After being in this career for more than 23 years I have seen MULTIPLE trends and tech stacks that have promised less errors and, cheaper software, and yet… here we still are.
Really, among other errors null references have been in languages for decades, now all of a sudden the heavens are opening up and AI is going to solve a software problem that’s existed for decades? LOL
Back in the late 90’s they were talking about AI and lisp eliminating the need for devs except to periodically change the input parameters of a program as it “mutates” code to a final product. There was CASE tools back then, and editors where you just drag and drop controls and “fill in the blanks” to get the behavior, and on and on.
Here is the biggest source of skepticism: how do you alter or fix a program that generated? It’s not possible to give an AI specifications without knowing what is already there and how it works. Even at this point, you can ask chatGPT to write a program that does ‘X’ and each time you ask it the same thing you won’t always get the same result. There are so many issues with chatGPT that extend passed it just being an engineering issue. Machine learning itself won’t allow companies to stop hiring devs. When we finally invent a true generalized AI then the conversation about whether or not devs are needed changes.
Here is where I do see see AI like chatGPT fitting in. It would be a tool for devs that’s essentially an extension of just a code template in your editor. It would build the higher abstracted building blocks of the program and you will have put it together.