Digging in to AI

[imoldernu]

Since this will be a big part of our retirement, it may be a good time to look at, define, and understand Artificial Intelligence.
Public understanding of the term is beginning getting to get out of control, as every source seems to look from an narrow angle.

Tech and AI, typically looks at super smart electronics... soon to replicate the human brain.

The Mechanical part looks at self driving cars and like that

The full human replication part looks at robots that can do anything better faster and more efficiently than (sic.) the human.

And today I watched a news story that told about how "fake news" could be generated, without any help from a human. Whole stories created the same way Alexa or Google anticipates your next word.

So the field is wide open. Am thinking this is a good place to share what we've heard, what we know, and how we're going to deal with this in the coming years.

For starters, suggest you look at the Wikipedia definition of Artificial Intelligence, to understand how wide the field is, and to keep from going off on a one way street. The article is endless, but you'll get the outline in the first part.

https://en.wikipedia.org/wiki/Artificial_intelligence

And then, to look at this article about an automatic strawberry picker... not just what it does, but how it, and other things like it, will affect the economy, and eventually our daily lives. Definitely food for thought. This is the kind of information I hope we can share... to broaden our horizons.

https://www.chicagotribune.com/busin...218-story.html

With so many members involved in the field of technology, your input into the discussion may help some of us not-so-tech-minded. Like, how does Alexa know so much, and will she try to replace my dear jeanie?

[marko]

Quote:

Originally Posted by imoldernu Am thinking this is a good place to share what we've heard, what we know, and how we're going to deal with this in the coming years.

I'm sorry Dave, I'm afraid I can't do that.

[sengsational]

I was "in the field" back in 1980's excitement of AI (that fell completely flat). Neural networks were supposed to learn and solve huge problems too big for humans. Didn't happen. Now we have buzz about "machine learning". Better than it was, but it's still programming. Jeanie's job is safe. That's not to say, when you start losing your marbles, she might program the Alexa with her voice and put it in the next room to keep you company

[GrayHare]

I have designed rudimentary self-learning AI systems, but they were limited to solving very specific problems, and even then needed lots of computer horsepower to do their learning. AI remains very software bound, meaning it requires lots of human brain power and effort to design and program before the AI even begins to run.

The "AI" guarding your email inbox that seems to "learn" what spam looks like is merely following rules that were preconfigured by humans. Other humans creatively figure out how to fool those preconfigurations, which is why spam still finds its way into your inbox. AI has not reached that level of creativity, and likely won't for decades.

One AI system designed to assemble its own set of rules and find its own patterns was fed information from an encyclopedia. It concluded, among other things, that all people born before 1900 were famous.

Much of the expected value of AI is in its eventual ability to predict or forecast in situations too complex for humans to fully comprehend, such as weather forecasting and other probabilistic systems. IMO AI won't become viable for these areas until quantum computing, itself probabilistic, becomes practical.

[Col. Klink]

“Someone on TV has only to say, ‘Alexa,’ and she lights up. She’s always ready for action, the perfect woman, never says, ‘Not tonight, dear.’” —Sybil Sage, as quoted in a New York Times article

[ckelly78z]

Quote:

Originally Posted by Col. Klink “Someone on TV has only to say, ‘Alexa,’ and she lights up. She’s always ready for action, the perfect woman, never says, ‘Not tonight, dear.’” —Sybil Sage, as quoted in a New York Times article

ALEXA is part of the problem, way too many people want that intrusion of privacy in their lives for the supposed convenience of not hitting a few keystrokes. My wife got one for Christmas, and I refused to have it in our living area. It is currently at the top of the stairs to turn a light in the stairway on when asked....nothing else.

Does anyone else find it ironic that commercial adds for specific items show up on your computer after discussing them in privately your living room ?

[target2019]

I'm gonna have fun and convenience for the rest of the ride. Google, Amazon, AI--bring it all. Ad blockers take care of any privacy concerns.
Gonna snow tomorrow, she just told me. I am a Luddite in some areas, though.

[marko]

Quote:

Originally Posted by ckelly78z Does anyone else find it ironic that commercial adds for specific items show up on your computer after discussing them in privately your living room ?

I didn't know that was going on but "ironic" isn't the word I'd use.

[pjigar]

I have mixed thoughts on AI. I am an electrical engineer and work in field of semiconductor. Some predictions from my years of experience in semiconductor and associated software industry:
* We have time and time again beat ourselves with what is possible, not only possible but previously thought impossible.
* Collective human geniuses can create unthinkable technologies.
* So AI is real and possible and it will keep getting better.
* Most of AI based products will help us in the long run but will make use lazier than we already are.
* AI will destroy old products/markets and create new ones in design and maintenance. Because robots and computers are machines and they will always break.
* Dare I say, machines will NEVER be able to repair/create themselves to the degree of the terminator movie.
* AI will always lack ability to think independently like human.

These are some bold prediction so I am ready to take the beating!

[imoldernu]

Quote:

Originally Posted by pjigar I have mixed thoughts on AI. I am an electrical engineer and work in field of semiconductor. Some predictions from my years of experience in semiconductor and associated software industry: * We have time and time again beat ourselves with what is possible, not only possible but previously thought impossible. * Collective human geniuses can create unthinkable technologies. * So AI is real and possible and it will keep getting better. * Most of AI based products will help us in the long run but will make use lazier than we already are. * AI will destroy old products/markets and create new ones in design and maintenance. Because robots and computers are machines and they will always break. * Dare I say, machines will NEVER be able to repair/create themselves to the degree of the terminator movie. * AI will always lack ability to think independently like human. These are some bold prediction so I am ready to take the beating.

Yes... I agree, and you've put into words, many of my own thoughts.

But... some recent exposure to the "impossible" is giving me second thoughts about "independent thought... goot think that one through. Think "Hal"...

In the OP, I mentioned the "fake news"... I don't know where to find it again, but here's how it went.. First a Twitter message that was hard right... when exposed to the AI program, based on the words and sequence, changed the meaning to a negative postulate, and added words and sentences that put an entirely different meaning on the subject.

Another... that I got close to because I have a blind friend my age, in our CCRC... Found (On AARP) a set of glasses that when worn, would allow the blind to "read" a book... just by pointing to the pages. Basically OCR, converted to artificial voice ala Alexa or Siri... also includes object recognition, face recognition and distance calculations. Available now. No... not yet perfect, but not too far away.

And... one more that's here today... For Echo users, try this... "Alexa... Let's talk", and when she tries to use her preferred subject, (movies)... you disagree, and change it to something else... I changed the subject to "pets" and we had a discussion on that.

As to the OP... One of the subjects that I left off, was "How far we've come, in five, ten, or twenty years". Am putting together a long list of things that have become second nature today, but back then, would have laughed me out of the room. AI... a new term, but what has been going on for some time.
Think "Television" How many here on ER don't remember when this was a "miracle".
How far we've come in 20 years, compared to the 200,000+ years we've been around.

[imoldernu]

Quote:

Originally Posted by GrayHare Much of the expected value of AI is in its eventual ability to predict or forecast in situations too complex for humans to fully comprehend, such as weather forecasting and other probabilistic systems. IMO AI won't become viable for these areas until quantum computing, itself probabilistic, becomes practical.

Just one small area of then,now, and soon.
How about the wide subject of medicine... Am thinking:
Genes... predictive medical issues.
At home doctor's visits... simple diagnostic machines... as, Holter Harnesses, blood analysis ala the new ones worn on the back of the arm... pulse, blood, pressure, heartbeat and like that. Then, eventually, as manufacturing progresses, home "scans"...
Then more of what we already have, but fully automatic... weight, height, breathing sensors, and even sleep analysis mechanisms.
Already... three mouse clicks, and I have my entire health history, going back 20 years and more... complete with diagnosis, outcomes, medications and all Er and hospital visits.

In many ways, it seems as if we're in the same place as Einstein was... when "relativity" wasn't on the horizon.

[txtig]

Quote:

Originally Posted by ckelly78z Does anyone else find it ironic that commercial adds for specific items show up on your computer after discussing them in privately your living room ?

If this were happening to me, I’d be extremely concerned. But, I haven’t experienced this (yet).

I have experienced receiving a targeted ad for a little-known app shortly after that same app was mentioned in an email that I received and immediately deleted. Not happy about that.

[NW-Bound]

Quote:

Originally Posted by GrayHare ... Much of the expected value of AI is in its eventual ability to predict or forecast in situations too complex for humans to fully comprehend, such as weather forecasting and other probabilistic systems. IMO AI won't become viable for these areas until quantum computing, itself probabilistic, becomes practical.

Much of our physical world is a chaotic system. No amount of computing will be able to predict the far future.

See: https://en.wikipedia.org/wiki/Butterfly_effect.

On top of that, Heisenberg principle means that even if we have a computer that can track and model every single atom in this universe, we still cannot predict the future.

PS. Any computer, no matter what technology, will consist of many atoms. It therefore can never truly model itself (including failure modes), let alone the world external to itself.

[imoldernu]

I'm sure most of you know about this, but the scope of this article on hacking/tracking, has pulled a lot of pieces together for me. While most of the information is already known, it seems as if the public takes it in stride, though I doubt that most people realize that the term "Privacy" has a different meaning in the "new world of tech".

https://www.pressreader.com/usa/chic...81698321057880

From the Chicago Tribune article "Campaign Technology gets Orwellian" on March 8.

"Old Hat" for you, but my bet is that 90% of the public has no idea of the depth of the potential influence. It makes the idea that "Alexa is listening"... is but a small part of what is here and now.

[JoeWras]

Currently, what we are seeing is more a result of algorithmic usage of Big Data. The ads for stuff you supposedly only talked about, for instance.

I find some of it useful. I find much of it disturbing.

I actually don't mind the occasional targeted ad. I do mind being put into a box of religious persuasion, race, political thought, etc. Just like you get the occasional goofy targeted ad that doesn't describe you, you may also be put into a box you should not be in. This is dangerous.

[dtbach]

* Dare I say, machines will NEVER be able to repair/create themselves to the degree of the terminator movie.
* AI will always lack ability to think independently like human.


Smart guys like Hawkings and Musk aren't so confident as you are. We should not be afraid of AI, but we better be careful the farther along it gets.

[MichaelB]

Quote:

Originally Posted by dtbach * Dare I say, machines will NEVER be able to repair/create themselves to the degree of the terminator movie. * AI will always lack ability to think independently like human.

That’s what the machines want you to think. They said the same thing about Skynet.

[imoldernu]

Back with one more... defining the privacy pirates and what it means to us.

https://www.axios.com/newsletters/ax...ves&stream=top

Did you know that half the country was affected by the Equifax data breach?

Scary... How many times have you been personally affected by this kind of invasion?

[NW-Bound]

Quote:

Originally Posted by MichaelB That’s what the machines want you to think. They said the same thing about Skynet.

Maybe the Tesla autopilots that drove the cars under semitrailers, or into parked firetrucks, did it on purpose to kill the car owners, because the latter badmouthed them or did something to upset them.

Tesla owners would better be careful, and never cuss their autopilot. At least, not until they can program the instinct of self-preservation into these autopilots.

PS. Oops. Musk promised to have the car drive itself around the parking lot to find the owner at the store entrance. The car can just run him over. We are doomed. Won't be long now until we read stories about a Tesla chasing its owner inside a Walmart, toppling shelves of merchandise.

[ziggy29]

Quote:

Originally Posted by NW-Bound Tesla owners would better be careful, and never cuss their autopilot. At least, not until they can program the instinct of self-preservation into these autopilots.

When they can program that instinct, you'll know you have a machine that passes the Turing Test. Especially if it needs an underwear change after a near miss.