Comment by qingdao99
4 years ago
- as probably everyone here would agree, AI-generated content will flood the Internet.
- the above will lead sites to desperately search for better bot prevention tools, eventually culminating in a few years with the introduction of systems where online activity is tied to a person's real identity via their government (privately - we will still be behind usernames but the site will be able to verify with the person's government that they are a real person) and passing off AI-generated content as human content will constitute a rule violation.
- AIs will be "let loose" on virtual machines, where they will experience many lifetimes' worth of computer use in a short time, learning to fully operate computers and start the process of replacing jobs not by having a technically capable person operating an AI, but by simply having an AI fully replace a person's job by acting identical to a remote worker from a manager/coworker's point of view. After AI gets intelligent enough, it will replace all desk jobs.
- After it becomes clear that AI will replace large portions of jobs (which I think will happen in 2023 when the next few iterations of AIs are released) there will be a push for UBI which will be supported by both sides of the political divide, as well as both the rich and the poor, and divided along the lines of who "believes" in AI and who doesn't
Contrarian take: AI continues to fizzle. ChatGPT is very impressive but so were the last ten breakthroughs in AI and none of them have proven very compelling. At least not for consumers. We'll see advances in some business processes related to text processing but that's about it. Who is going to make practical use of AI art? Maybe the low-end gaming market? I'm also bearish on writing even for low-touch marketing copy. If you've ever sat in a meeting to discuss marketing copy you'd know writing takes up about 1% of the cost and the rest is debate.
OpenAI is the apotheosis of regression analysis. It can only repackage what humans have already created. Over reliance will absolutely stultify every creative endeavor and we'll be stuck in a loop. I foresee a lot of adoption followed by a lot of disillusionment.
All that being said, I do believe it's only a matter of time before all that happens. Just not 2023.
On Linux VM, if a program is given root privilege and it just execute whatever command ChatGPT spit out, then you'll got hell a problem solver/havoc wreaker.
The hit rate of AI is still in that phase of being right almost all the time and being spectacularly wrong 2% of the time. So actually pretty close to a good DevOps engineer.
You think all of this will happen in 2023? I could see AI becoming more powerful and more of an issue for content moderation/spam/propoganda, but leading to social, political, and economic change? I'm doubtful.
Yeah, I think AI will lead to huge change starting next year when the next iterations on ChatGPT and whatnot are released and they jump from being good most of the time to being excellent almost all of the time. When that happens next year I think the large-scale changes will start as people realize what the future will be like.
*of course it will take at least a couple of decades for the large scale changes to take place but I think we're right on the edge of it starting
Spot on analysis. Especially the last point. The divide will indeed be between believers and non believers much more than rich and poor.
The rich and poor will both have among their lines people who think the AI will uplift the masses and others who think it is a tool of oppression. The real battle would be on those who think they can control it for the betterment of humanity. And those that do no.