I wonder if this AI (statistical software) hype is just like the 3D videos, or the blockchain, or the VR glasses, and so forth. I mean I am thinking what can they be useful for. Even if these AI can produce a lot of cool images and text, I mean what is it useful for? Cheating on your exams? Sure. Creating more spam-bot content to fill up websites quickly? Sure. For fun for a few days to generate random shit? Sure.
What else?
Maybe the only two useful things would be:
What else?
Maybe the only two useful things would be:
- Coding and technical answers. So things that you can check. I tried using AI to help me push commits to the AUR, and was mostly helpful. Providing command line tips and all that. But since it is at times inaccurate you can also mess things up. On top of this if you use them to create pieces of code can be decent from my tests, but still primitive in many regards.
- Quick unimportant answers. Like recipes, or whatever else you may need that does no have to be that factual.
like this
muppeth
in reply to Tio • • •As for usage of current AI stuff. For example in gaming, AI could be creating in realtime dialog lines. Worlds could dynamically change depending on your actions in the game thanks to realtime analyzing and adjusting the story/world. For general use, from coding, designing, writing, you could use AI to create things for you by just talking to it and describing what you need. We are currently at the early stages of things. So surely current AI’s can’t create complex code etc and are prone to failure but things are improving really fast. It does not mean you woldn’t ne... show more
As for usage of current AI stuff. For example in gaming, AI could be creating in realtime dialog lines. Worlds could dynamically change depending on your actions in the game thanks to realtime analyzing and adjusting the story/world. For general use, from coding, designing, writing, you could use AI to create things for you by just talking to it and describing what you need. We are currently at the early stages of things. So surely current AI’s can’t create complex code etc and are prone to failure but things are improving really fast. It does not mean you woldn’t need specialists anymore, but such tool in hand of a specialist could allow him to be fast, more accurate and generally better at their craft and for avarage folk, it will allow them to materialize their ideas/thoughts without prior skills needed to be acquired.
muppeth
in reply to muppeth • • •Tio
in reply to muppeth • •muppeth
in reply to Tio • • •Tio
in reply to muppeth • •Also don't forget that these bots will allow people to quickly create "original" content, which means those SEO vampires will create even more shitty websites/content and much faster. I know people whose entire job is just that. And then in the future these bots will feed from their own vomit 😁. A positive feedback loop.
muppeth
in reply to Tio • • •Tio
in reply to muppeth • •Tio
in reply to muppeth • •AI requires a lot of resources and data, to produce nothing of true original value. By that I mean sure, an AI can create some cool poems, and images, but since they are AI generated the more you create and the easier, the less their value. Would you read a book that was created with such a piece of software? I personally would never.
I can see some uses when this type of software acts like a translator, not of words, but actions. Such as maybe I know how to design websites but not to code them, so it can code a simple one for me. Even there it will... show more
AI requires a lot of resources and data, to produce nothing of true original value. By that I mean sure, an AI can create some cool poems, and images, but since they are AI generated the more you create and the easier, the less their value. Would you read a book that was created with such a piece of software? I personally would never.
I can see some uses when this type of software acts like a translator, not of words, but actions. Such as maybe I know how to design websites but not to code them, so it can code a simple one for me. Even there it will become a mess the more you rely on such systems.
Use in games, idk, maybe that could be feasible. I don't see that as useful personally. This still looks to me like more hype than revolutionary. Will see....
muppeth
in reply to Tio • • •Eg. Ubisoft is already working on implementing AI for dialog creation. So instead of few lines or repeatable text from NPC's you could have proper discussions. Add to it AI generating voice and facial expressions (already exists) and you can have discuission ingame with any character you meet on your way.
I dont think the idea of AI is that it will write books or make music and we will all love it. It's not what it's meant to be IMO. It's another tool for us to make use to make cooler shit.
Tio
in reply to muppeth • •Where are all of these now? I felt for many of these, I wrote books about these, and I feel tricked about them now. There is some value in all of them, and usefulness, but the hype created by the companies selling them, does not match the reality at all.
Time will tell with these AI's. My feeling now is that they are yet another hype, and will be only u... show more
Where are all of these now? I felt for many of these, I wrote books about these, and I feel tricked about them now. There is some value in all of them, and usefulness, but the hype created by the companies selling them, does not match the reality at all.
Time will tell with these AI's. My feeling now is that they are yet another hype, and will be only useful in certain narrow domains.
muppeth
in reply to Tio • • •Sure it did. Maybe not what mainstream media was promising riding a hype train but I assure you prototyping has never been easier, faster and cheaper.
VR glasses also are quite widely used but not for casual everyday use. It's technology that needs still lots of work but it does make significant impact in engineering na dproduct design.
Self driving Musk promise I put together with 3d movie hype. But all others are pretty much relevant and shaping your world too. Same as AI. If you expect AI to do your laundry, write a book and compose music like what media tries to sell not understanding what AI is, then surely you are right. AI will however be implemented in about everyday use for universe of things. As end user you might not even realize it as it won't be front facing but the potential for technology is huge and it will have significant impact in few years. We are just scratching the surface here.
F... show more
Sure it did. Maybe not what mainstream media was promising riding a hype train but I assure you prototyping has never been easier, faster and cheaper.
VR glasses also are quite widely used but not for casual everyday use. It's technology that needs still lots of work but it does make significant impact in engineering na dproduct design.
Self driving Musk promise I put together with 3d movie hype. But all others are pretty much relevant and shaping your world too. Same as AI. If you expect AI to do your laundry, write a book and compose music like what media tries to sell not understanding what AI is, then surely you are right. AI will however be implemented in about everyday use for universe of things. As end user you might not even realize it as it won't be front facing but the potential for technology is huge and it will have significant impact in few years. We are just scratching the surface here.
For me the question is whether to jump on the train while you still can to:
1. Stay relevant.
2. Participate in the shaping of new reality and try to prevent the evils of the world to ruin it by co-creating it. To prevent situation where we play never ending chasing game (like with big tech social media race).
Or to slowly get familiar with the fact world has gone off, and you decided to skip the train. This means staying with outdated tech and having fun with it, not really being interested in the current stuff. This is what happened to our parents and their parents and I think we are witnessing this for our generation too.
Where do I stand, I dont know yet myself, but I can't ignore the fact this is quite a breakthrough technology that will drive the change for years to come.
Tio
in reply to muppeth • •Well they were hyped for years and in a credible manner I'd say. Like IBM Watson and such. Regardless my point was that compared to what most media projected about these tech, the reality of how they ended up is very very different. And I suspect the same with these pieces of software that are neither artificial or intelligent, nor that we know what intelligence means.
They are very hyped-up by the companies that sell them, which is normal. For me the way to deal with this influx of information is to learn how they work to be able to appreciate them. And from my readings and courses that I took, they are less impressive than the hype around them. They are regurgitating data statistical pieces of software. Apply filters to them, set goals, rinse and repeat, and you will get some interesting results. The way chatgpt works, from my understanding, is based on what word statistically goes with what other. So it is good at creating sentences that make sense... show more
Well they were hyped for years and in a credible manner I'd say. Like IBM Watson and such. Regardless my point was that compared to what most media projected about these tech, the reality of how they ended up is very very different. And I suspect the same with these pieces of software that are neither artificial or intelligent, nor that we know what intelligence means.
They are very hyped-up by the companies that sell them, which is normal. For me the way to deal with this influx of information is to learn how they work to be able to appreciate them. And from my readings and courses that I took, they are less impressive than the hype around them. They are regurgitating data statistical pieces of software. Apply filters to them, set goals, rinse and repeat, and you will get some interesting results. The way chatgpt works, from my understanding, is based on what word statistically goes with what other. So it is good at creating sentences that make sense, despite them being incorrect at times.
So yea, in order for us to have any views about this "AI" is to learn how it works. Plus, it is not "a thing" it is many things, many of which we even use today, from text prediction to the normal searches that take into account your history, and so forth.
muppeth
in reply to Tio • • •Isaac Kuo
in reply to Tio • • •In particular, I'm thinking of tabletop style role playing games where a Game Master can quickly whip up some background art for a room or other setting. The AI could offer, say, a dozen images and if none of them are satisfactory it's no big deal - just fall back onto the normal role playing technique of no artwork.
There are ethical issues of the raw input data set, but I see this as solvable in a straightforward way - limit input data to public domain images (older art and various public domain photos etc).