• Guest, you're invited to help build our new TBT time capsule! It contains three parts, with some of its elements planned to open in 2029 and others not until the distant future of 2034. Get started in 2024 Community Time Capsule: Blueprints.

Have you seen more AI?

OswinOswald

Senior Member
Joined
Jan 9, 2015
Posts
1,521
Bells
289
Carnival Coins
0
Switch
1920-3285-9320
RAFFLE: Sunday's Loyalty Bounty
Old Gameboy
Arcade Scenery
Goomba
Invader
Smash Ball
Pinky
White Crescent Moon
Yellow Crescent Moon
Pink Crescent Moon
Green Crescent Moon
Blue Crescent Moon
There was a recent flurry of discussion about AI with ChatGPT and those photo generation AIs, but I would argue that AI has been around for a long time and has been used for a long time. Like the Facebook algorithm is a form of AI that is learning from all our clocks and what our eyes are watching. I read a cool article about an AI in Japan that takes a picture of the bread you get at the bakery and figures out how much it costs, based on the different shapes of the rolls, etc. And it is supposedly quicker and more accurate than a human ringing you up.
Recently at my work (at a hospital), we started testing out an AI scribe. It's like a dictation service but it summarizes off of conversational speech. It seems like it's saving me a lot of time when it comes to paperwork and note taking, although I definitely have to go double check and edit some things.
So where else have people been seeing AI/computer learning in their lives? Is it all scary? Or is some of it making your lives better!
 
One time when I was at the mall I saw this café kiosk where they had a robot that could make espresso or something like that. I didn’t order from them but would like to sometime.

Other than that I can’t remember any other interesting encounters I’ve had with AI. I guess I could speak about Pokémon Violet but I chose Scarlet as my version so it doesn’t count. And whenever I think about AI I tend to think about the stock market because one of my parents likes to watch videos about the stock market on TV and the videos are usually either about Tesla and Elon Musk or AI in general.
 
i’ve seen a lot of ai on tiktok like ai song covers it’s crazy how it can replicate a specific artists voice so accurately (but sometimes the original voice in the song the ai is covering outshines it lol)
 
Well I use ChatGPT at work and for school, but only because it's so easy to find information with it.

I was looking for a specific Excel. Google just showed me Microsoft sites that weren't completey related to my issue. ChatGPT immediately knew what I wanted, showed me the Excel formula I needed and explained it.

Mostly I just use it to give me ideas.
 
There are lots of websites that can automatically detect the background of an image and crop it out really well! And there's stuff like fingerprint detectors on phones, Google Image Search, Cortana/Siri/Alexa and other search assistants with voice recognition.
 
The AI song covers of cartoon characters have been fun as well as the presidents play video games trend on youtube. That said, I'm nervous about AI and what will happen when it becomes just as good as humans. Will it replace people's jobs forever? Kinda worried in that regard.
 
Yes. I'm beginning to think though that @Dunny County was right when he told me that producing decent AI art is an art form in itself though. When he does it the results are beautiful and when I do it the results are like this:

4a9c0ffbafe74ba29800f1ab154664a2.png

I asked it to draw an Eastenders character kicking a postman over a wall and this happened. I was very confused!
 
Yes, it's all about getting good with the 'prompt' - which you failed to do with this bizarre nightmare.

I love AI though, I especially love AI art in it's little baby stage, getting bigger and stronger every day.
 
Yes, and it's driving me nuts on how often I'm hearing that term. For example, a company advertising a product and make it blatantly obvious by tacking on "with AI!!!" in its description. News flash: I don't really care. I used to associate AI with video games only, but now it's got to a point where I'm annoyed hearing the term every day.
Also, random thought, but instead of wasting $70k on a useless bachelors at a university, why not have some random AI bot teach me critical facts and I check them? Seems much simpler than having some old dude talk for four hours straight a day, gives you no free time, and expect you to know everything by the exam in which you MUST pass with an A otherwise you'll receive academic expulsion.
 
I love technology and I'm always excited to see new developments in it, especially in the fields of medicine and biotechnology (my professional background). The most exciting thing I've seen recently I'm not sure I'm allowed to speak about yet.

Artificial intelligence has a place in the world. The area of AI I'm most interested in outside of my field is companion robots. Specifically, the ethical and social implications companion robots could have on our society. To give an example of what I mean, here's an excellent summary ChatGPT produced when I asked it about the ethical concerns of companion robots in elderly care:

ChatGPT said:
You
What are the ethical concerns of artificial intelligence in companion robots designed to care for the elderly


ChatGPT
Artificial intelligence (AI) in companion robots designed to care for the elderly raises several ethical concerns, given the vulnerable nature of the population being served. Here are some of the key ethical concerns:
  1. Autonomy and Decision-making: Companion robots may be tasked with making decisions on behalf of elderly individuals, such as managing their medication schedules or determining when to call for emergency assistance. Ethical questions arise regarding who should have the authority to make such decisions and how to ensure that the autonomy and preferences of the elderly person are respected.
  2. Privacy and Dignity: Companion robots often collect sensitive personal data about their elderly users, such as health information and daily routines. There's a concern about maintaining the privacy and dignity of the elderly individuals, particularly regarding how their data is collected, stored, and shared. Proper safeguards must be in place to protect their privacy rights.
  3. Quality of Care: While companion robots can provide valuable assistance and support, there's a risk that they may not always deliver high-quality care, especially in complex or unpredictable situations. Ethical considerations include ensuring that the elderly receive adequate care and support and that the robots are equipped to handle various caregiving tasks effectively.
  4. Emotional Well-being: Companion robots designed to provide emotional support to the elderly raise questions about the authenticity of the interactions and the potential for emotional manipulation. Ethical concerns include the need for transparency about the robot's capabilities and limitations and ensuring that interactions with the robot do not replace meaningful human connections.
  5. Dependency and Isolation: There's a concern that elderly individuals may become overly dependent on companion robots for companionship and assistance, leading to a reduction in human-to-human interactions and feelings of isolation. Ethical considerations include promoting social connectedness and ensuring that the use of companion robots complements rather than replaces human care and support.
  6. Equity and Access: There's a risk that companion robots designed for the elderly may exacerbate existing socioeconomic inequalities if access to these technologies is limited to those who can afford them. Ethical considerations include ensuring equitable access to companion robots and addressing disparities in healthcare and social services for the elderly.
  7. End-of-life Care: Companion robots may be involved in providing care and support to elderly individuals nearing the end of their lives. Ethical considerations include respecting the wishes and preferences of the elderly person regarding end-of-life care decisions and ensuring that the use of companion robots does not compromise their comfort, dignity, or quality of life.
Addressing these ethical concerns requires careful consideration of the values and preferences of elderly individuals, as well as collaboration among stakeholders, including caregivers, healthcare professionals, technologists, ethicists, and policymakers, to ensure that companion robots designed for the elderly uphold ethical principles and promote the well-being of their users.

The above is also a good example of the type of thing I use ChatGPT for: quick summaries of a topic. For example, the last thing I used ChatGPT before this was to provide a summary of the events of the Battle of Chibi (208 AD). Not all information ChatGPT produces is correct, but when you're interested in a topic and aren't sure where to start it's a good way to get an idea of what it is you need to research further.
 
There are big broad philosophical questions surrounding AI when it comes to art I suppose. Does art come from the soul, from the hand or from a combination of both? Could there have been fantastical dreamers throughout history whose artistic visions would have superceded those of the great masters but they never had the required physical dexterity to create them in physical form?

Could this become a new form of human creatitivity, with AI merely as a tool to replace the brush? It'll only be a few more years before the ability of AI to accurately and exactly represent whatever you imagined will be reality - it wont stop at images, it'll become movies, it'll become games, it'll be pretty much every form of art there is. But who will be the 'artist'? You or the computer? That would depend - what if you envisaged, pictured and then described every single facet that comes out the other end? An entire movie for example, line by line, scene by scene, shot by shot - and it was then immediately produced by the computer, precisely as you imagined it, nothing additional added - straight from brain to screen. ...Who was the artist then?

Could the argument be made that we're on the verge of a completely new artistic revolution - a more literal one than ever before - with the artistic field being levelled for the first time in history and the greatest dreamers rising to the top?

I like to think so. But then it might also just be loads of people going "create me as an anime character" and the finished products being displayed in The Louvre. 🤷‍♂️
 
There are big broad philosophical questions surrounding AI when it comes to art I suppose. Does art come from the soul, from the hand or from a combination of both? Could there have been fantastical dreamers throughout history whose artistic visions would have superceded those of the great masters but they never had the required physical dexterity to create them in physical form?

Could this become a new form of human creatitivity, with AI merely as a tool to replace the brush? It'll only be a few more years before the ability of AI to accurately and exactly represent whatever you imagined will be reality - it wont stop at images, it'll become movies, it'll become games, it'll be pretty much every form of art there is. But who will be the 'artist'? You or the computer? That would depend - what if you envisaged, pictured and then described every single facet that comes out the other end? An entire movie for example, line by line, scene by scene, shot by shot - and it was then immediately produced by the computer, precisely as you imagined it, nothing additional added - straight from brain to screen. ...Who was the artist then?

Could the argument be made that we're on the verge of a completely new artistic revolution - a more literal one than ever before - with the artistic field being levelled for the first time in history and the greatest dreamers rising to the top?

I like to think so. But then it might also just be loads of people going "create me as an anime character" and the finished products being displayed in The Louvre. 🤷‍♂️
At first I was like "What in the nerd is this?" :geek:
But then I lol'd at that last line, so you're good.😎
 
I was incredibly annoyed to learn recently that a local, highly esteemed art competition has decided to allow AI art in their contest this year.. and it's not even in its own category??? So odd. And the rules require you to have a live subject, so... how would that even work lmao? Describing the subject to the program? Bizarre. Oh, and keep in mind that the prizes are thousands of dollars.. yeah.

Don't even get me started on my thoughts on AI art in general.

I'm sure there are lots of ways that AI can be useful, but it seems to be hit or miss right now. Like I use the self checkouts a lot at the grocery store and they have TWO cameras watching me scan, and sometimes the AI thinks I put something in the bag without scanning it but I did! I'm really not out here trying to steal one red capsicum, Woolworths. Geez

AI kinda seems like it's starting to become a buzzword, too. Every company seems to be trying to integrate some kind of AI into it. I wonder if AI really will get better; if there will be some kind of AI renaissance that will take us into the next era of technological advancement, or if this is simply just a fad of the 2020's.

Anyway sorry for the long rant, every now and then I guess I gotta get my AI grievances out.
 
I was incredibly annoyed to learn recently that a local, highly esteemed art competition has decided to allow AI art in their contest this year.. and it's not even in its own category??? So odd. And the rules require you to have a live subject, so... how would that even work lmao? Describing the subject to the program? Bizarre. Oh, and keep in mind that the prizes are thousands of dollars.. yeah.

Don't even get me started on my thoughts on AI art in general.

I'm sure there are lots of ways that AI can be useful, but it seems to be hit or miss right now. Like I use the self checkouts a lot at the grocery store and they have TWO cameras watching me scan, and sometimes the AI thinks I put something in the bag without scanning it but I did! I'm really not out here trying to steal one red capsicum, Woolworths. Geez

AI kinda seems like it's starting to become a buzzword, too. Every company seems to be trying to integrate some kind of AI into it. I wonder if AI really will get better; if there will be some kind of AI renaissance that will take us into the next era of technological advancement, or if this is simply just a fad of the 2020's.

Anyway sorry for the long rant, every now and then I guess I gotta get my AI grievances out.
So what you're saying is, I should submit my masterpiece to that contest and win thousands of dollars 💰

I hope it's still thousands of pounds when I exchange it for sterling 🇬🇧
 
I have seen a lot more AI, and my general feelings towards it are... complicated? As someone who's in both creative spaces and who is in computer science.

I think the specific types of AI need to be differentiated better when talking about it, though I also understand that's probably impossible. I think when most people are talking about this type of topic, they're talking about generative AIs and learning models (both of which take training data and try to spit out some result).

AI itself is a bit broad, as it covers things from the AI of characters in games (which goes down to thing as simple as pathing or things like what actions enemies take in games): that in itself has been around for a long time and is harmless and mostly just algorithmic. And then there's thinks like automatic closed captioning or real-time translations, which is certainly helpful in terms of accessibility, and in terms of art, smoothing algorithms and the likes which are also fine.

When it comes to things like generative AI/learning models: I think it can be helpful and has its uses, but the issues I have come from the ethics and misuse of it and so I'm just generally very much against it. When used properly, I think it's great as an assistive technology. However, for things like AI art, it's fundamentally flawed: the training data consists of stolen art, and the intention isn't usually to help artists but to replace them -- there's a lot of artists I've seen who have their art stolen and fed into AI. The intent of this wasn't to help the artist in any way, but as a sort of "You can use this instead of commissioning them". This use case is insanely ethically wrong to me, as it's directly harming an individual, and most uses of genAI for art seem to be for similar scenarios. But this tends to be where the majority of its use goes for art, so as a whole: I really hate AI art. I also hate how I can't talk about "AI" in other regards without people thinking I support AI art lmao.

The "ethical" way of using genAI for art would be: someone trains their own model with their own style and use it to handle the tasks that they want to simplify the process of (e.g. if lineart is something that someone finds very time consuming, then using it so help generate the lineart and then cleaning it up manually might be a better way). In an industrial setting, it would be using this to reduce the burden on artists for tasks like... removing stray hairs in movie frames and the likes (which is just an example I pulled having had a friend do a role and cry about having been stuck in a dark room for hours doing cleanup like this -- as an artist, it wasn't creatively fulfilling at all and it was maddening for them). Tasks that aren't meant to replace creatives, but rather to help reduce repetitive work (that may be important but not something they want to physically suffer for).

Unfortunately, with the way things have been released, it's largely impossible to achieve this. It's incredibly hard to enforce that tools are used in an ethical manner, and to enforce that the training data was ethically sourced. The premise of it is flawed, so the tool as a whole is dead to me and I wish there were some way to shove it back in its pandora's box.

When it comes to things like generative AI for text (ChatGPT and the likes): I'm of a mixed opinion. I think there's a large subset of users who will take the generated answers assuming they're always accurate, which is a large problem and comes down to irresponsible use. But I think it can be helpful if someone needs a concept explained to them in an interactive way but they don't have the time to research things fully on their own or the resources to do so.
 
I was at a restaurant with no server. Food and Drinks were entered through buttons and delivered by a robot on wheels. A real person could still be alerted for assistance.

It was a novelty and I've never seen it anywhere else. I'm not sure how I would feel if that were the norm.

Screenshot 2024-03-20 12.23.12 AM.png


It resembled this a little bit.
 
Back
Top