Would you believe it? Artificial Intelligence.

[maik 455]

DALL-E has placed me in a futuristic Blade Runner-style universe 😀

[blue_flag 2,008]

Similar text prompt with Photoshop.

[blue_flag 2,008]

[blue_flag 2,008]

 

I have been playing with one of the on-line Artificial Intelligence image generator websites and started to wonder where the original 'faces' and 'bodies' came from.  All the above had the same description of the person and with the exception of the first image had the same description of the background.  The only changes in the text prompts related to the clothing.  The last few were done on Photoshop.

Some of the faces were almost too unblemished to be genuine and looked artificial; others not.  My question is still where do the original faces and bodies come from?

Edited November 5, 2024 by blue_flag

[blue_flag 2,008]

Flying over Thailand.  The birds with my photographs however the backgrounds were AI generated.

Edited November 16, 2024 by blue_flag

[blue_flag 2,008]

From Soi 6 to another bar.

[blue_flag 2,008]

Everybody is texting.

[blue_flag 2,008]

[baeegew 0]

this is awesome

[sailingbill 1,264]

On 11/5/2024 at 1:41 PM, blue_flag said:

have been playing with one of the on-line Artificial Intelligence image generator

Please continue playing, maybe sans clothes? 

[teelack 10,546]

I haven't really given AI much thought. I do however like YouTube a lot and the videos can be a real eye opener at the moment.

I mention this as there seems to be so little genuine intelligence out there at present, I guess that we should be thankful for the artificial sort!

Sent from my CPH2669 using Tapatalk

[yselmike 448]

17 hours ago, teelack said:

I haven't really given AI much thought. I do however like YouTube a lot and the videos can be a real eye opener at the moment.

I mention this as there seems to be so little genuine intelligence out there at present, I guess that we should be thankful for the artificial sort!

Sent from my CPH2669 using Tapatalk
 

There's a clip doing the rounds.Its a Trump appointment who's charged with AI in education.The thing was she was calling it A1.

[blue_flag 2,008]

Photoshop Beta has now incorporated Google Nano Banana as an add on.  This is a significant improvement on Firefly 3, which Photoshop previously used.  The starting image was as below.  One question I pose;  does anyone recognise the market; it could be artificial or based on a real image.

Edited September 27 by blue_flag

[blue_flag 2,008]

from:-

[LocalYokul 395]

On 4/20/2025 at 3:25 PM, yselmike said:

Trump appointment who's charged with AI in education.The thing was she was calling it A1.

she did it repeatedly, so completely out of touch

but it reminds me of a friend of mine who called Wi-Fi wiffee

 

[yselmike 448]

2 hours ago, LocalYokul said:

she did it repeatedly, so completely out of touch

but it reminds me of a friend of mine who called Wi-Fi wiffee

 

It's also how the Dutch say it 

[Bruce Mangosteen 1,275]

8 hours ago, LocalYokul said:

she did it repeatedly, so completely out of touch

At least she has a degree and background in education.  Compare that to the extent of the "qualifications" of the last Transportation Secretary, Pete Butt-I-Gag:  Having given blowjobs in the men's toilets of the South Bend IN airport.  

Personally, I think AI has no place in education, at least for now.  In fact, it's causing more harm than good at the moment.

https://www.edweek.org/technology/new-data-reveal-how-many-students-are-using-ai-to-cheat/2024/04

Edited Saturday at 07:53 PM by Bruce Mangosteen

[LocalYokul 395]

The students are obviously using it to cheat, the teachers will likely need to be trained to figure that out, or the Artificial Intelligence will have to be trained to not lets students take advantage like that.

 

[Bruce Mangosteen 1,275]

1 hour ago, LocalYokul said:

The students are obviously using it to cheat, the teachers will likely need to be trained to figure that out, or the Artificial Intelligence will have to be trained to not lets students take advantage like that.

Well, first I don't see how that would even be possible, especially if students are rewriting the output from AI sources.  The right answer is the answer the teacher is expecting (which is usually but not always the factually correct answer) regardless of the source.  It's a problem as old as education itself, but at least in the past teachers had recourse to check.

In less fact-driven disciplines, students will be infused with the viewpoints of the people who curated the data used to train the AI model.  Unlike with a human teacher, who can be interviewed and questioned, an AI system can't be held responsible for what children "learn" from it.

We're having enough trouble with human teachers as it is.

https://www.nea.org/resource-library/educator-support-transgender-nonbinary-and-intersex-students-setting-record-straight-k-12-schooling
https://www.justiceinschools.org/school-walkouts-civil-disobedience
https://www.nea.org/nea-today/all-news-articles/meeting-challenge-teaching-climate-change
https://rethinkingschools.org/articles/reproductive-justice-and-our-classrooms/

Whatever your opinion of those topics, you should be able to control and set limits on how they are taught to your children.  Teachers already refuse to accept that; if they outsource children's education to robots, that control will become even harder.

What computer is going to jail for this?

https://www.nbcnews.com/tech/tech-news/family-teenager-died-suicide-alleges-openais-chatgpt-blame-rcna226147
https://apnews.com/article/chatbot-ai-lawsuit-suicide-teen-artificial-intelligence-9d48adc572100822fdbc3c90d1456bd0
https://www.rnz.co.nz/life/wellbeing/the-ai-chatbot-helped-her-write-a-suicide-note

AI needs to be kept far away from kids IMO, in and outside of schools, at least until these and many other questions have been answered.

Edited Saturday at 03:25 PM by Bruce Mangosteen

[LocalYokul 395]

1 hour ago, Bruce Mangosteen said:

I don't see how that would even be possible, especially if students are rewriting the output

Some teachers could spot plagiarism from a 111 Meters, but some may need to be trained.

 

[Bruce Mangosteen 1,275]

1 hour ago, LocalYokul said:

Some teachers could spot plagiarism from a 111 Meters, but some may need to be trained.

Is copying AI "plagiarism"?  Whose work are you stealing?  Isn't it yours?  Are the pics posted on this string "plagiarized"?

[LocalYokul 395]

 

At the lowest level, a book report is just reporting on what you thought you read.

When it comes to solving complex Math problems, there are usually examples available online, so copying that is not solving the problem, it is just copying someone else's method of solving it.

At the end, the student did not think their way out of it, other than using the internet to find an answer.

In the real world, in a working environment, the internet is not always going to solve the problem.

This is why Mechanical Engineers should know how to do the Math that the CAD programs are doing for them.

 

[Bruce Mangosteen 1,275]

6 minutes ago, LocalYokul said:

At the lowest level, a book report is just reporting on what you thought you read.

But you had to read it first.  Not with AI.

7 minutes ago, LocalYokul said:

When it comes to solving complex Math problems, there are usually examples available online, so copying that is not solving the problem, it is just copying someone else's method of solving it.

But the point was to learn the method, which you need to do with your particular example, unless you find an exact duplicate online, the chances of which are nil.  Not with AI, you give it your particular example, and it spits out the equations, so you learn nothing.

8 minutes ago, LocalYokul said:

In the real world, in a working environment, the internet is not always going to solve the problem.

Which is why AI is so dangerous, that and it CAN spit out "solutions" to problems, which might not be correct or even VALID.

9 minutes ago, LocalYokul said:

This is why Mechanical Engineers should know how to do the Math that the CAD programs are doing for them.

And why AI must be kept out of schools:  So that we're not graduating students who know how to do nothing but feed AI robots their assignments, then turn them in.

In my opinion, AI will be dangerous as long as it's an imperfect simulation of human intelligence. 

After that, it will be even more dangerous.

The challenge is and will always be how to harness AI in a way where the dangers can be controlled.  Simply turning it loose on a world full of random quarter-wits and evil fucktards doesn't seem to be a good way to do that, at least so far.

[LocalYokul 395]

 

I was recently researching specifications for external blu-ray readers/ DVD writers & gOOgle's Artificial Intelligence is a liar.

It claimed there were performance differences in two drives with very similar model numbers (off by one Letter).

Well, it turns out it is just a Regional difference, they are the same, but sold in different parts of the world.

That happens with home cinema blu-ray players too, there will be models in Europe that are identical, except for the hard-coded region & Voltage.

but, to be fair to gOOgle, they base that bullshit off of input from real people & sometimes people talk shite

My Sony UBP-X700 home cinema blu-ray player gets touted as being region free, but that is a big fat lie!

 

[Bruce Mangosteen 1,275]

2 hours ago, LocalYokul said:

but, to be fair to gOOgle, they base that bullshit off of input from real people & sometimes people talk shite

Yep, as I said, AI is only as good as the data it's trained on, which in turn is only as good as the people who are feeding it the data.  That is, if anybody is responsible for curating what these models are learning, as opposed to them simply scraping the internet without direction. 

If you "train" "AI" on the "data" from random idiots and lunatics on the internet, the result will be Artificial Idiocy or Artificial Insanity.  Which appears to be what you discovered.

Edited Saturday at 08:04 PM by Bruce Mangosteen