Well I spent a few minutes asking #chatgpt what features it’s next version would have.
I then asked it to pretend that it was that version and what the next version would have… I did this about 15 times until I got to chatgpt-20.
I was too scared to continue this process!
Some pretty wild stuff.
#futurism #ai #Transcendence #Omnificence #Enlightenment #cosmos
I really hope automation replaces all labor jobs eventually. But I also hope that capitalism is toppled and therefore the loss of that aspect of human labor has no negative consequences for workers.
ChatGPT is only one example of an AI that can do a job convincingly. Even if its wrong, even if everything it says are baseless lies, the fact that it is convincing enough to look and sound like something a human would write is pretty amazing. Previous forms of “AI” couldn’t write a single sentence that wasn’t obviously a product of an automated process. And gpt isn’t static. This is a relatively early iteration of it, we are only going to see it’s competency continue to improve by leaps and bounds over time.
Its frustrating to see innovation in technology that everyone hates because of how it has negative implications under capitalism. Maybe instead of being angry at the software engineers designing artificial intelligence, you should be angry at the capitalists who have not instituted universal basic income and who therefore expect you to starve to death if you are no direct benefit to them.
First a point of clarification. The term AI has become overloaded recently which leads to a lot of confusion. AI as originally invisioned, and what most people imagine when you say AI would be more accurately called a general purpose AI. It’s something of human level intelligence (or greater) able to demonstrate self awareness and the ability to reason. Think HAL from 2001 a space odyssey. To date nobody has ever created a general purpose AI or even anything that comes remotely close to one.
What most people call AI these days would be more accurately referred to as machine learning or ML. Unlike an AI a ML algorithm is created for a specific task and can only do that specific task. ChatGPT is a ML designed to parse written text and output written text. There have similarly been several ML algorithms created recently that do a very good job of generating human sounding speech or images.
I am not angry at software engineers designing ML (or even ones trying to design AI). I am a software engineer. I have studied and even written some ML myself. What I am angry about is people who either have no idea what they’re talking about, or actually do but are trying to drum up page views making wildly unhinged articles either claiming “AI” will lead us to a utopia soon or is on the verge of starting the machine uprising to wipe out humanity. I am further angry at CEOs just pretty much in general and the post-capitalist hellhole western democracies are quickly turning into, but that’s a rant for another day.