AI investigation

[rebrecs]

Since AI, as a moniker won't go away, I decided I would go ahead and start work on the
Arificial Intelligence Quotient. After all, how will we know it we can trust it if we don't know its AIQ?
First, I need to take a look at non-artificial intelligence. Should we call that real intelligence?
Given a real intelligence, what factors go in to forming the IQ of it ?
Would an AI be measured the same way ( in order to get its AIQ? )
if not, then it must be different than a real intelligence. In what way ?
If it is different, it has to have a different set of evaluations to arrive at an AIQ ?
If that is true, is it actually an intelligence? Or something else?
A new kind of intelligence? Yet still an intelligence.
So, if all intelligences get their own evaluation criteria, is anything an actual intelligence?
Does The word intelligence have any meaning without a standard test that all intelligences get measured by -

As you may can tell, I'm just at square one with this project.

 

[Melensdad]

Dr Jordan B Peterson has been doing an investigation with AI and specifically Chat GPT.

He found it fascinating and also found that, when he directed it to respond with scholarly citations and back up its statements with proof that the Chat GPT would, approximately 85% of the time, actually provide legitimate URLs to scientific studies, but that perhaps 15% of the time the citations would be fake/made up by the AI. As he double checked every source, Dr Peterson would admonish and correct the AI and Chat GPT would react in an apologetic and contrite way, and then would return to providing accurate information.

I suspect that many clerical jobs, including low to Mid-Level LEGAL, MEDICAL and other PROFESSIONAL jobs will be replaced by Chat GPT and other AI alternatives. Trade jobs will be safe. High level legal, medical and professional jobs will likely be safe, because they will lobby for their job security and use the guise of need. In fact high level legal and medical jobs already enjoy that status, granted monopolies by mere fact of their licensing requirements. But para-legals, and the like, enjoy no such protection and those types of jobs will vanish as they are replaced by AI.

Companies will save BILLIONS, perhaps TRILLIONS in labor costs.

But what do we do with the new unemployed that will follow in the next 10 years?

 

[chowderman]

just what we need . . . a AI that does "citations would be fake/made up by the AI" - that'll work out so well, especially in the medical profession.

unless and until AI is iron-clad honest, it'll go nowhere.

 

[rebrecs]

just what we need . . . a AI that does "citations would be fake/made up by the AI" - that'll work out so well, especially in the medical profession.

unless and until AI is iron-clad honest, it'll go nowhere.

it could, for example, moderate a website ?
Hah! (couldn' resist)
I'm skeptical of the cost savings. The figures I see don't have supporting data. A simple example to illustrate where I think the AI marketing people are abusiveL I recall a case in San Francisco, CA where some toll booth attendants were replaced by an automated system. In that article, cost savings were cited to justify the switch. I don't know how much 8 toll booth operators cost. I do however have a pretty good idea that the human operatorcost has to be compared to software updates by trained IT people, , on call techs, and supporting network infrastructure, etc. If history repeats itself- they will get sick of all that "fixed in the next rev" crap and start hiring their own support people.
I have observed that cycle for 25 years.

I hope they didn't put gates on those toll lanes. The guy with no sticker, nor any change, whose pregnant wife has got to get to the hospital right now, will be without the decision making capability a human would provide. Thus that gate will be history within seconds.

 

[rebrecs]

Dr Jordan B Peterson has been doing an investigation with AI and specifically Chat GPT.

He found it fascinating and also found that, when he directed it to respond with scholarly citations and back up its statements with proof that the Chat GPT would, approximately 85% of the time, actually provide legitimate URLs to scientific studies, but that perhaps 15% of the time the citations would be fake/made up by the AI. As he double checked every source, Dr Peterson would admonish and correct the AI and Chat GPT would react in an apologetic and contrite way, and then would return to providing accurate information.

I suspect that many clerical jobs, including low to Mid-Level LEGAL, MEDICAL and other PROFESSIONAL jobs will be replaced by Chat GPT and other AI alternatives. Trade jobs will be safe. High level legal, medical and professional jobs will likely be safe, because they will lobby for their job security and use the guise of need. In fact high level legal and medical jobs already enjoy that status, granted monopolies by mere fact of their licensing requirements. But para-legals, and the like, enjoy no such protection and those types of jobs will vanish as they are replaced by AI.

Companies will save BILLIONS, perhaps TRILLIONS in labor costs.

But what do we do with the new unemployed that will follow in the next 10 years?

Well I hope they don't fire the paralegal folks too quickly. Don't forget, somebody with a good understanding of how things go in the profession has to operate the AI software. The attorneys, accustomed to barking orders and disappearing are (IMHO) not going to want to sit down with the thing and hope they can frame their "search strings" in a useful way. I think the para is much better suited for that. Paralegals just need to see it coming so they can use it. It would be just another office machine to them. One new additional trick the computer can do. The fact that it runs so fast ( like lightning compared to historical search engines) and has night-versus-day improvements over historical inference engines, means they (the paralegals) get done more quickly. And possibly use that time to do more higher level "legal" stuff?
If at all, I would think the attorney's job would be the one to worry about. Especially given all the starving would-be actors and actesses we have who could step right in , as directed by the paralegal and their new toys ....

There is another alternative outcome to consider. This would be the phase 2 scenario where paralegals, attorneys, judges, juries, courtrooms - all fade away. The accused will simply be convicted, or not, if the AI software says so.

The only thing on earth that never gets questioned from an ethical or human rights perspective is technology advancement. Never has, never will. Putting your mind toward that- can take you down some interesting paths.
I have my own answer to that. But that is a different thread.

have a great day while you still can.

 

[Melensdad]

I'd say that PARALEGALS will be on the chopping block as much of that work can be done by AI

Lawyers and judges are a different matter entirely. Their careers are actually protected by law. So they won't be going away. At least not many of them.

 

[chowderman]

AI will not be able to moderate a forum - forums have different focus and tolerance - especially in 'off topic' areas.
an AI would need to 'observe' the human moderated forum for an extended period to learn the flavor of that specific forum.

Pennsylvania Turnpike eliminated all toll both attendants and went with the transponder / toll-by-plate routine.
they are now losing millions per year in tolls not paid. the one-off out of state vehicles one can understand - but the Turnpike people, knowing when and where the scofflaws enter and when and where the scofflaws exit,
day after day,
can't figure out how to catch them.

I submit AI is more needed in a politician / bureaucrat than in normal people's lives . . . .