Ai Uses

I think as artists we are looking at such a small part of what AI is, and may become, and forget that it is already being used for other purposes. Medicine and research is one of the areas I am sure will continue to be benefited by this growing technology. I probably should say, hopefully without the dumbing down of the human race.

I was disappointed that David closed that thread.

I have already found using Ai for help generating descriptions fascinating, if not worthwhile. And of course the imagery has been wonderful, if not exactly perfect, but I am warming up to it again. For now, it is a fun place to "start."

I hinted in David's thread that I would like to see an image to text generator. Some have said, they'd rather not beat their brains trying to write a description. Especially for an abstract. I agree, I'd rather make art, not try to describe it but some of you are excellent storytellers. I am not.

Not to be negative, but as Val hints, and I quipped in the 6 Word Story thread, the other side of the coin is "I found an alternative to thinking."

Accurate and timely medical diagnosis. My guess is it's already being used this way to some degree.

Oh yeah... autonomous driving. Like good autonomous driving where cars and traffic systems are networked together. Not the current stuff on the market. One day we won't need steering wheels in cars.

I'm a big fan of self-driving cars. Especially for large and busy cities :-)

I regularly watch videos on YouTube where people let their car drive (mainly Tesla FSD). Tesla operates a really large data center for their AI projects.

Yes, I enjoy driving myself but not all the time. I can't wait to have such a car one day!

I think the major problem with autonomous driving currently is that it's mixed in with human drivers and roads that just aren't designed around it. I believe once roads and highways are built for it and cars are on the same page, communicating with each other (i.e. negotiating position, speed, etc), it will be vastly improved. Probably not in my lifetime though. People love their cars and I see fear politics playing a huge part is slowing down the entire process.

You kind of see it already. Every time a Tesla crashes, naysayers are gleefully posting about it, like they're so very happy about it for some reason.

When I first started to list here on FAA, I was overwhelmed and stressed out by having to write descriptions.
I asked a lot of questions and some folks gave me some great suggestions and tips.
Even with all that information, it's still became overwhelming, and I found myself not listing as many things because of how stressed out I got every time I had to look at the descriptions and create something that made sense and please google.

As a person who has been struggling with dyslexia, I have figured out ways to get by, but it is still never easy for me when it comes to writing. Even with my dictation program, my subscription to Grammarly. With all of these things.

AI: ChatGPT is a game changer for me. Although it is not perfect, and I still have to edit and start with my own basic idea of what I want to talk about, it helps me rearrange the words, so they make sense and gives me some words that I would never have come up with on my own based on my limitations.

I am extremely grateful to David for starting the conversation about ChatGPT!

Perhaps the problem is we humans, not so much technology. Technology can be used for good or for bad.

Yes, did you see the CNN report yesterday? Surprisingling they were very supportive it, saying they have their own data bases. AI was so accurate even in medical schooling exams. Actually really scary to me that young people could be whizzing through to their medical degree and didn't have to gain the knowledge at all. Just ask AI. I don't want that doctor. But then on the other hand, if it's that accurate, just get an account and ask it to make your medical diagnosis.

I would like to believe that AI will free us as Humans to:

REACH FOR THE STARS!!

I find more and more that Alexa is becoming my best friend. As I get older my memory of trivial things gets worse and all I have to do is tell Alexa to remind me to turn on the dishwasher or turn off the oven or to call my sister on her birthday and so many other things. Not to mention she amuses me when I ask her to do things like recite limericks or tell me silly jokes. Just asking her to spell certain things it makes me burst into laughter. Try giving her words like "Supercalifragilisticexpialidocious"

I also enjoy that she can operate my iRobot and turn my lights on and off even if I am not home. I call the robot Betty and I just have to say to Alexa that I think it is time to send Betty to work and before I know it that little machine is wandering around the living room.

There are uses for it, like sorting out medicine, that idea that it can see if a person said two different things in legal cases. Finding patterns to disease, and the like. Maybe a smarter search. I don't think the writing is there yet and don't get me started on the art part...

It has its uses. I don't trust and won't ever trust it in a car. But to give me the right amount of soda or ice cream that's fine. I think its a bad idea to trust it to do all coding, that's a recipe for disaster. Smart cooking to cook something just right would be nice to have. I thought it was weird that it got answers wrong on a medical test it took. But to see common patterns to things and thinking up solutions that don't involve internet ideas, but actual medical texts, that may be a good thing.

And I guess add a better brain to those silicone love doll things so people can do more than well, you know with it, and they can talk about their day and complain about coworkers and such.


----Mike Savad

I got a kick out of Leslie's description of her iRobot, Betty. There is some AI built in of for sure, like knowing when to re-do an area with extra grime. I operate our robot with my phone - but will now (thanks, Leslie) figure out the Alexa routines. LOL

Self-driving cars have fascinated me for a long time but I don't think most of the motoring public is ready. But who knows. right? Medicine, for sure, is a big win with AI and I'm sure it will only get better.

But one of my long-time fascinations has always been astronomy and outer space. I can see AI being used to explore where no man has gone before - without putting real humans in harm's way.

I would love to have a self driving car, but there would have to be some really heavy convincing to show me its safe. Like show me what it does on ice or snow, or just like a low lying truck, or a child in the street. I hate driving, don't do highways. If it can drive me some place, that's great. But I don't think we are really ready yet, even though there are cars that can do it on its own more or less.


----Mike Savad

I wrote a whole post and realised i was not logged in any more

I want a self driving car. Very much. But i also want to race Audi's on the country roads like i do now, so either I'd have to have rally driving mode, or manual takeover.

I have a semi retired robot vacuum. Two Shepweillers mean too much hair for the poor little thing.

I just don't want to be retired off myself. Too early. Unless they pay me to live where i want, read, paint, garden, and have lots of animals, whilst ai does everything else.

 

Companion robots that will appear to be listening, just like their human counterparts in the real world.   ;)

They'll smile, nod, or as appropriate, frown, pout, widen eyes in amazement, stifle their yawns, etc.

Their body movements to be the paradigm of empathetic, consoling/celebratory/shocked (etc.) gestures.

AI, AI — Oh!

Definitely meant to be a positive, as studies are ongoing to vet robots in seniors' homes already— the lonliness in many instances is palpable.

AI will reduce the number of tasks a human has to do: that is not the same as reducing the number of jobs available.
If someone replaces you in your job, it is because s/he knows about AI!

AI will allow us to render the minutiae of "everyday mundane reality" a thing of the past; so we can concentrate
on the vulnerable, the sick, the impoverished, the uneducated etc etc

AI has great social possibilities - no doubt a better future all round!

I've heard too many stories of how the car didn't think the truck was an obstacle, and it cut people in two. Videos of children (dummies), being hit. And one guy proving that it will stop for a kid, using his own kid as a dummy. It did luckily stop. Can't imagine what pot holes, snow etc would do. I'll let the rest test that out before I ever use it. Not sure if I would even trust it to park.

Those little AI vacs.. How good a job does a randomly controlled thing do? I've seen too many videos of it running over the dog doo. Then spreading it all over the house in comical brown tracks.


----Mike Savad

In 8½ years I'll be 100 years old... The reason I intend to live indefinitely is because I can't wait to see what's next in technology.

And I promised myself decades ago that I'm not going ANYWHERE until teleportation is a common means of travel.

@Jason and @Matthias - I think self-driving would improve SO much on the roads, if ALL cars were self driving and all the infrastructure was networked together. That includes the cars, all traffic lights, stop signs, yield signs, etc. In fact, if it was done properly, traffic lights wouldn't even be necessary. The system would know that if it slowed Car A down by 0.7614kmph and sped Car B up by 0.19762kmph, then both could cross through an intersection without having to stop at a light and without hitting each other. Extend that to the thousands or millions of cars on the road, and it would be able to adjust all of their speeds up or down by minute amounts to keep all the traffic flowing. But I think we're still a long way away from that.

@David - how are you using AI to generate descriptions of images? I'm looking for something where I can either upload an image or give it a direct link to one, and have it generate a description. So far, I still have to type stuff, something like "generate a description of this image that shows a waterfall cascading down a cliff in winter"... in other words, I have to describe what I want it to describe.

Restoring, conserving, and preserving our cultural artifacts. How many sites are damaged or destroyed due to things like a storm or a fire? Things happen to irreplaceable items, from buildings to paintings to audio tapes. It seems like conservators, restorers, and archaeologists would welcome the assistance!

How about helping us to find sources of information? I cannot reliably find information by searching the internet using search engines, but I can ask ChatGPT for references to information. They key is to ask for information, not the answer. And remember, ChatGPT is trained on the internet. It's answers are as truthful as we humans are.

I have another: diffusing bad situations! What if you could ask for differing points of view, and get as many as can be found? People trust computers more than other people, maybe it would stop some of the anger when a computer points out how many opinions there are on a subject other than your own.

I saw somewhere that there is a system that people just talk to. No answers, no questions, just to talk to. How would that help? Think about recent events! Illness that prevents contact with other humans, traveling to other planets with only a few people... Our mental health is just as important as our physical health.

Theoretical science: there is much we do not know. When things are discovered it's sometimes "wow, I wouldn't have thought that!" Have a problem? That's where the bizarre responses would be useful! Ask for all possibilities, and then look at them and see what we can find out about them. A lot of theoretical science is based on guesswork, and sometimes really strange guesses can lead to a breakthrough.

AI could reduce the need for living experiment subjects, if it was sufficiently advanced. Imagine if we could model the effects of a neurosurgical operation on an AI "brain" before it's tried on an actual person (not with what we have now, but in the future).

I could use an AI thesaurus too.

Mike - there's now a robotic vac that detects and avoids dog poo - problem solved. Now they just need to make one that can get itself up the stairs and clean in corners properly, which I'm sure will happen at some point.

ChatGPT, I have discovered, is great at creating keywords for images. Great time saver!

I guess I already have an AI thesaurus!

@John -

I have (upon Richard Reeves suggestion) prompted ChatGPT to describe my image by directly feeding it the image URL. While it gave me back a description, it wasn't really looking at the image itself but extrapolating from the URL. I think. Anyway it didn't really give me what I'd hoped for. So, yes you have to describe your image to get a description. Sort of like - garbage in garbage out, if you remember that old axiom.

I also asked ChatGPT if there are image to text generators. Here's what I got back:


Is there an image to text generator?

Yes, there are various Image to Text (OCR) generators available both as free and paid tools. Some of the popular ones include Adobe Acrobat, ABBYY FineReader, Readiris, OCR Terminal, and Google Drive. These tools use optical character recognition (OCR) technology to extract text from images and convert it into editable text format.


Do you mean image captioning?

No, I was referring to Optical Character Recognition (OCR), which is a technology used to convert scanned images or pictures of text into editable text.

Image captioning, on the other hand, is a field of computer vision and natural language processing that focuses on automatically generating a textual description of an image. It involves analyzing an image and producing a corresponding textual description, often in the form of a complete sentence. There are various Image Captioning models available, both pre-trained and custom, that can be used to generate captions for images.

I've yet to test any of this further.


And by the way, I'm anxious to try out a self-driving car. Can I rent a Tesla for that?


When I think of other Ai uses, I think about the movies on the subject - such as "Artifical Intelligence" "Blade Runner" "I,Robot" "Terminator" and of course "2001-A Space Odyssey" among others...

Buckle Up

Ah yes, "2001-A Space Odyssey"... I remember it well.

I see what you did there Dave....






"Open the pod bay doors, HAL."

"I'm sorry, Dave. I'm afraid I can't do that."

@James B
from the article
"programs exist that can tell you that a play is written in a style closer to a particular playwright’s than to those of hundreds of his peers."

Now what happens if I tell AI to write a play "in the style of Lorca" and place it somewhere strategic, will another AI "discover" it as a new Lorca play?

GOOD POINT, JAMES. The clues would be there. hmmmmmm !

USE FOR Al ? ::::::::
The Entertainment Director, here at my 'residence', visits me before his 'rounds' at the Group and Al came up.
He signed up for $50 AU, sent 10 photo/pix of himself, and that Al company produced 10 different Al pix of him.
So, there's a 'use' for it: get yourslef 10 different ID's......! the mind boggles. Scammers who prey on people,
would love 10 different IDs. and the Al ? it just rearranges,etc., yet leaves a resemblance: I saw all 10. This
is NOTHING compared to what Ai can also do, rearranging.

I just the movie "Megan" with my son for family time. A bit of AI horror is always good entertainment.

Statement 2 by Doug is the most important

Try unplugging the internet. Can't be done now.
Maybe at a personal level but even that, try conducting business on a day to day basis.

It's best to leverage AI now. For me, AI is this long anticipated arrival of a horse of no color. Yes, it's a figurative concept. It's not the horse to be feared. It can't be allowed to be ridden by any one individual.

Fear is not the way; knowledge is a gift and it can not be filtered in such a way to manipulate the masses to any single individual's ideas of a conceived utopian society.

Rather than allowing ourselves to be lead by the proverbial wizard behind the curtain, strip the wizard naked and expose anyone who attempts to turn knowledge into a commodity to be cornered.

Is AI a turning point?

I could plug this into any one of at least 3 threads currently running about Ai. And, it encompasses many of the elements of discussion going on in all of them. Perhaps as well as Abbie's Image and Literature discussion, regarding writing a story.

I chose this one because we're hypothetically speaking.

If inappropriate, please delete (or ask me to and I will). But, to me, this is the FUN part about Ai. Too much seriousness is a downer.

I dreamt that in the not-so-distant future, an exoskeleton has been developed that learns "your" body. It doubles, triples, quadruples your strength. It protects you from the elements and works in hostile environments. Yet it adjusts daily to your needs and wants. Is easy to put on and take off. It varies the degree of physical input "resistance" to develop your level of strength allowing you to become stronger over time so that you don't wither away because the machine is doing all the work. Particularly nice for those who are handicapped or catastrophically injured (like Robocop) but is a much more refined and nimbler machine. In some circumstances it can be worn indefinitely, and it takes care of your biological needs. All of them. It learns from you, and you learn from it. You are connected to it. It senses you. Eventually you both work as one. And still, you can take it off at night and give it a break. All enhanced by Ai of course.

I also asked ChatGPT to expand on this, here's what it gave me:

An exoskeleton that is designed to be customized to the user's body, providing numerous benefits to the wearer. It enhances strength, providing the user with a 4x increase in strength, allowing them to perform physical tasks with greater ease. The exoskeleton protects against harsh environments, insulating the wearer from extreme temperatures and other environmental factors, making it ideal for use in hazardous work conditions. The exoskeleton adjusts to the wearer's needs and wants on a daily basis, ensuring that the user remains comfortable and safe at all times. It is designed to be easy to put on and take off, and it gradually increases the degree of resistance to physical input, helping the user to build their strength over time and avoid muscle atrophy. The exoskeleton is especially useful for people who are handicapped or who have suffered catastrophic injuries, providing them with a level of mobility and physical ability that they may not have otherwise. The exoskeleton is designed to be worn indefinitely in some circumstances and provides the wearer with all their biological needs, including the ability to learn from and connect with the exoskeleton.

I used a shortened version as a prompt to generate an Ai image. I was given many great alternatives that would fit this story, but since this is not an image thread, here is a link:

https://fineartamerica.com/featured/the-ultimate-body-david-manlove.html

Of course, this technology is already here and the refinements on our doorstep.

I wondered when J.A.R.V.I.S. would make an appearance

Now google has one called bard, and china came out with one too. Weird how they all came out at the same time. Its like the AI that created the AI wanted it that way.


----Mike Savad

The main worry I guess is that people will get complacent with the search. Forget how to do any research. And decide that if it came from an AI its telling the truth, that is the answer, doesn't matter what it feeds you or who. Right now people are getting facts from memes, this is the next step to that. Pretty sure this is step one to world domination.

I wonder if its smart enough to write complete software. One guy made a pluggin for wordpress that did a certain thing. Could I make a better AI, something that added spy code, then tell it to copy over itself? I wonder how much autonomy it has?


----Mike Savad

Matrix anyone?

There is a possible chance that we are in the matrix already. Just look at tik tok for evidence. There are people trying to prove that we are in the matrix by some how breaking it. Not sure how.


----Mike Savad

AI is scary in how well and fast, it creates art on command. It has the potential to decimate art markets if people are silly enough to gravitate toward it. I'm of the opinion that most folks who truly love art will seek out real artists and avoid AI-created pieces. I think where AI will be big is with advertising and marketing. That sort of thing. The ability to create a visual and fast from an idea would be very attractive to those groups.

@sean, except for the possible copyright violations a company could get if they used AI art. But that's off topic and discussed to death. Mostly.


----Mike Savad

I just had another AI application ideas….

Back during Covid they started giving every student at my son’s school a Chromebook. What if every student was issued an AI “tutor”?

A little robot that could help them with their courses? The school systems complain a lot about not having enough help. This could greatly help. It would be like having a teacher’s aid for each kid.

What would be super cool, and i think that nvidia made something about that, i'd like to see AI look at a scene, then ID everything in the scene up the amount of words needed, putting important words up front. Like it would have to know what the scene is about and list all the words and alt words I could use. That is a good use for AI. And I suppose make a story about the scene itself. But it would really have to be smart.


----Mike Savad

Well, Brian, maybe I could actually learn some maths with my dyscalculia. There would be no shame, or embarrassment or feelings of letting someone down. Just acceptance of a condition and lessons to fit

OMG I LOVE IT

Very cool idea, I have given real life and online classes, I can see some of that kind of material being summarised, converted to text, put in a chatbot, or whatever else you can imagine. The most interesting classes are the ones who have the most questions, need the most help (ones academically frowned upon). Kids questions would provide a big input.

Kids like me weren't even taught. Maths classes i was put in another room with a pack of cards and a couple of others, and told to play cards ... This could be a real game changer

Is advanced AI actually smart? 'No, it's using the same system as a pigeon', study finds

https://news.sky.com/story/is-advanced-ai-actually-smart-no-its-using-the-same-system-as-a-pigeon-study-finds-12805120

Basically its not thinking. Its a search that finds phrases, programs etc on the net. Where it takes chunks of it and adds a few letters, really no different than buying a term paper and changing a few things. What I think makes it smart is understanding roughly what we want.

Like amazon could really use an AI search. Look for a part number and it comes up with a 1000 part numbers, none are mine.

I can see AI doing searches like - my couch is red, I like ducks, I also like ham, find something that matches this in the form of art on a wall. Enter that in google and you'll get all sorts of weird stuff like couches and ducks (just tried it now, its what I thought I'd see). But I think if the AI was smart, it would find duck and ham related art with a red theme. Not sure if we are there yet though.


----Mike Savad

What is interesting is that A.I.s are being compared to living warm blooded animals such as birds. Ravens and Parrots are highly intelligent.

Enjoy:






And then there's the pigeon:

We know what artificial intelligence is in digital gadgets. It's when, presented with data and a choice, they select an option that looks like what we would do. That makes good programming, but the second question is OUR intelligence. Human brains and behavior are conditioned not just by mathematical logic but by experience and the emotional impact of that experience and the outcome of our previous attempts to understand and deal with recurring situations.

Some of this certainly is intelligence, some of it might be intelligence, some of it is something that satisfied our emotional urges and some of it is completely bonkers. I worry about those urges, not to mention the completely irrational impulses. Nevertheless, it's what probably what will get programmed into a hypothetical AI device that's supposed to emulate intelligence....we are the model, the only one we have.

AI X-ray screening tool is twice as effective at discovering lung cancer as doctors, study suggests - https://www.dailymail.co.uk/health/article-11724057/AI-X-ray-screening-tool-twice-effective-discovering-lung-cancer-doctors.html

Lung cancer is the leading cause of cancer death worldwide. An estimated 238,340 PEOPLE will be diagnosed with lung cancer in 2023 in the U.S. 1 IN 16 people will be diagnosed with lung cancer in their lifetime – 1 in 16 men, and 1 in 17 women.

We can use AI to design better AI........
Oops! IT'S already is doing that.

https://www.technologyreview.com/2021/05/27/1025453/artificial-intelligence-learning-create-itself-agi/

Human growth patterns are based on education through life experiences and a desire to self improve.

Give an AI a phish but teach and AI to Phish. Thatz........well another thing entirely.

I wonder what kind of safeguards did they put in these Ai's and how many are being removed when they program themselves... Like can some teen ask for hacking software? Or how to make a bomb? Etc will it just give the data?

I know there is a reddit threaddit that has that very thing and looking for ways to get around some of the safeguards. I just wonder how free and loose with the info these chats are.


----Mike Savad

Https://www.vice.com/en/article/qjk745/ai-police-sketches

then there is this thing, they are making police sketches from I guess random bits from the net. But they are afraid of both racial profiling and seeing a complete image and that image becomes the bad guy in their head.

When asking DALLE for gang members, they were all one color.

It would be interesting to see the AI and a real sketch artist using traditional tools and the computer version some stations use, and see if they all get a similar face in the end.


----Mike Savad

News from Israel -> SpotitEarly, which is building an early cancer detection system based on dogs’ superior sense of smell combined with artificial intelligence.

See NEW STARTUP IS TRAINING LABRADORS TO DETECT CANCER (https://www.israel21c.org/new-startup-is-training-labradors-to-detect-cancer/).

Best of three worlds: humans, other animals and AI :-)