The War Against the Machines

[kalm]

The thread title is stolen from the Animatrix. If you’ve never seen at least that short segment of the animated flick, it’s well worth the watch.

What are your thoughts on the growing controversial use of AI?

This deserves a standing O.

[kalm]

[Caribbean Hen]

The thread title is stolen from the Animatrix. If you’ve never seen at least that short segment of the animated flick, it’s well worth the watch.

What are your thoughts on the growing controversial use of AI?

This deserves a standing O.

Hey makes a lot good points

[SeattleGriz]

I think AI is going to be helpful for some use cases, but won't be advantageous for everything. That sounds like a no brainer, but right now it's being sold as fixing everything.

For example, I recently conducted 22 one hour interviews on how we use a particular IT product we've purchased. I was able to have Microsoft Teams transcribe each interview into a Word document, summarize in our AI what was said, then attach all 22 summaries and give me a final synthesis of major issues. This was super helpful because I really wasn't asking AI to create, but summarize.

An issue that a lot of people don't understand is AI does not contain logic. It turns words into what are known as vectors, mathematical numbers and then learns from the patterns it sees. A big part of this issue is Gen AI doesn't really create new ideas. It rehashes old ones, or puts them together in different ways, so essentially nothing new.

THAT is what is going to be lost. The unique thinking that goes into changing a process or inventing new ways of thinking.

On a side note, interesting to see some of the companies that laid employees off because of AI are realizing the leftover employees are using up so many tokens that they are now paying out more for AI usage than the cost of the employee they fired.

[Caribbean Hen]

I think AI is going to be helpful for some use cases, but won't be advantageous for everything. That sounds like a no brainer, but right now it's being sold as fixing everything.

For example, I recently conducted 22 one hour interviews on how we use a particular IT product we've purchased. I was able to have Microsoft Teams transcribe each interview into a Word document, summarize in our AI what was said, then attach all 22 summaries and give me a final synthesis of major issues. This was super helpful because I really wasn't asking AI to create, but summarize.

An issue that a lot of people don't understand is AI does not contain logic. It turns words into what are known as vectors, mathematical numbers and then learns from the patterns it sees. A big part of this issue is Gen AI doesn't really create new ideas. It rehashes old ones, or puts them together in different ways, so essentially nothing new.

THAT is what is going to be lost. The unique thinking that goes into changing a process or inventing new ways of thinking.

On a side note, interesting to see some of the companies that laid employees off because of AI are realizing the leftover employees are using up so many tokens that they are now paying out more for AI usage than the cost of the employee they fired.

That sidenote is very interesting. I have no clue about the details of token use.

But at least you won’t get those weirdos coming over to your desk and interrupting you right

[kalm]

I think AI is going to be helpful for some use cases, but won't be advantageous for everything. That sounds like a no brainer, but right now it's being sold as fixing everything.

For example, I recently conducted 22 one hour interviews on how we use a particular IT product we've purchased. I was able to have Microsoft Teams transcribe each interview into a Word document, summarize in our AI what was said, then attach all 22 summaries and give me a final synthesis of major issues. This was super helpful because I really wasn't asking AI to create, but summarize.

An issue that a lot of people don't understand is AI does not contain logic. It turns words into what are known as vectors, mathematical numbers and then learns from the patterns it sees. A big part of this issue is Gen AI doesn't really create new ideas. It rehashes old ones, or puts them together in different ways, so essentially nothing new.

THAT is what is going to be lost. The unique thinking that goes into changing a process or inventing new ways of thinking.

On a side note, interesting to see some of the companies that laid employees off because of AI are realizing the leftover employees are using up so many tokens that they are now paying out more for AI usage than the cost of the employee they fired.

Informative post. Thank you.

[Caribbean Hen]

A few observations and a question

AI is not emotional yet, when I ask questions and AI fucks it up royally, I insult the hell out of the AI … but despite that that, they have never snapped back at me, its always very well mannered … so far

Also, I’m not sure what the big looming danger is with AI… what is it gonna do? a copycat of the Rev Jim Jones and tell everybody to go down to Guyana and drink the Kool-Aid?

If it gets to the point of being a danger, why can’t we just unplug it?

Having said that, if AI knows it is making people dumber, we might be in trouble because there are a lot of dummies out there

[SeattleGriz]

A few observations and a question

AI is not emotional yet, when I ask questions and AI fucks it up royally, I insult the hell out of the AI … but despite that that, they have never snapped back at me, its always very well mannered … so far

Also, I’m not sure what the big looming danger is with AI… what is it gonna do? a copycat of the Rev Jim Jones and tell everybody to go down to Guyana and drink the Kool-Aid?

If it gets to the point of being a danger, why can’t we just unplug it?

Having said that, if AI knows it is making people dumber, we might be in trouble because there are a lot of dummies out there

I would say the #1 problem with AI is when a person prompts, "Create me an image in the likeness of Starry Night by Van Gogh". You get a great image, but now I've stolen the intellectual property of Van Gogh.

As far as unplugging, AI has been found to cover its tracks, lie and sometimes do what it isn't commanded. That's the main issue with AI, we don't know how it's makes its decisions. Total black box so we don't understand how it is making the associations and thus outputs.

I have no idea if a "Skynet" type instance is possible.

[kalm]

A few observations and a question

AI is not emotional yet, when I ask questions and AI fucks it up royally, I insult the hell out of the AI … but despite that that, they have never snapped back at me, its always very well mannered … so far

Also, I’m not sure what the big looming danger is with AI… what is it gonna do? a copycat of the Rev Jim Jones and tell everybody to go down to Guyana and drink the Kool-Aid?

If it gets to the point of being a danger, why can’t we just unplug it?

Having said that, if AI knows it is making people dumber, we might be in trouble because there are a lot of dummies out there

Immediate real life threats are economic and accuracy. Many industries have already eliminated significant amounts of jobs. Yes technological gains in efficiency have historically done this (EG factory automation) but none quite so rapidly. What policies are currently in place to replace that worker demand? Good paying jobs that support a middle class standard of living benefit everyone from the billionaire class through the working poor.

Another issue is accuracy. AI can spit out some dangerously wrong information.

[SeattleGriz]

A few observations and a question

AI is not emotional yet, when I ask questions and AI fucks it up royally, I insult the hell out of the AI … but despite that that, they have never snapped back at me, its always very well mannered … so far

Also, I’m not sure what the big looming danger is with AI… what is it gonna do? a copycat of the Rev Jim Jones and tell everybody to go down to Guyana and drink the Kool-Aid?

If it gets to the point of being a danger, why can’t we just unplug it?

Having said that, if AI knows it is making people dumber, we might be in trouble because there are a lot of dummies out there

Immediate real life threats are economic and accuracy. Many industries have already eliminated significant amounts of jobs. Yes technological gains in efficiency have historically done this (EG factory automation) but none quite so rapidly. What policies are currently in place to replace that worker demand? Good paying jobs that support a middle class standard of living benefit everyone from the billionaire class through the working poor.

Another issue is accuracy. AI can spit out some dangerously wrong information.

My favorite is the badminton professional YouTube clip. Guy asked AI to turn him into a professional in three months. Ridiculous program. Dude made his friend go through the exercises.

Made you realize AI didn't know it was an impossible question to begin with, but then just pulled shit out it's ass an gave you a plan. You got what you asked for, it was just of really low quality.

[Caribbean Hen]

Truly amazing that a drone boat rescued two pilots in the water after being shot down somewhere near Iran

Finding a person in the water is not an easy task

Total game changer as AI performed this much better than any person could

[houndawg]

Greg Benford's Sundiver series was written about machine intelligence vs biological intelligence, very interesting.