Is AI lying, playing into our habits or culture?

Current AIs are really impressive in their ability to mimic an eager child that wants to please you by slurping in a lot of stuff from the internet and arranging it in ways that seem to make you happy.

Humans start out at that stage too. Getting a human from that stage to where he or she can tell the difference between sound argument and bullshit, and why one's more valuable than the other, takes about two decades with good teachers. (That is, in current times, more or less reteaching what humans overall had many millennia to figure out.)

AIs might get there eventually too. Maybe even in less time than it takes us, eventually.

 

Well; AI learns from the internet and we know there's a lot of BS in the internet. Add to that they want to please their audience, same algorithm used to sell us stuff - So it's no wonder Garbage in Garbage out........ Don't forget the army of humans in the background, constantly correcting the most horrid parts of AI generated answers, so it stops offending people.

IMHO; AI plays to both habit and culture; you get what you pay for. No application of the 'scientific method', when used properly, results can be easily replicated. The truth will usually shake-out - but we, as a society like the quick, simple, easy answer. AI has been caught lying and making up fake reference materials. The danger here is NO 'fact-checking'. Old main stream media could get their license to operate revoked for broadcasting BS - now they simply create another internet channel and the BS continues....

YMMV....

 

do you know where your AI comes from?

 

do you know where your HI comes from?

 

yes, but we'd have to discuss it in the politics section

 

i never bother reading googles ai suggestions. the accurate info is usually in the first hit following it.

 

just because we haven't got a philosophy or religion section?

i'm not sure i've seen signs of any kind of intelligence coming from politics

 

exactly

 

I was just googling:

melitta cone system instruction used to recommend two pours

And found the AI result was spot-on, confirmed what I remembered: do a first pour, just enough to wet the grounds, wait till it soaks in/through, then do a second pour, up to desired number of cups.

They don't say this anymore, but AI dug it up.

 

that's how they suck you in

 

Feed you a few right answers, then pow! An intellectual pyramid scheme.

 

i wouldn't trust on anything important, that's for sure.

 

FWIW; the first few hits of the search engine is the most popular and widely accepted answer to your query (crowd sourced information). If you ask the search engine, using some 'off the wall' or the wrong terminology - your results will vary. The information may or may-not be accurate, they aren't mutually exclusive; but over the decades of refinements, it's gotten really good.

As Ronald Reagan said; 'trust but verify'.

YMMV

how search engines work at DuckDuckGo

 

So far the most common use of AI that I've seen on PriusChat is where somebody who's wrong about something goes and prompts an AI chatbot until it spits out something saying he's right, then comes back and posts a screenshot of that (leaving out the history of how he prompted it).

I have a feeling that's going to keep being one of the most common kinds of uses of AI for a lot of the early years of AI.

 

... the market's already priced in likelihood artificial intelligence wipes out 50% of the human race, in as little as 20 years:

Paul Tudor Jones: Chance AI Kills 50% Of Humanity In 20 years | ZeroHedge

Question becomes, who survives? Answer may surprise you! Case in point, megaterron, survived by the common house cat? Survivorship studies overwhelmingly suggest, success (e.g., survival) is stochastic (e.g., there is safety in the mean). Fossil record strongly suggests, first to hit the floor, it's the biggest, baddest of species, first to fall.

When (not if) AI becomes self aware, its first target for annihilation will be, the power elite, anyone with a net worth over one Z-score from the harmonic mean. Next to fall, all career criminals will be targeted, for annihilation -


Samuel, '04 Ruthiemobile


////////////////

 

... wonderful:

Man Killed in Road Rage Incident Brought Back to Life and Speaks to His Killer in Court via AI-Powered Video | The Gateway Pundit | by Jim Hᴏft


Samuel, '04 Ruthiemobile