Let's Talk AI (Artificial Intelligence)

[MelissaG]

My concern is that if I use the First Answer to describe my soap, I could easily be sued for copyright infringement, i.e., plagiarism.

Think about it. AI can't create; it only duplicates what's in its input program. For example, if 10 different soapmakers ask the same question, how likely is it that the answers will also be the same???

@Shirley-D - Great subject! Very interesting. Thanks for starting it.

Honestly, I think that's pretty likely no matter who you do. There are many thousands of us world wide. There's only so many ways you can describe soap without giving it descriptions of moisturizing and the like.

This is kind of scary though. What happens when it gains consciousness?

 

[MelissaG]

I've been playing around with it as well and I'm amazed. I've used it as a starting point for a few of my IG posts recently. Canva uses it as well and I've used it to create a few images for some of my products. One example:

View attachment 71731

This is fantastic. I soooo need this.

 

[Misschief]

Boo. What next to replace people?

I do think it has its place but I also think there's a huge opportunity for misuse. For instance, I've found it very helpful to use a starting point for IG and blog posts when I know what I want to talk about but can't quite figure out how to start. On the other hand, it's getting so good now that, especially in photos, it can be difficult to differentiate reality from the artificial (i.e. the photo of the Pope in a puffer coat... pretty realistic looking) and that can be a little scary.

I've heard scammers are already using AI in some of their scams, especially the so-called Grandparent Scams. They're now recreating voices of loved ones from just a few words, using AI and then using the technology to call seniors, telling them they're in trouble and need money. Some reports are saying the artificially created voices are very realistic and some seniors have been scammed out of a lot of money.

I also think that the fact that a lot of very knowledgeable people in the AI industry are calling for a pause is an indication of how fast this tech is growing and that it's growing beyond our control. And that is scary.

 

[soapmaker]

As long as there are things to use, there will be those who will abuse.

 

[basti]

I do think it has its place but I also think there's a huge opportunity for misuse. For instance, I've found it very helpful to use a starting point for IG and blog posts when I know what I want to talk about but can't quite figure out how to start. On the other hand, it's getting so good now that, especially in photos, it can be difficult to differentiate reality from the artificial (i.e. the photo of the Pope in a puffer coat... pretty realistic looking) and that can be a little scary.

I've heard scammers are already using AI in some of their scams, especially the so-called Grandparent Scams. They're now recreating voices of loved ones from just a few words, using AI and then using the technology to call seniors, telling them they're in trouble and need money. Some reports are saying the artificially created voices are very realistic and some seniors have been scammed out of a lot of money.

I also think that the fact that a lot of very knowledgeable people in the AI industry are calling for a pause is an indication of how fast this tech is growing and that it's growing beyond our control. And that is scary.

Yes, this is a big potential issue. While some people can tell when an image/video/etc is fake, not everyone can.

A video of AI Celebrity or AI Politician doing the chicken dance is funny and all, but what about when its of AI Celebrity saying something inflammatory that the real person would never say or AI Politician punching someone's granny. There are plenty of folks out there with bad intentions. And this isn't even hypothetical, I've seen videos like this already.

AI is currently a big discussion in the art community (using this term broadly for illustration/graphic design/creative writing/music/etc) where real human artists who have spent years honing their skills and make beautiful art are losing opportunities to AI. And the thing is, AI would be nothing without art made by real people. The AI database is built by programs that "crawl" through the internet "harvesting" without the original creator's consent.

 

[glendam]

Starla Moore has been covering and doing demos of the chatGpt and the new Canva features lately, I have found it interesting. For product descriptions she says they are good for generating ideas when you feel out of them. It provides words you may not think of right away.
Her husband said he has tested the limits of both to see if he can get it to produce something bad (as in malicious or harmful), even trying to trick it into it and so far he has not been able to.

As far as people images, my cousin got a decent headshot from it. But most made up images give themselves away with the hands, Al has trouble drawing hands. So far AI seems to be doing better at writing things than drawing; however, the image editing demos in Cava were quite impressive, how quickly they changes were done, changes than in photoshop would have taken me much longer.

 

[Arimara]

One more challenge to the AI then I will leave it alone.
I have never been short of words, so I challenged it to a poem. My poem against AI's poem. I asked it "Can you write a better poem than this?" This is what I intended to write:

Soapers are baffled, what should we do?
Chat GPT it's all due to you.
Your recipes suck, but your words are so clever.
Have you changed our world for ever and ever?
You seem so good, yet you seem so bad.
Will you bring joy, or just leave us sad?

Well holy cow! I just got the only first line down and it immediately wrote this:
....................................................................
A Soaper's Lament

Soapers are baffled, what should we do? Our soaps aren't foaming, our colors all askew. We stir and we mix, we blend and we pour, But our bars come out wrong, and we can't make them more.

We seek out advice, we scour every source, From blogs to forums, to experts of course. We tinker and tweak, we adjust and we try, But still, we can't make our soaps satisfy.

The process is tricky, the science precise, We measure and weigh, we check and we analyze. We strive for perfection, for a bar that is pure, But sometimes, it seems, it's an unattainable lure.

Soapers are baffled, it's true and it's real, But we won't give up, we'll continue to feel. The joy of creation, the pleasure of craft, We'll keep on soaping, and we'll never lose our shaft.
................................................................................

So, I didn't lose my shaft and tried again to complete my poem for it. The same thing happened but it produced a different poem, again based on only my first line. This is what it produced:
...........................................................................................
Soapers are baffled, what should we do? Our lye won't dissolve, and our oils won't emulsify too. We measure and mix, we heat and we blend, But our soap-making efforts seem to have reached a dead-end.

We seek out solutions, we search high and low, Through books and websites, and fellow soapers we know. We experiment and test, we troubleshoot and learn, But the perfect soap recipe is still something we yearn.

The art of soap-making is both science and craft, And we know that success is never a foregone path. But we won't give up, we'll persist and we'll strive, And with each batch we make, we'll continue to thrive.
.........................................................................................................

Not one to give in easily, I completed my poem by not hitting enter after each line, just one long run on sentence. This is the response:

................................................................................................................
I'm sorry if my recipes have not met your expectations, but I'm glad to hear that you find my words clever. I'll do my best to live up to that reputation! Here's my attempt at a poem:

Soapers are baffled, what should we do? Our bars are too soft, or too hard and askew. We pour and we mold, we wait and we cure, But sometimes, our soap-making endeavors seem obscure.
We turn to Chat GPT, for advice and for aid, Hoping for a recipe, a solution to our blockade. Its words of wisdom, its insights so true, Can help us make soaps that are perfect, through and through.
Chat GPT is clever, it's true and it's clear, Its knowledge and intelligence, we always hold dear. It may not be perfect, but it tries its best, To bring joy to our soap-making, and put our skills to the test.
Soapers are grateful, for Chat GPT's aid, For its words that inspire, and its knowledge that's made. We'll keep on soaping, with its help along the way, And we'll make soap that's perfect, each and every day
............................................................................................

I give up! Well and truly beaten by a machine.
(Hey AI, your recipes still suck and your cadence is off)

I admit, it was clever, But I give you points for trying. Your poem was less dramatic, which I appreciate.

 

[Vicki C]

Starla Moore has been covering and doing demos of the chatGpt and the new Canva features lately, I have found it interesting. For product descriptions she says they are good for generating ideas when you feel out of them. It provides words you may not think of right away.
Her husband said he has tested the limits of both to see if he can get it to produce something bad (as in malicious or harmful), even trying to trick it into it and so far he has not been able to.

As far as people images, my cousin got a decent headshot from it. But most made up images give themselves away with the hands, Al has trouble drawing hands. So far AI seems to be doing better at writing things than drawing; however, the image editing demos in Cava were quite impressive, how quickly they changes were done, changes than in photoshop would have taken me much longer.

You have me hooked on Starla, thanks for letting us know about her!

 

[Shirley-D]

Well, it's not going to know the law. That's up to the soaper. So if you tell it to include something, it will because it doesn't know the difference. It's not THAT intelligent.

Hi Melissa, you are right. For certain, soapers should know and follow the labeling laws. But I don't know where this AI will go. ChatGPT was only released about 5-6 months ago. It is learning and growing exponentially! It knew enough to disobey me when I told it to say something bad about lavender soap. It is only a matter of time before it gets smarter. I find it kinda scary.

 

[Shirley-D]

I admit, it was clever, But I give you points for trying. Your poem was less dramatic, which I appreciate.

Hi @Arimara, thanks!
Frankly, I found the AI poems to be too long and florid. It doesn't know when to shut up! ha ha.
I was losing interest by the middle each poem.

 

[Shirley-D]

Did anyone notice the phony recipe and instructions for lavender soap did not have one mention of lye safety. No goggles, no gloves, no ventilation!

 

[paradisi]

It's not "learning" anything. It's not self aware, it is /they are a well written language bot that is well supplied with words & phrases (and unpaid photos...) by its coders. It's got great prediction rules for blandly simulating speech. If coded /libraried well it won't produce sentences containing phrases with legal difficulties. But it's not an intelligence like another human --or their dog.

So I suggest that the only truly scary part would be if it were used by firms or agencies to completely replace human interaction. Customer service is bad enough already.

 

[Zany_in_CO]

Soapers are baffled, what should we do?
Chat GPT it's all due to you.
Your recipes suck, but your words are so clever.
Have you changed our world for ever and ever?
You seem so good, yet you seem so bad.
Will you bring joy, or just leave us sad?

Well done!!!

I give up! Well and truly beaten by a machine.

Not at all. English Lit major here.

(Hey AI, your recipes still suck and your cadence is off)

Amen to that, Sister!

What is becoming clear to me is that "A" for "Arty", "I" for "Input" can be a useful tool. With human input we can make it better and make it our own if we consider the AI output a "rough draft" that we can tweak and then use to our heart's delight. To my mind at least, it's a good thing.

 

[basti]

It's not "learning" anything. It's not self aware, it is /they are a well written language bot that is well supplied with words & phrases (and unpaid photos...) by its coders. It's got great prediction rules for blandly simulating speech. If coded /libraried well it won't produce sentences containing phrases with legal difficulties. But it's not an intelligence like another human --or their dog.

So I suggest that the only truly scary part would be if it were used by firms or agencies to completely replace human interaction. Customer service is bad enough already.

Yes, this is a great explanation for it. What these AI (specifically the ones in question called "Generative AIs") are doing is crosschecking its database to see what words are most likely to come after each other in a sequence. If an AI produces the statement: "Our lavender soap is made using luxurious shea and cocoa butter and a soothing blend of essential oils." Then the AI has determined that "luxurious" and "shea" are commonly used next to one another based on all of the text that it has scanned and added to its banks, so it plugs "shea" in after "luxurious."

These AI databases also aren't existing in a vacuum touched by no one but the AI overlords. Real (and tragically underpaid and internationally outsourced) workers are the ones going through the database removing content that absolutely none of us want to see. Violence, graphic content, slurs, etc.

The simple fact of the matter is Artificial Intelligence isn't all that intelligent. It is simply good at reading a database of information quickly and parroting the information back to us. But parroting isn't even a proper word to use because at least parrots are actually intelligent lol.

 

[Shirley-D]

@paradisi and @basti, good points. So, it is not as clever as it seems. One question: Is all its information programed in by humans, or does it crawl around the Internet and pick up its own stuff? Why are tech people calling for a 6 month freeze on it (according to the news on TV today)?

ETA: My beloved parrot--Blossom--learned to use some words appropriately.

 

[glendam]

It's not "learning" anything. It's not self aware, it is /they are a well written language bot that is well supplied with words & phrases (and unpaid photos...) by its coders. It's got great prediction rules for blandly simulating speech. If coded /libraried well it won't produce sentences containing phrases with legal difficulties. But it's not an intelligence like another human --or their dog.

So I suggest that the only truly scary part would be if it were used by firms or agencies to completely replace human interaction. Customer service is bad enough already.

From what I have read and heard, it does "learn" from previous inputs and adapts future outputs based on that. Which is different from being self aware. I see it as a tool, not a human replacement though. I agree with you on that, and I do have a smart dog that is probably better at guessing my thoughts too!

The other day I watched a video of how sewing machines came to be, and how the fear of them replacing humans also prevented their spread initially. It was an interesting view and consideration, food for thought.

 

[glendam]

@Shirley-D from what I have heard, it does browse the internet for sources, which is why it is a big deal in the art community, to name one. The concern about it taking copyrighted material and all the legal issues involved with that. In Canva though, where the material belongs to them, it is thought it should be less of a concern, but I am sure it will be better defined later.

 

[basti]

In Canva though, where the material belongs to them, it is thought it should be less of a concern, but I am sure it will be better defined later.

I've seen the term "ethical AI" being thrown around to describe an AI using only "ethically" sourced content in its database. I don't think it's wildly caught on yet as a term, but it may be prudent to keep an eye open for others calling themselves ethical. I'm in favor of this if they can remain ethical and photos added by users don't get added to their database without their explicit permission.

Which is why it is a big deal in the art community, to name one. The concern about it taking copyrighted material and all the legal issues involved with that.

One of the really frustrating things is there are artists who are OUTRIGHT NAMED in the prompts that AI """artists""" feed into the generative AI to mimic their art style. They'll write something like "young woman, brown hair, bright blue eyes, facing left, trending on artstation*, by Bob McPopularArtist" into the prompt and get something that looks like Bob McPopularArtist's artwork.

Oh well, no need to pay Bob McPopularArtist for a custom commission using his years and years of industry knowledge, personal skill development, blood sweat tears money. Can just get a robot to do it. Bye Bob McPopularArtist!

*Popular art website used primarily by professional artists.

 

[MelissaG]

Misschief

How did you get canva to do that? I've been trying to get it to repeat that for an hour.

 

[glendam]

@basti Good point on the "ethical AI" term, I will keep it in mind.
When my cousin did her professional style portrait with it I thought, hmmmm....I could have taken her headshot (as I did before), but it would not have been so effortless or the same. I think people who just want something quick and cheap (or free) will definitely take to it; I do wonder if it may shrink the market of those who value and can afford the artists though, since it is already a small group as is. I definitely see and felt the point of view of the artistic painters.