Master AI (Artificial Intelligence) Discussion/News Thread

[rdenney]

I was also thinking about a summary. English is not my native language, and although I understood @rdenney Rick's message (more or less ) I just wanted to be sure so i ask Chatgpt to annalyse Ricks reply looking for a relatively short summary that still does it justice.

Chatgpt Revised Short Summary:​

Rick talks about the problem of who decides what is true in the world of AI. If only approved information can be used, then those in control may push their own views. But letting users check AI output doesn't work either people use AI because they don’t already know the facts.

He says AI should be more transparent, but right now, the sources it gives are often weak or don’t match the answers. Rick compares this to how journalism and science used to follow strict rules to separate facts from opinions rules that AI does not follow.

Even though many people are thinking about these issues, Rick is not sure a good solution will be found. He worries that control over AI may end up in the wrong hands.

There is a lot that is wrong with that summary, and a lot that it misses. Example: It's not that control over AI ends up in the wrong hands, it's that there may be no "right" hands.

Rick "thinking summaries should retain meaning and not just words" Denney

 

[Chr1]

Reckon AI definitely has it's uses. And it's misuses.
Trouble is, there's no real distinction being made presently.

And little regard for the consequences.
My 2c.

 

[Petrushka]

I can put it in three. Prediction: AI will be a tool of propagandists as easily as for educators. Both will be morally challenged by the technology. Users will be unable to tell the difference between propaganda and education, or perform their own validation.

Rick "a point of view you might not get from ChatGPT" Denney

I’m rather old. Most of what I was taught in school is now considered propaganda.

I suspect most of what is now taught will someday be considered propaganda.

This is true even in the hard sciences. String theory is kaput, and there hasn’t been a new and useful idea in 50 years.

I blame the Trisolarians.

 

[Chr1]

I blame it on the Boogie.
Mostly.

 

[Snarfie]

There is a lot that is wrong with that summary, and a lot that it misses. Example: It's not that control over AI ends up in the wrong hands, it's that there may be no "right" hands.

Rick "thinking summaries should retain meaning and not just words" Denney

A summary in this case is difficult to make also by Chatgpt i guess understanding and including every complex detail isn’t easy. For me personally responding is difficult English isn’t my native language, so it’s harder for me to fully express or respond to all the points you made. But I see your point, and you’re right that some of the deeper meaning didn’t come through in Chatgpt summary respons.

 

[rdenney]

A summary in this case is difficult to make also by Chatgpt i guess understanding and including every complex detail isn’t easy. For me personally responding is difficult English isn’t my native language, so it’s harder for me to fully express or respond to all the points you made. But I see your point, and you’re right that some of the deeper meaning didn’t come through in Chatgpt respons.

I understand the language issue, and I did not mean to minimize that.

But having spent my adult life explaining technical concepts with writing, I tend to prefer my own modes of expression, even though I do realize that sometimes puts a burden of understanding on the reader. That is particularly the case when writing for a forum about one of my hobbies. Writing at work is much more strict. Sometimes, those modes of expression are easter eggs (read: little gifts) for readers who enjoy what I imagine to be a witty turn of phrase. That those attempts usually fall flat bothers my self-image not at all But more often they work at multiple levels of meaning, and an AI summary might only catch the surface meaning.

That is perhaps the greatest sense of threat posed by large language models to those who value the writing process--it trivializes what took many of us who are not naturally talented writers years of dedicated practice to reach competency. Many engineers and other people in technical professions struggle with it, and I'm sure they are hoping the LLMs will be their equalizer. I fear they hope in vain--good writing is meaningless if the ideas expressed have no point of view, and that's one attribute I do not see in the LLMs at present. But it may be good enough for their purposes.

Rick "sometimes having a point of view is a detriment" Denney

 

[Snarfie]

I understand the language issue, and I did not mean to minimize that.

But having spent my adult life explaining technical concepts with writing, I tend to prefer my own modes of expression, even though I do realize that sometimes puts a burden of understanding on the reader. That is particularly the case when writing for a forum about one of my hobbies. Writing at work is much more strict. Sometimes, those modes of expression are easter eggs (read: little gifts) for readers who enjoy what I imagine to be a witty turn of phrase. That those attempts usually fall flat bothers my self-image not at all But more often they work at multiple levels of meaning, and an AI summary might only catch the surface meaning.

That is perhaps the greatest sense of threat posed by large language models to those who value the writing process--it trivializes what took many of us who are not naturally talented writers years of dedicated practice to reach competency. Many engineers and other people in technical professions struggle with it, and I'm sure they are hoping the LLMs will be their equalizer. I fear they hope in vain--good writing is meaningless if the ideas expressed have no point of view, and that's one attribute I do not see in the LLMs at present. But it may be good enough for their purposes.

Rick "sometimes having a point of view is a detriment" Denney

Got your point. I had to look up ‘detriment’ (disadvantage). I’ll avoid getting witty in Dutch. But I do see how your style adds layers that AI might miss.

 

[xanalog]

Can you ask AI to condense that post into 4 sentences?

This stuff is developing so much faster than a speeding bullet that I need a new assistant to summarize the developments with all these agents running amuck.
I fired my last assistant because s/he forgot to notify me that Zhipu is now Z.ai...

 

[amper42]

AI is gutting workforces—and an ex-Google exec says CEOs are too busy ‘celebrating’ their efficiency gains to see they’re next

Google’s former chief business officer Mo Gawdat says that the notion AI will create jobs is ‘100% crap’. He warns it’ll wipe out jobs, with even ‘incompetent CEOs’ and ‘evil’ government leaders on the chopping block.

www.yahoo.com

This article makes it sound like working with AI is only temporary - until the AI kicks you off the job.

 

[amirm]

Funny bit...

Had a nice family photo that otherwise had a very out of focus image of someone's head on the bottom 10% of the picture. Didn't want to crop it out so asked Photoshop (or was it Lightroom?) to use its AI to fix it. I said it was going to search around and do that. After about 10 seconds, it replaces that fully out of focus head, with a perfectly in-focus picture of a random person!!!

 

[DVinylfan]

Funny bit...

Had a nice family photo that otherwise had a very out of focus image of someone's head on the bottom 10% of the picture. Didn't want to crop it out so asked Photoshop (or was it Lightroom?) to use its AI to fix it. I said it was going to search around and do that. After about 10 seconds, it replaces that fully out of focus head, with a perfectly in-focus picture of a random person!!!

It happened perhaps because AI has absolutely no concept of the idea of family or the importance of people close to us.
Which is exactly why I never, and never will, use it

 

[Chr1]

Funny bit...

Had a nice family photo that otherwise had a very out of focus image of someone's head on the bottom 10% of the picture. Didn't want to crop it out so asked Photoshop (or was it Lightroom?) to use its AI to fix it. I said it was going to search around and do that. After about 10 seconds, it replaces that fully out of focus head, with a perfectly in-focus picture of a random person!!!

Wow.
Getting weird when other people's random photos (soon videos too, presumably), might happen to feature you. Or your peeps.
(And, damn. Might have to shave more regularly too now.)

Strange times indeed

 

[Salt]

Funny bit...

Had a nice family photo that otherwise had a very out of focus image of someone's head on the bottom 10% of the picture. Didn't want to crop it out so asked Photoshop (or was it Lightroom?) to use its AI to fix it. I said it was going to search around and do that. After about 10 seconds, it replaces that fully out of focus head, with a perfectly in-focus picture of a random person!!!

Scaring

 

[Timcognito]

I'm sure there will be a big market for photos that make one thinner, younger, taller, tan, bigger eyes, smaller nose, thicker hair...

 

[Chr1]

Already is. My new phone appears to have most of the functioninality of Photoshop.
Plus simplified AI version thereof too.

 

[amirm]

I'm sure there will be a big market for photos that make one thinner, younger, taller, tan, bigger eyes, smaller nose, thicker hair...

I am about to purchase such a software for the wedding I just shot. It is called Evoto. It has full knowledge of human anatomy and can apply enhancements to any such component! Makes retouching portraits so much easier than screwing around with Photoshop.

It is one of the best uses of AI I have seen.

 

[Power Pop 23]

I am about to purchase such a software for the wedding I just shot. It is called Evoto. It has full knowledge of human anatomy and can apply enhancements to any such component!

Amir,

Please let the members know when Evoto can enhance the memory component of the brain

thank you !

 

[Chr1]

Apologies as slightly off topic, but I was impressed with what I can get straight from my new phone. (Taken and processed in 3 minutes on Xiaomi 14)...

 

[Power Pop 23]

I am about to purchase such a software for the wedding I just shot. It is called Evoto. It has full knowledge of human anatomy and can apply enhancements to any such component! Makes retouching portraits so much easier than screwing around with Photoshop.


So, instead of going on a GLP-1 medication before the wedding, an invited guest can now request 'post-production enhancement' ?

 

[pablolie]

I am about to purchase such a software for the wedding I just shot. It is called Evoto. It has full knowledge of human anatomy and can apply enhancements to any such component! Makes retouching portraits so much easier than screwing around with Photoshop.

It is one of the best uses of AI I have seen.

Personally I like the fact I grew up in an age where my pictures reflects the reality of the moment.

Optimizing the output of the camera (contrast, exposure, saturation etc) is very different from modifying appearance to adjust it to some standard of eye size or white teeth or how many wrinkles are allowed. Done in moderation... why not. But new tools often get used to ridiculous excess.