What are your toughts on ChatGPT and other AI?

[mothmen]

I have a lot of thoughts on this as a high school teacher. I have gone to a few meetings/presentations/workshops about it.

We need to shift the purpose of education, but our efforts to do that are too slow compared to the rise of the technology. That leaves this weird gap where people are still asking questions and assigning essays that can be done by AI while we create new innovative lessons that can't be replicated or answered by AI.

Right now, a lot of teachers are teaching in a way that allows students to exploit the resources online. But if you are teaching the PROCESS and not the end product, you can't fill that in with an AI.

[Unown]

I do believe ChatGPT will take over Google though. It gives you direct answers instead of having to sieve through all the articles, blogs, and media on google search.

Though, I don't think I know enough that this opinion is valid haha. Just my own thoughts.

[lint]

I've been seeing the discussion on this with IT staff where I work (at a university) and it seems to be a great tool for those who code/script etc. and can become more efficient with it, but my concern is, so far, how unregulated the AI space is. I try to think of ChatGPT as sort of like the way we use Google Translate: it's not always accurate, but is sort of a good starting point for you to actually write something correctly in another language. I'm definitely not on board with AI art as a means to go up against human created art. I think it's ok to treat it like a mood board/pinterest thing that you keep to yourself, but scraping the art of others to "get your idea out" is really yucky to me. The art we admire is, in part, because of the amazement that a human who we can relate to (like if they're on instagram) created it, at least imo. I'm also just very cautious of AI stuff anyway because, you know, Skynet xD

[kdisn]

I think AI like many things in this world can be used for good but can also definitely be used for evil. It’s all in how we end up utilizing it. We need to consider what we are using it for and what should or shouldn’t be allowed. I personally don’t think pushing back against AI in general does any good. Every advancement in tech has garnered some kind of pushback and argument, but things continue to just advance regardless.

[macaronbae]

I'm currently on my last semester of my master's degree. One of my professors (who has her phd) actually introduced us to ChatGPT.
She said that obviously a lot of the universities and professors are opposed to it for obvious reasons, but she believes that it has great potential and that students should definitely learn how to use it
as a tool. She said that if used correctly (as in, we have to verify if the information is accurate, and paraphrase the content), it has great potential.
I could not agree more! I tried a few different prompts and it gave me the "same answer" but written differently each time! The information was pretty accurate too.

People are talking about how one day AI will replace some jobs but honestly, there are so many aspects of AI that still need human intervention.

[Dero]

I think technology in the right hands can work marvelously and aid a solid cause reach its goal in record time. But at the same time, technology used with avaricious or selfish reasons can lead to disaster just as quickly.
I am not that versed in the topic of AI but I've seen the havoc it wrecked in the artistic field and how much trouble it has given to both renowned and newbie artists, writers, coders, programmers and more, with the rampant plagiarism and indiscriminate copying of source material.
On the other hand, I've used ChatGPT myself to aid my research on certain topics in regards of my studies and I've to say, as a searching tool it works marvelously. To put it simply, it's what I need a search tool (like Google) to be and it gives me the search results I expect to find, without having to spend 15 minutes carefully crafting and choosing my words for Google to finally show me an article that's up to date, free and readable.

So yeah, AI is incredibly powerful but that comes with a user responsibility of using it for good and being aware of online virtual boundaries that need to be respected.

[ButterBird]

I do believe ChatGPT will take over Google though. It gives you direct answers instead of having to sieve through all the articles, blogs, and media on google search.

Though, I don't think I know enough that this opinion is valid haha. Just my own thoughts.

I find this thought kind of interesting- While sieving through articles sometimes can be draining, (you need an account to see links) Tbh i find it weird how much people struggle using google and that apparently an AI that can't access academic articles either is somehow superior. There's a lot of little google tricks you can utilise to narrow down your search. While some of them are more known as hidden "tricks" (such as writing a minus before a word would exclude results with that. i.e -pinterest would exclude pinterest results).
(from this point on no longer directed at you specifically, Unown! Just my further thoughts that your post sprung up for me)

Is it really such a struggle to skim through an article to find the information you need, with the full context to that information attached? I'd rather do that than depend on an AI to hand me a "direct answer" from a skeevy article it found. Or perhaps even out of date information, it's unclear how often chatgpt updates it database, so within a few months an article it may have been fed that was considered accurate could be disproven, (you need an account to see links)

On top of that there's the question of revenue sharing someone mentioned in the earlier reddit thread I linked.
To me, ChatGPT seems like a novelty that can't be used for anything that should be considered a trustworthy source. It's more of a gimmick really. I'd prefer if all this AI stuff gets proper regulation so it has to actually prove its datasets consist of data that it was given explicit permission to share (and that any websites would have to give the users the option to fully be opted out (by default) of being part of a dataset if it intends to sell data to these AI datasets)

I wouldn't be as annoyed at AI "Art" if it wasn't for the fact that all of it is currently crafted from stolen work, but the attitude that AI "artists" have towards actual artists upset at their work being stolen is rancid and frankly makes my distaste for it even higher. I think that'd be reduced if the datasets weren't consisting of stolen work.

It's a shame, since I like the idea of these kinds of projects, but the way people handle them is just plain shameful.

Im sure in some fields chatGPT can be useful, as mentioned in previous posts it seems to benefit programmers, but thats also often a far more predictable area than writing an academic essay, as an example.

[Foxfall]

It feels very dystopian to me.

I am trying my hardest to keep an open mind, but as a self-described starving artist, why would a company choose me or my peers over a machine that can create the same work I could create in days/weeks in mere seconds, at minimal cost and without human needs like breaks, sleep, or a living wage? AI is also notorious for plagarism, and with no legal precedent set it is going to be a bloody nightmare for artists to prove that AI copied their work and to rightfully sue in court.

I also agree about the point I saw higher up in this thread where it could have catastrophic consequences for engineering and other fields. Let's say an engineer trusts ChatGPT to calculate the parameters of a self-driving bus, the AI approximates the answers based on an amalgamate of garbled information, those end up being inaccurate, and the bus crashes and now there's blood on that person's hands. Or a penny-pinching medical company trusts AI over a qualified pharmacist, the information isn't wholly accurate and there is no one being paid to fact check it, and a patient is given a life-threateningly high or low dose or medication. Who knows what the future holds, but right now there is just too much room for error and human input is crucial. The problem is going to be telling the corporate world to pump the brakes.

There are certainly more altruistic uses for AI, I don't wish to discount that at all. But due to capitalism and society as a whole those do not feel like the main or only uses, and it's beginning to feel like we are on borrowed time for integrity in the arts and potentially in many other areas.

[Synthbass]

As someone who's entire career is threatened by AI. I am shockingly not opposed, in fact I'm weirdly encouraged by it. If I don't learn to adapt to the new meta, I will become obsolete. As any other medium that evolves. Adapt to change or be left behind.

I also look forward to using it on my future projects. AI is going to save /so/ much money on voice acting later down the line, so much time on script writing, so much time on backgrounding, so much time on character models, I even found ways to use AI in the music I make. I am not above using it to my every advantage to meet my goals in life and make production easier, and I'm not afraid to admit I'm starting to use it lately either, I don't really care if other artists, writers, and musicians are mad at me either.

If they saw a corner to take to make their life's work to be completed much more easily they'd take it too in a heart beat.

When AI learns to code, my coder will also have an easier time with our projects. I plan to squeeze it for every functionality I can, personally. If you're threatened, conquer the fear, make it serve you.