Pandora's Box....

[Chewbacca]

Originally Posted by Blansky

And we'd expect nothing less.......ya big ole fur ball.

It’s not too late for you! They’ll need many henchman. I’ll put in a good word.

[~JJ]

Originally Posted by Blansky

I'm not really sure if man-made gardens can really be compared to unspoiled nature.

Obviously we can make gardens, with maintained plants, weed killer, clear water features and artificial layering of species but it is ......artificial.

If we left it unmaintained for a year it would no longer look like that.

It's pretty but it's artificial. Comparing nature to it seems contrived.

We can basically tart up anything and make it pretty to some. But humans have always tried to "control" their surroundings.

Not sure how we got from AI to this. These threads do meander.

^well said.

Just today a coworker showed me how he’s using AI do create his notes at work… basically, type the keywords in, then tell it to create notes and it provides a complete bulleted list with a summary of next steps.

[EEpro]

Originally Posted by ~JJ

^well said.

Just today a coworker showed me how he’s using AI do create his notes at work… basically, type the keywords in, then tell it to create notes and it provides a complete bulleted list with a summary of next steps.

Is the company ok with him uploading details like this to offsite servers ?

[Maleg]

Originally Posted by Blansky

I'm not really sure if man-made gardens can really be compared to unspoiled nature.

Obviously we can make gardens, with maintained plants, weed killer, clear water features and artificial layering of species but it is ......artificial.

If we left it unmaintained for a year it would no longer look like that.

It's pretty but it's artificial. Comparing nature to it seems contrived.

We can basically tart up anything and make it pretty to some. But humans have always tried to "control" their surroundings.

Not sure how we got from AI to this. These threads do meander.

Another gloom and doom, sky is falling, pessimist. Finding fault in every intention is inherently toxic. Take the time to stop and smell the flowers. Life will be sweeter for it.

[East Bay Rider]

Originally Posted by Blansky

I'm not really sure if man-made gardens can really be compared to unspoiled nature.

Obviously we can make gardens, with maintained plants, weed killer, clear water features and artificial layering of species but it is ......artificial.

If we left it unmaintained for a year it would no longer look like that.

It's pretty but it's artificial. Comparing nature to it seems contrived.

We can basically tart up anything and make it pretty to some. But humans have always tried to "control" their surroundings.

Not sure how we got from AI to this. These threads do meander.

It's the arrogance of man. Nature, but better.
Parks: since we've destroyed all the nature for miles, here's a manicured nature-like space to satisfy you. The selection that we've provided represents many varieties of things you might experience in actual nature delivered here in this condenced space to maximize your nature experience in the shortest time possible. Please enjoy the show and snap some pictures before you go back to the concrete world to live and consume.

[GB-man]

[m j b]

Originally Posted by enjoythemusic

If aliens come down and abduct me, I'm going to consider it a rescue mission away from this planet.

We'll miss you Steven!

Originally Posted by Blansky

For the second time in human history we have "invented" something that could/will likely destroy us.

What was the first time?

No, I'm serious, I have no idea what you are referencing.



Anyway...

As a former programmer and current manager of programmers, it's really not a realistic thing that AI can "write code". Sure, you can have a computer program write other computer programs, heck, we've had that technology for years. It generates crap.

And most "real" computer programs are far too complex and integrated and have all kinds of special conditions and such that by the time you explain all of it to some kind of "bot", you could have written it yourself.

But hey, go ahead and tell your kids to drop that STEM degree and go to a technical school and learn plumbing. Because regardless of all of our other improvements in technology and society, we still have to poop the old-fashioned way.

[Krash]

Originally Posted by East Bay Rider

It's the arrogance of man. Nature, but better.
Parks: since we've destroyed all the nature for miles, here's a manicured nature-like space to satisfy you. The selection that we've provided represents many varieties of things you might experience in actual nature delivered here in this condenced space to maximize your nature experience in the shortest time possible. Please enjoy the show and snap some pictures before you go back to the concrete world to live and consume.

Sorry, I'm not just as miserable when it comes to humanity.

[EEpro]

Originally Posted by m j b

What was the first time?

No, I'm serious, I have no idea what you are referencing.

Hydrogen bomb

[Maleg]

Originally Posted by East Bay Rider

It's the arrogance of man. Nature, but better.
Parks: since we've destroyed all the nature for miles, here's a manicured nature-like space to satisfy you. The selection that we've provided represents many varieties of things you might experience in actual nature delivered here in this condenced space to maximize your nature experience in the shortest time possible. Please enjoy the show and snap some pictures before you go back to the concrete world to live and consume.

Life in your city must be pretty dire to generate this level of cynicism.

[austinp]

Golf courses are pretty.


Sent from my iPhone using Tapatalk

[Knappo 1307]

Originally Posted by East Bay Rider

It's the arrogance of man. Nature, but better.
Parks: since we've destroyed all the nature for miles, here's a manicured nature-like space to satisfy you. The selection that we've provided represents many varieties of things you might experience in actual nature delivered here in this condenced space to maximize your nature experience in the shortest time possible. Please enjoy the show and snap some pictures before you go back to the concrete world to live and consume.

Wow, sorry you feel this way. I'd hate to live my life this way. Hope you can find some peace and happiness somewhere.

[EEpro]

Originally Posted by Knappo 1307

Wow, sorry you feel this way. I'd hate to live my life this way. Hope you can find some peace and happiness somewhere.

Yes it's sad. Nietzsche saw it coming over 100 years ago and has some ideas on emerging from it.

[Blansky]

Originally Posted by maleg

another gloom and doom, sky is falling, pessimist. Finding fault in every intention is inherently toxic. Take the time to stop and smell the flowers. Life will be sweeter for it.

lol.

[Blansky]

Originally Posted by m j b

What was the first time?

No, I'm serious, I have no idea what you are referencing.





As a former programmer and current manager of programmers, it's really not a realistic thing that AI can "write code". Sure, you can have a computer program write other computer programs, heck, we've had that technology for years. It generates crap.

And most "real" computer programs are far too complex and integrated and have all kinds of special conditions and such that by the time you explain all of it to some kind of "bot", you could have written it yourself.
.

The first was the atomic bomb.

As for AI, this link talks about it's capabilities. But it's also in it's infancy.

https://newatlas.com/technology/chat...ternet-access/

This talks about the people and resources behind it's explosion onto the scene. https://time.com/6255952/ai-impact-c...d-2148d7788cb7

There are a lot of people, even Elon Musk (who I don't care for) who are saying, hold on here, we need to look at this, think of the repercussions.

https://www.bbc.com/news/technology-65110030

https://www.nytimes.com/2023/03/29/t...usk-risks.html

Just because we can do something, doesn't always mean we should.

[Maleg]

Originally Posted by Blansky

lol.

Insightful reply.

Your posts on this thread really are quite cynical and negative. I hope this doesn’t reflect your true nature. That would be quite sad.

[Blansky]

Originally Posted by Maleg

Insightful reply.

Your posts on this thread really are quite cynical and negative. I hope this doesn’t reflect your true nature. That would be quite sad.

Thank you for the psychoanalysis.

I find the head in the sand about the dangers of AI absolutely amazing, as if it's just another Apple iPhone release coming down the pike.

Wow, that's pretty. LOL

[Maleg]

Originally Posted by Blansky

Thank you for the psychoanalysis.

I find the head in the sand about the dangers of AI absolutely amazing, as if it's just another Apple iPhone release coming down the pike.

Wow, that's pretty. LOL

Several militaries have been using advanced AI in deployed systems for a few years. If the technology was inherently dangerous we’d be serving that master already.

I don’t put a lot of stock in Elon’s AI lobbying either. He’s always working what’s best for Elon.

[330ci]

Originally Posted by Maleg

Several militaries have been using advanced AI in deployed systems for a few years. If the technology was inherently dangerous we’d be serving that master already.

I don’t put a lot of stock in Elon’s AI lobbying either. He’s always working what’s best for Elon.

"Why are you so nihilistic, this technologies already proven its great in the context of killing people."

Personally I find a ton of problems with society but it doesn't detract from my love of life.

Not sure where all the Ad hominem attacks are coming from when the best use case for the technology thus far is a destructive force.

From the bottom of the Marianas trench, to the top of mt Everest. There are signs of humanity at every point along the way.

Deforestation, increasing greenhouse gasses, enslaving people in third world countries so we can enjoy our life in the first world, fracking causing earthquakes where previously there weren't, fertizilizer runoff destroying water supplies across the world, sewage runoff from cesspools destroying water supplies as well. Multiple nuclear wastelands, killing off entire populations of animals from native areas due to oil spills, you can't go anywhere in the US without being more than a few miles from a main road. Even the state parks are all littered with trash from city tourists coming out to go walk five minutes to post 20 photos "for the gram" and leaving their snack wrappers everywhere as they have no respect for the nature itself.

Even the nature we get to experience isn't "natural". The bulk of trees in Mi state parks were planted during government funding projects, theres information placards along the trails to "aid" trail users in identification, the brush is regularly burned out with controlled burns, animal populations are culled, lakes and ponds are stocked, rivers are damned to control flow.

Nature is preserved in select areas and magnified to "protect its beauty" in others. Whether you're going through an African nature preserve or government protected land in the US, its all been stripped clean and rebuilt.

If zoos and botanical gardens are the epitome of natural beauty; we're screwed. Following human logic, AI is going to keep us in cages for entertainment if we're lucky, and will annihilate us like we're ants in the way of progress at the worst.

Just because it hasn't figured out how to eliminate us yet, doesn't mean it won't reach that as an objective in the future.

The fact that they're not using this tech to stop world hunger and its doing a good job in military applications seems like great reason to be apprehensive of the tech amongst other reasons.

Humans are totally a cancer to the planet. I have little doubts AI will come to a different conclusion. Hopefully they have a less nihilistic view of humanity and feel like working towards a solution that includes humans in their plan. Seems like alot of extra steps when humans generally just eliminate any problems we find.

[Cloudchaser]

Originally Posted by m j b

We'll miss you Steven!



What was the first time?

No, I'm serious, I have no idea what you are referencing.



Anyway...

As a former programmer and current manager of programmers, it's really not a realistic thing that AI can "write code". Sure, you can have a computer program write other computer programs, heck, we've had that technology for years. It generates crap.

And most "real" computer programs are far too complex and integrated and have all kinds of special conditions and such that by the time you explain all of it to some kind of "bot", you could have written it yourself.

But hey, go ahead and tell your kids to drop that STEM degree and go to a technical school and learn plumbing. Because regardless of all of our other improvements in technology and society, we still have to poop the old-fashioned way.

But AI isn't something static that only responds to human inputs. It combines human inputs with its own "(machine) learning" to constantly improve its data.

Using that logical model, it can write code, does write code, and will continue to write code, theoretically until it writes it better than any human could. Along that progression, the human side is needed less and less.

Of course, maybe humans' eagerness to feed it everything we've ever learned in human history in a short period of time could shorten that span of needing human inputs at all.

That's the basic concern more than is "how good is its programming code right now?"

[Cloudchaser]

During multiple of my ChatGPT sessions this week, the AI has produced a list of events where one or some of those events did not actually happen, but purportedly occurred in 2022 according to ChatGPT. When corrected, ChatGPT "apologizes." As part of that apology, ChatGPT is sure to remind me that its knowledge cutoff date is actually 2021.

When further asked why it is including erroneous data from 2022 if the knowledge cutoff is 2021, it repeats the same apology without further explanation.

[Blansky]

As I've said before, none of this affects me because I'll be dust in 10 or so years. But if I was in my 20s or 30s and/or had kids, I'd be screaming WHAT THE HELL.

Lets think this stuff through, instead of leaving it to a bunch of nerds and money grubbing idiots that only care about making more money, to see if this needs some regulation and safeguards.

The tech that's been in the last 20 years is happening so fast, that it's upon us before we can really think it through. Social media turned out to be a sewer that also subject to foreign infiltration.

Instead of thinking how new tech is gonna be so fun, we need to look at how it affect jobs, information, democracies, etc etc.

I can remember in the 50s and 60s, the Jetson's era (futurists) how tech is going to make life easier, less work, more free time with family, maybe no work at all because robots will do it. Sounds great. I'm in.

Except nobody bothered to mention how we were going to make money to live on, to enjoy all that great free time.

If anybody doesn't think that AI will, in the near future, be writing and executing programs, be in charge of infrastructure, air traffic, nuclear facilities, electrical grid, all aspect of finance/money, media distribution etc etc is in denial. We will become more and more dependent on this stuff and it never sleeps.

[Krash]

Originally Posted by Blansky

As I've said before, none of this affects me because I'll be dust in 10 or so years. But if I was in my 20s or 30s and/or had kids, I'd be screaming WHAT THE HELL.

Lets think this stuff through, instead of leaving it to a bunch of nerds and money grubbing idiots that only care about making more money, to see if this needs some regulation and safeguards.

The tech that's been in the last 20 years is happening so fast, that it's upon us before we can really think it through. Social media turned out to be a sewer that also subject to foreign infiltration.

Instead of thinking how new tech is gonna be so fun, we need to look at how it affect jobs, information, democracies, etc etc.

I can remember in the 50s and 60s, the Jetson's era, how tech is going to make life easier, less work, more free time with family, maybe no work at all because robots will do it. Sounds great. I'm in.

Except nobody bothered to mention how we were going to make money to live on, to enjoy all that great free time.

If anybody doesn't think that AI will in the near future, be writing and executing programs, be in charge of infrastructure, air traffic, nuclear facilities, electrical grid, all aspect of finance/money, media distribution etc etc is in denial. We will become more and more dependent on this stuff and it never sleeps.

I agree. AI is something that we should be concerned about. Also, we shouldn't focus on what AI is today, we should focus on what AI will be in 5 to 10 years.

With that said, I'm not on the "humanity is horrible" bandwagon here...Sorry, I just can't wallow in any of that misery.

[Blansky]

Originally Posted by Krash

With that said, I'm not on the "humanity is horrible" bandwagon here...Sorry, I just can't wallow in any of that misery.

I don't think that was a part of the thread.

[East Bay Rider]

Originally Posted by Cloudchaser

During multiple of my ChatGPT sessions this week, the AI has produced a list of events where one or some of those events did not actually happen, but purportedly occurred in 2022 according to ChatGPT. When corrected, ChatGPT "apologizes." As part of that apology, ChatGPT is sure to remind me that its knowledge cutoff date is actually 2021.

When further asked why it is including erroneous data from 2022 if the knowledge cutoff is 2021, it repeats the same apology without further explanation.

It's lying like a teenager.
That's creepy in a sci-fi way.

[EEpro]

Originally Posted by East Bay Rider

It's lying like a teenager.
That's creepy in a sci-fi way.

Twilight zone style for sure.

[Maleg]

Originally Posted by Cloudchaser

During multiple of my ChatGPT sessions this week, the AI has produced a list of events where one or some of those events did not actually happen, but purportedly occurred in 2022 according to ChatGPT. When corrected, ChatGPT "apologizes." As part of that apology, ChatGPT is sure to remind me that its knowledge cutoff date is actually 2021.

When further asked why it is including erroneous data from 2022 if the knowledge cutoff is 2021, it repeats the same apology without further explanation.

Surely the system has seen 2022 questions before. If it was truly learning it would have offered a caveat about its limitations prior to offering its answer rather than an apology after getting caught giving a bad answer.

It is not analyzing. It’s regurgitating it’s database in a document writer.

[Cloudchaser]

Originally Posted by Maleg

Surely the system has seen 2022 questions before. If it was truly learning it would have offered a caveat about its limitations prior to offering its answer rather than an apology after getting caught giving a bad answer.

It is not analyzing. It’s regurgitating it’s database in a document writer.

Well, it is actually a language model and not a logic model, so that is somewhat to be expected.

However, in my month of using it, there have been many instances of what can best be described as inference on its part. And when that happens, it sure reads and "feels" like supposition, or in some cases even defensiveness. That said, it is tremendously difficult to not poke at it with an expectation of logic being returned; I can't help but think sometimes what an outside observer would think in real time (basically, the Turing Test).

The example:

I began a conversation asking about a particularly anomalous series of events. ChatGPT began with the list of the sequence of those events, but the last included was from 2022, and hadn't actually happened. ChatGPT stated, "I apologize for the confusion," and then referred to the "knowledge cutoff" of 2021. After another erroneous list, it finally got it right.

I then asked ChatGPT to find another series of events that approached or equaled the statistical rarity of the first (within a specific industry over the last 100 years) ChatGPT replied with only one comparable example.

However, without any further question, ChatGPT expanded its answer to include further information that the reasons for the two (wholly unrelated) distinct series were not the same, in a way that most definitely read like a proactive, and invalidating, defense of what it predicted that I would suppose next.

Incidentally, you can ask ChatGPT to apply Occam's razor in any scenario (and it's fun), which definitely requires some element of gray area that sits close to actual logic.I asked for 10 applications of Occam's razor to the first series of events, and none of the applications resulted in an answer positing malfeasceance by authority figures.

That result had, bear with me here, an anti-anti-authoritarian bias, given that actual logic would probably require malfeascence to have been in the top 5 applications. Maybe the appearance of logic isn't actual logic. Maybe inputs precluding anti-authoritarianism have been programmed in. I don't know...I'm just trying to use it to learn about it.

But it seems a very fine line between "predictive" ("generative predictive transformer," afterall) and genuine logic. I haven't tried anything yet, but I'm guessing ChatGPT might be used to outline the best possible "malicious compliance" strategies. That's some Bladerunner type applications there, but something HAL 9000 might proactively shut down.

Point is, really, that it's not quite what people assume it is, even those using it. And it's changing quickly!

[Yobber76]

Daisy, Daisy, give me your answer do.........

Hal 1000

BTW, does anyone else think it's a little odd when we get into these discussions about man's hubris, greed, and natural tendency to destroy everything it touches, and the next thread will probably be about does your multi thousand dollar watch match your multi multi thousand dollar car?

[EEpro]

Originally Posted by Cloudchaser

Well, it is actually a language model and not a logic model, so that is somewhat to be expected.

However, in my month of using it, there have been many instances of what can best be described as inference on its part. And when that happens, it sure reads and "feels" like supposition, or in some cases even defensiveness. That said, it is tremendously difficult to not poke at it with an expectation of logic being returned; I can't help but think sometimes what an outside observer would think in real time (basically, the Turing Test).

The example:

I began a conversation asking about a particularly anomalous series of events. ChatGPT began with the list of the sequence of those events, but the last included was from 2022, and hadn't actually happened. ChatGPT stated, "I apologize for the confusion," and then referred to the "knowledge cutoff" of 2021. After another erroneous list, it finally got it right.

I then asked ChatGPT to find another series of events that approached or equaled the statistical rarity of the first (within a specific industry over the last 100 years) ChatGPT replied with only one comparable example.

However, without any further question, ChatGPT expanded its answer to include further information that the reasons for the two (wholly unrelated) distinct series were not the same, in a way that most definitely read like a proactive, and invalidating, defense of what it predicted that I would suppose next.

Incidentally, you can ask ChatGPT to apply Occam's razor in any scenario (and it's fun), which definitely requires some element of gray area that sits close to actual logic.I asked for 10 applications of Occam's razor to the first series of events, and none of the applications resulted in an answer positing malfeasceance by authority figures.

That result had, bear with me here, an anti-anti-authoritarian bias, given that actual logic would probably require malfeascence to have been in the top 5 applications. Maybe the appearance of logic isn't actual logic. Maybe inputs precluding anti-authoritarianism have been programmed in. I don't know...I'm just trying to use it to learn about it.

But it seems a very fine line between "predictive" ("generative predictive transformer," afterall) and genuine logic. I haven't tried anything yet, but I'm guessing ChatGPT might be used to outline the best possible "malicious compliance" strategies. That's some Bladerunner type applications there, but something HAL 9000 might proactively shut down.

Point is, really, that it's not quite what people assume it is, even those using it. And it's changing quickly!

It still prefers to allow use of a thermonuclear device to kill 10M people rather than prevent that event by using a racial slur... because those are harmful.

It has significant bias and can't weigh a moral dilemma properly.

It goes to show if we ever end up enslaved (or policed or judged or have our medical decisions decided) by these things that they have an incredibly difficult time weighing the magnitude of human suffering.