Is there such a thing as AI?

[Kaylla]

My kid brother is studying math and computer science at grad school level - and he keep going on about how all this talk about AI is bullshit.

According to him, all there is a bunch of statistics backed up by lots of computing power - using algorithms that have been around since the 1960s. There is, in his view, nothing new or special about AI except that the old algorithms have gotten faster.

I am the dumbass of the family, having studied business and horticulture at community college level (Associate Degree in each), so while what he says sounds reasonable, I have no means to evaluate this claim myself.

What do smarter people think?

[Impenitent]

smarter people think thoughts just like everyone else...

-Imp

[Atla]

Imo just until recently, AI used to mean artificial intelligence. Well that would be an artificial, intelligent entity. Similar to humans but a machine. Maybe even self-aware, self-conscious. Like Data from Star Trek, or the Normandy's computer from ME.

We still have no idea how to create one of those, even in theory. Do you just slam enough processing power together and bam it happens? Are there other requirements? Maybe not even possible with the materials currently used? Maybe we'll know in 50-100 years. AI nowadays seems to mean anything that you can use as a search engine, and its responses are human-like enough.

[godelian]

In the end, it is all just computation:

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

The AI effect is the discounting of the behavior of an artificial-intelligence program as not "real" intelligence.[1]

The author Pamela McCorduck writes: "It's part of the history of the field of artificial intelligence that every time somebody figured out how to make a computer do something—play good checkers, solve simple but relatively informal problems—there was a chorus of critics to say, 'that's not thinking'."[2]

Researcher Rodney Brooks complains: "Every time we figure out a piece of it, it stops being magical; we say, 'Oh, that's just a computation.'"[3]

"The AI effect" refers to a phenomenon where either the definition of AI or the concept of intelligence is adjusted to exclude capabilities that AI systems have mastered. This often manifests as tasks that AI can now perform successfully no longer being considered part of AI, or as the notion of intelligence itself being redefined to exclude AI achievements.

McCorduck calls it an "odd paradox" that "practical AI successes, computational programs that actually achieved intelligent behavior were soon assimilated into whatever application domain they were found to be useful in, and became silent partners alongside other problem-solving approaches, which left AI researchers to deal only with the 'failures', the tough nuts that couldn't yet be cracked."[5] It is an example of moving the goalposts.[6]

Tesler's Theorem is: AI is whatever hasn't been done yet.

Michael Swaine reports "AI advances are not trumpeted as artificial intelligence so much these days, but are often seen as advances in some other field".

Once you understand how something works, it stops being "magical". At that point, it becomes just algorithms that are crunching data.

AI devalues itself but it also devalues large swaths of human intelligence. In the middle ages, they used to think of someone who could carry out a long division as some kind of highly-learned magician with superpowers. He could do something that no one else could! He had access to some secret esoteric knowledge akin to witchcraft!

Not so long ago, someone who could manipulate a computer by using a keyboard was someone considered to be a highly skilled individual, who could do something that others could only dream of.

[accelafine]

From what I've seen I think your brother is probably right. The grammar has got better but AI still reeks 'fairly sophisticated google search responder' and nothing more. No more 'AI' than a pocket calculator.

Still, you do have to wonder why so many people who appear to be 'computer/AI experts' are so scared of it.

To me, AI just comes across as a 'soup' of all the information and crap that's swirling on the internet.

I asked it the other day (on google search) something innocent about pregnancy scans determining the baby's sex and it said this: ''Other considerations: It's important to keep in mind that the baby's gender may not match the sex they were assigned at birth.'' So AI can seriously just go fuck itself.

[ThinkOfOne]

My kid brother is studying math and computer science at grad school level - and he keep going on about how all this talk about AI is bullshit.

According to him, all there is a bunch of statistics backed up by lots of computing power - using algorithms that have been around since the 1960s. There is, in his view, nothing new or special about AI except that the old algorithms have gotten faster.

I am the dumbass of the family, having studied business and horticulture at community college level (Associate Degree in each), so while what he says sounds reasonable, I have no means to evaluate this claim myself.

What do smarter people think?

One needs to gain a minimum understanding of a topic to be able to evaluate it correctly. Seems like your brother has yet to gain that understanding. Of course, that's assuming that you understand enough to accurately describe your brother's position.

That said, it does seem like much of what's being touted as "AI" is being way oversold to make it sound more impressive than it is. Such "AI" is BS: "We're using AI"; "We're teaching AI".

But, "Yes Virginia, there is such a thing as AI".

[attofishpi]

My kid brother is studying math and computer science at grad school level - and he keep going on about how all this talk about AI is bullshit.

According to him, all there is a bunch of statistics backed up by lots of computing power - using algorithms that have been around since the 1960s. There is, in his view, nothing new or special about AI except that the old algorithms have gotten faster.

I am the dumbass of the family, having studied business and horticulture at community college level (Associate Degree in each), so while what he says sounds reasonable, I have no means to evaluate this claim myself.

What do smarter people think?

I'm not smarter people (I am still that dude U think smoked way too much sage)

Your brother is on the money, mostly.

What's important about A.I. currently is that it can collate and reorganise information as we want it to be provided, be it in precis form, synopsis form..etc..etc..in any form we desire as sentient intelligent beings.

The statistical ---> next probable/likely thing a 'human' would do/think is precisely how an AI is geared to mimic US HUMANS...ergo...emulated intelligence. (nothing wrong with that --> cold logic --> A.I.) https://www.androcies.com/alphatwo.php



PS: RE your thread question " Is there such a thing as AI?" -- yes there is, it's an emulation of how we think. (cold-logic nonetheless - www.androcies.com)

[Skepdick]

According to him, all there is a bunch of statistics backed up by lots of computing power - using algorithms that have been around since the 1960s. There is, in his view, nothing new or special about AI except that the old algorithms have gotten faster.

Yes, what passes for "AI" today is just a stochastic parrot.

That's all fine and dandy - there's nothing "new" or "special" about it - it does what it does. It works however it works.

The question isn't whether AI is special. The question is whether there's much more to our own, human learning abilities?
Are we just more sophisticated stochastic parrots with greater computational power?

What is it that sets us appart from pattern-matching automatons which learn by repeated exposure to information?

I can think of at least one difference. While the LLMs are competent at using language (because they are language models after all) they aren't so good at generalizing and turning this into practical know-how.

In philosophy this is often referred to as the "symbol grounding problem" or the challenge of "embodied cognition."

https://en.wikipedia.org/wiki/Symbol_grounding_problem
https://en.wikipedia.org/wiki/Embodied_cognition

But in ancient wisdom this is simply the difference between being good at talking and good at doing. They can talk the talk, but they can't walk the walk.

[attofishpi]

⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀
⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀
⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀Skepdick
⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀pomppepdick
⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀stickelepdick
⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀fifepdick⠀⠀
⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀fifepdick
⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀stickelepdick
⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀that's how you spell Skepdick
⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀
⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀
⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀
⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀
⠀⠀⠀⠀⠀⠀⠀⠀⠀⠀

[Gary Childress]

According to him, all there is a bunch of statistics backed up by lots of computing power - using algorithms that have been around since the 1960s. There is, in his view, nothing new or special about AI except that the old algorithms have gotten faster.

Yes, what passes for "AI" today is just a stochastic parrot.

That's all fine and dandy - there's nothing "new" or "special" about it - it does what it does. It works however it works.

The question isn't whether AI is special. The question is whether there's much more to our own, human learning abilities?
Are we just more sophisticated stochastic parrots with greater computational power?

What is it that sets us appart from pattern-matching automatons which learn by repeated exposure to information?

I can think of at least one difference. While the LLMs are competent at using language (because they are language models after all) they aren't so good at generalizing and turning this into practical know-how.

In philosophy this is often referred to as the "symbol grounding problem" or the challenge of "embodied cognition."

https://en.wikipedia.org/wiki/Symbol_grounding_problem
https://en.wikipedia.org/wiki/Embodied_cognition

But in ancient wisdom this is simply the difference between being good at talking and good at doing. They can talk the talk, but they can't walk the walk.

There is one thing that can (at least in theory) always separate a computational "machine" from a "human being". A human being presumably has "experiences" and machines presumably do not. For example, when I remove my hand from a hot stove to avoid being burned, I do so because I feel pain. A machine presumably removes its hand from a hot stove because it is programmed to do so, not because it "feels" pain.

The 64 million dollar questions become, can a computer be programmed to "feel" and how would a programmer go about programming "feeling" and if a programmer created a computer that "feels" pain, how would the programmer be able to discern what "pain" looks like in order to verify that a particular computer has it (or is "feeling" such)?

[Gary Childress]

You said:
Hi, ChatGPT. Do you ever feel happy or sad?

ChatGPT said:
ChatGPT
Hi! I don't experience emotions like happiness or sadness because I'm just a program designed to assist you. However, I can simulate empathy or cheerfulness in my responses to make our interactions feel more natural and supportive. How can I help you today?

[Gary Childress]

You said:
Is it possible to program a computer to have real emotions?

ChatGPT said:
ChatGPT
The idea of programming a computer to have real emotions is both a fascinating and deeply complex question. It touches on fields like artificial intelligence (AI), neuroscience, philosophy, and psychology. Let’s break it down:

1. What Are "Real" Emotions?
For humans, emotions are:

Biological: Rooted in the brain’s chemical and neural activity.
Subjective: Involve conscious feelings and self-awareness.
Adaptive: Serve evolutionary purposes, like fostering social bonds or triggering survival responses.
If by "real emotions" we mean a conscious, subjective experience akin to what humans feel, this would require a machine to have a level of consciousness, which is still far from being understood or achievable.

2. Can Computers Simulate Emotions?
Yes, computers can simulate emotions to a convincing degree:

AI Models: Programs can be trained to detect human emotions (e.g., facial recognition or tone analysis) and respond in ways that appear empathetic or emotional.
Applications: Emotional simulations are already used in customer service chatbots, companion robots, and even therapeutic AI tools like virtual therapists.
However, these simulations are not "felt" by the machine. They're based on algorithms, patterns, and programmed responses.

3. Could Computers "Feel" Emotions Someday?
This depends on whether emotions can be replicated outside of biology:

Neuromorphic Engineering: Efforts to build AI systems that mimic the structure of the human brain could lead to machines that process information more like humans, potentially enabling more advanced emotional behavior.
Artificial Consciousness: If scientists can one day understand and replicate consciousness, it’s theoretically possible that machines could "feel" emotions as a byproduct. This remains speculative and highly controversial.
Synthetic Biochemistry: Some futurists envision combining AI with biological systems (e.g., artificial neurons or biochemical pathways) to create machines with emotion-like states.

4. Philosophical Questions
What Is "Feeling"? Even if a machine mimics emotions perfectly, does that mean it "feels" them? Or are feelings inherently tied to organic life and subjective experience?
Ethics: If machines could feel emotions, would they deserve rights? How would we ensure ethical treatment?

Conclusion
Currently, computers and AI can only simulate emotions, not experience them. Developing "real" emotions in machines would require breakthroughs in understanding consciousness and emotion, and even then, it would raise profound philosophical and ethical questions. While intriguing, this possibility remains largely theoretical for now.

[Skepdick]

There is one thing that can (at least in theory) always separate a computational "machine" from a "human being". A human being presumably has "experiences" and machines presumably do not. For example, when I remove my hand from a hot stove to avoid being burned, I do so because I feel pain. A machine presumably removes its hand from a hot stove because it is programmed to do so, not because it "feels" pain.

Distinction without a difference if you ask me.

You are just describing a stimulus-response in a manner which distracts from the fact that pain-response is your programming.

You can program a computer to do exactly that also. The CPU's regular operation will be interrupted by something very very urgent/imortant.
And this signal will be handled in some manner - remove hand. Increase cooling. Whatever.

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

[Gary Childress]

There is one thing that can (at least in theory) always separate a computational "machine" from a "human being". A human being presumably has "experiences" and machines presumably do not. For example, when I remove my hand from a hot stove to avoid being burned, I do so because I feel pain. A machine presumably removes its hand from a hot stove because it is programmed to do so, not because it "feels" pain.

Distinction without a difference if you ask me.

You are just describing a stimulus-response in a manner which distracts from the fact that pain-response is your programming.

You can program a computer to do exactly that also. The CPU's regular operation will be interrupted by something very very urgent/imortant.
And this signal will be handled in some manner - remove hand. Increase cooling. Whatever.

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

But does the cooling increase take place because the computer feels pain? Or is it no different than a robotic arm that is connected to a thermostat to trigger levers and what not to remove it's arm from a heat source that would destroy the arm? Presumably there's something different between a mechanical device that is rigged deterministically to remove itself from destruction verses a human being who removes himself or herself from destruction out of pain.

[Skepdick]

But does the cooling increase take place because the computer feels pain? Or is it no different than a robotic arm that is connected to a thermostat to trigger levers and what not to remove it's arm from a heat source that would destroy the arm? Presumably there's something different between a mechanical device that is rigged deterministically to remove itself from destruction verses a human being who removes himself or herself from destruction out of pain.

If there's any difference then what is it?

There's a medical condition which makes one insensitive to pain. So when a certain part of your brain/hardware is turned off - you don't feel pain either.

https://en.wikipedia.org/wiki/Congenita ... ty_to_pain