AI

Unfocussed Mike said

ANDY00 I am saying that if you want to understand it properly, you should tune out the CEOs and actually look at the research.

 

ANDY00 That's not a 30 second sound-bite from marketing folks, but from the people actually involved in the design and/or training of the various systems. As Unfocussed Mike points out, there is a lot of pop-culture coverage that's simply incorrect. False-to-fact. Wrong. And that's from many mainstream sites and publications in addition to the many people parroting what they think they heard from sources that were oversimplifying things to make it easier to "understand". (See "Lies to Children"--watered down overviews that aren't actually correct but might make something vaguely close to the truth sound kind of understandable by those without the background to understand the more complex reality.)

Whether it's "real AI", machine learning, large language models, generative AI, etc, there seems to be more incorrect statements posing as facts than there is accurate information: simplified or not.

Kevin Connery said

Unfocussed Mike said

ANDY00 I am saying that if you want to understand it properly, you should tune out the CEOs and actually look at the research.

 

ANDY00 That's not a 30 second sound-bite from marketing folks, but from the people actually involved in the design and/or training of the various systems. As Unfocussed Mike points out, there is a lot of pop-culture coverage that's simply incorrect. False-to-fact. Wrong. And that's from many mainstream sites and publications in addition to the many people parroting what they think they heard from sources that were oversimplifying things to make it easier to "understand". (See "Lies to Children"--watered down overviews that aren't actually correct but might make something vaguely close to the truth sound kind of understandable by those without the background to understand the more complex reality.)

Whether it's "real AI", machine learning, large language models, generative AI, etc, there seems to be more incorrect statements posing as facts than there is accurate information: simplified or not.

Look, I've been as honest as I can be from a personal point of view. I think these are intelligent systems (not sentient, but intelligent). It's not human or animal; it's artificial. So artificial intelligence is what it is, or the official term is ANI i believe (Artificial narrow intelligence) with the next stage in its evolution being AGI (Artificial general intelligence) Maybe that's not good enough for you guys—tough. That's my own personal opinion. You lot can call it what you like. It's allowing systems to find new antibiotics—fact. It's helping cure diseases—fact. It's deciphering animal language—fact. It can also be used for much more, from telling someone a recipe to controlling a billion-dollar F18, but this is just the beginning. Personally, I don't care what you call it. Call it Bob for all I care. I think it's an amazing step forward and will change life for generations in the future. i list and quote a load of experts in the field and im told well theyre all lying us lot on a photography site know better than them, now ive been very clear from the start im no expert, like most i follow what i read or come across on the internet including Google Russ, and its all saying this is AI and its developing fast, the only place where i see folk saying its not ai is here and if this site was called AI expert i would be much more receptive but its a photography site where people say dont listen to the rest of the planet its not really artificial intelligence its really dumb. well that dumb thing is doing some pretty intelligent stuff we couldn't before now :-) and i recon in future it will probably be doing a lot lot more in everything but remember when that system or robot takes your job its not AI its just a dumb fake system :-D

You know, just looking at what it can bring in simpler terms, like ChatGPT making assistants for the elderly or self-driving cars for those who can no longer drive, would make a huge difference in someone's life. That's not a gimmick to those people. The smart systems that can detect if your breathing is off or if your spending has changed in your bank—all AI, all pretty great steps forward

These new intelligent systems can make a massive difference to everyone from the huge corporation to the elderly person sitting alone, they can also be used for bad  purposes just like everything else but that the same with everything,

Edited by ANDY00

ANDY00 said

Kevin Connery said

Unfocussed Mike said

ANDY00 I am saying that if you want to understand it properly, you should tune out the CEOs and actually look at the research.

 

ANDY00 That's not a 30 second sound-bite from marketing folks, but from the people actually involved in the design and/or training of the various systems. As Unfocussed Mike points out, there is a lot of pop-culture coverage that's simply incorrect. False-to-fact. Wrong. And that's from many mainstream sites and publications in addition to the many people parroting what they think they heard from sources that were oversimplifying things to make it easier to "understand". (See "Lies to Children"--watered down overviews that aren't actually correct but might make something vaguely close to the truth sound kind of understandable by those without the background to understand the more complex reality.)

Whether it's "real AI", machine learning, large language models, generative AI, etc, there seems to be more incorrect statements posing as facts than there is accurate information: simplified or not.

Look, I've been as honest as I can be from a personal point of view. I think these are intelligent systems (not sentient, but intelligent). It's not human or animal; it's artificial. So artificial intelligence is what it is, or the official term is ANI i believe (Artificial narrow intelligence) with the next stage in its evolution being AGI (Artificial general intelligence) Maybe that's not good enough for you guys—tough. That's my own personal opinion. You lot can call it what you like. It's allowing systems to find new antibiotics—fact. It's helping cure diseases—fact. It's deciphering animal language—fact. It can also be used for much more, from telling someone a recipe to controlling a billion-dollar F18, but this is just the beginning. Personally, I don't care what you call it. Call it Bob for all I care. I think it's an amazing step forward and will change life for generations in the future. i list and quote a load of experts in the field and im told well theyre all lying us lot on a photography site know better than them, now ive been very clear from the start im no expert, like most i follow what i read or come across on the internet including Google Russ, and its all saying this is AI and its developing fast, the only place where i see folk saying its not ai is here and if this site was called AI expert i would be much more receptive but its a photography site where people say dont listen to the rest of the planet its not really artificial intelligence its really dumb. well that dumb thing is doing some pretty intelligent stuff we couldn't before now :-) and i recon in future it will probably be doing a lot lot more in everything but remember when that system or robot takes your job its not AI its just a dumb fake system :-D

I have been as fair and conversational and upbeat as I can, and I've tried to explain my arguments, but you're quite rudely misrepresenting what I am saying.

(FWIW I am not an expert in these current systems (I'm just a techie who has read a lot) and I've not claimed to be. I have, though, studied neural networks and AI stuff at degree level. 30 years before ChatGPT launched. I understand people are excited about these new LLM and GAN tools, but they are actually a pretty small offshoot of a very large discipline that has made slow, rigorous, painstaking progress over more than sixty years, and it's kind of disappointing to see the fuss about literal content generators suck all the oxygen out of it.)

Edited by Unfocussed Mike

Unfocussed Mike said

ANDY00 said

Kevin Connery said

Unfocussed Mike said

ANDY00 I am saying that if you want to understand it properly, you should tune out the CEOs and actually look at the research.

 

ANDY00 That's not a 30 second sound-bite from marketing folks, but from the people actually involved in the design and/or training of the various systems. As Unfocussed Mike points out, there is a lot of pop-culture coverage that's simply incorrect. False-to-fact. Wrong. And that's from many mainstream sites and publications in addition to the many people parroting what they think they heard from sources that were oversimplifying things to make it easier to "understand". (See "Lies to Children"--watered down overviews that aren't actually correct but might make something vaguely close to the truth sound kind of understandable by those without the background to understand the more complex reality.)

Whether it's "real AI", machine learning, large language models, generative AI, etc, there seems to be more incorrect statements posing as facts than there is accurate information: simplified or not.

Look, I've been as honest as I can be from a personal point of view. I think these are intelligent systems (not sentient, but intelligent). It's not human or animal; it's artificial. So artificial intelligence is what it is, or the official term is ANI i believe (Artificial narrow intelligence) with the next stage in its evolution being AGI (Artificial general intelligence) Maybe that's not good enough for you guys—tough. That's my own personal opinion. You lot can call it what you like. It's allowing systems to find new antibiotics—fact. It's helping cure diseases—fact. It's deciphering animal language—fact. It can also be used for much more, from telling someone a recipe to controlling a billion-dollar F18, but this is just the beginning. Personally, I don't care what you call it. Call it Bob for all I care. I think it's an amazing step forward and will change life for generations in the future. i list and quote a load of experts in the field and im told well theyre all lying us lot on a photography site know better than them, now ive been very clear from the start im no expert, like most i follow what i read or come across on the internet including Google Russ, and its all saying this is AI and its developing fast, the only place where i see folk saying its not ai is here and if this site was called AI expert i would be much more receptive but its a photography site where people say dont listen to the rest of the planet its not really artificial intelligence its really dumb. well that dumb thing is doing some pretty intelligent stuff we couldn't before now :-) and i recon in future it will probably be doing a lot lot more in everything but remember when that system or robot takes your job its not AI its just a dumb fake system :-D

I have been as fair and conversational and upbeat as I can, and I've tried to explain my arguments, but you're quite rudely misrepresenting what I am saying.

I'm sorry; I honestly didn't realize I had misrepresented you. That's not my intention. I've posted direct quotes from people, but I didn't think I changed anyone's quotes. I'm not an expert on AI or anything for that matter. I think it's a great new tool that is developing, that's all. People have a hate thing going on about it being called AI. Not much I can do about that. It's artificially intelligent, so artificial intelligence it is. Plus, literally the whole world's calling it that, so it's kind of weird to call it that non-artificial intelligence thing.

Edited by ANDY00

Unfocussed Mike to be clear as I'm not sure i have been, you may well know far more than me, you certainly know more about the inner working from what I've seen and I've enjoyed the conversation thanks, maybe we don't agree about everything but you've certainly given me things to think about and i will absolutely read more into it, I am sorry if along the way I have somehow said the wrong thing honestly that's not my intention, it was just a lively debate I thought,  to be honest in this day and age i get a little lost when everyone's saying don't believe them or believe them it rings back to the covid thing with so many points of view and experts, i just tend to follow what i see these days on official websites and stick to official narratives, maybe that's wrong - who knows

 

Edited by ANDY00

ANDY00 said

Unfocussed Mike to be clear as I'm not sure i have been, you may well know far more than me, you certainly know more about the inner working from what I've seem and I've enjoyed the conversation thanks, maybe we don't agree about everything but you've certainly given me things to think about and i will absolutely read more into it, i am sorry if along the way i have somehow said the wrong thing honestly that's not my intention, it was just a lively debate i thought, 

No need to apologise. Indeed, I should instead -- and I do, I am being a bit jumpy and it's not really warranted, so... sorry!

I think Kevin Connery hit the nail on the head, actually. It fits with my main concern here, which does really worry me on a deep level, that people might get cynically laid off from jobs based on the remarkable level of misunderstanding of what these two new AI techniques (LLMs and content-generating GANs) can actually do. 

As someone who is heading towards being an old bloke in a young man's game, I am actually scared about the lack of discipline in the tech industry and the number of people feeding the hype bubble (who are often the very same people who told us excitedly about Theranos, or cryptocurrencies, or NFTs).

Unfocussed Mike said

ANDY00 said

Unfocussed Mike to be clear as I'm not sure i have been, you may well know far more than me, you certainly know more about the inner working from what I've seem and I've enjoyed the conversation thanks, maybe we don't agree about everything but you've certainly given me things to think about and i will absolutely read more into it, i am sorry if along the way i have somehow said the wrong thing honestly that's not my intention, it was just a lively debate i thought, 

No need to apologise. Indeed, I should instead -- and I do, I am being a bit jumpy and it's not really warranted, so... sorry!

I think Kevin Connery hit the nail on the head, actually. It fits with my main concern here, which does really worry me on a deep level, that people might get cynically laid off from jobs based on the remarkable level of misunderstanding of what these two new AI techniques (LLMs and content-generating GANs) can actually do. 

As someone who is heading towards being an old bloke in a young man's game, I am actually scared about the lack of discipline in the tech industry and the number of people feeding the hype bubble (who are often the very same people who told us excitedly about Theranos, or cryptocurrencies, or NFTs).

Thankfully, things like Crypto are beyond me lol, but from what I've read and seen on multiple levels, these new technologies will, and in some cases already are, taking jobs. From the Figure 1 humanoid robot powered by ChatGPT that is due to start working in BMW factories and Amazon warehouses as early as next year, to the AI voice systems due to start answering calls for BT and NHS 111 and your doctors' surgeries, it's going to start filling a lot of jobs if it's all to be believed. That's why they need to start talking about a universal wage system now. but at same time i imaging coders, and tech guys will still be needed for a long time to come as well as engineers and specialists.

Edited by ANDY00

ANDY00 said

Look, I've been as honest as I can be from a personal point of view. I think these are intelligent systems (not sentient, but intelligent). It's not human or animal; it's artificial. So artificial intelligence is what it is, or the official term is ANI i believe (Artificial narrow intelligence) with the next stage in its evolution being AGI (Artificial general intelligence) Maybe that's not good enough for you guys—tough. That's my own personal opinion.

 

I certainly agree with the part in bold.

I've not been keeping as abreast of all the changes since the early days of neural networks (frequently touted as AI), machine learning (ditto) and fuzzy logic (ditto), than when I was working on them as part of my job, but I do know a number of folks in the field, and they're even more irked by the widespread-but-incorrect use of the AI label.

Language is fluid, though, and it's possible we'll see 'intelligence' and 'artificial intelligence' change definition through sheer repetition. (Much like 'organic food'--as if there were any other type--for example.)

ANDY00 said

It's allowing systems to find new antibiotics—fact. It's helping cure diseases—fact. It's deciphering animal language—fact. It can also be used for much more, from telling someone a recipe to controlling a billion-dollar F18, but this is just the beginning. 

Agreed.

I will note that the F18 was released in 1983, and their control systems were nowhere near AI at that time, nor were similarly- and more-advanced control systems through the early 2000's. But that label did get some minimal traction even then. (My stint with control systems and software was in the 1990s through 2002, where AI was slapped onto ML, neural nets, and fuzzy logic; I'm woefully out of date on the deep details, but do still try to keep track of the various offshoots.)

ANDY00 said

Personally, I don't care what you call it. Call it Bob for all I care.

If you're so passionate about something you don't care about, I'm impressed. :) But Bob is a perfectly fine label.

Seriously, though, none of the mass-market tools labeled AI today are what anyone in the field would call AI. That doesn't diminish in the slightest the advantages those current tools offer; it just recognizes that the AI buzz-word escaped the zoo, and is being applied to virtually anything with an algorithm more complex than 1+1=2. (I realize 1+1=2 isn't an algorithm; it's just being used as an analogy here.)

ANDY00 said

People have a hate thing going on about it being called AI. Not much I can do about that. It's artificially intelligent, so artificial intelligence it is. Plus, literally the whole world's calling it that, so it's kind of weird to call it that non-artificial intelligence thing.

Only if you use a unique definition of intelligence. But if you do choose such an unusual usage, sure, calling it AI would be correct. With the current usages and definitions of the word, though, we're not there yet.

Useful? Absolutely; the various tools are showing great promise and usage already. Intelligent, however, is not a term I'd use for any of the current systems. 'Bob' would be a more honest label.

Edited by Kevin Connery

To be honest, people think knowledge is intelligence, in that regard Google was the most intelligent entity some 10 years back.

AI is able to artistically create - music, images, video.

And to be honest this is pretty impressive:

 

ANDY00 said

Kevin Connery said

Unfocussed Mike said

ANDY00 I am saying that if you want to understand it properly, you should tune out the CEOs and actually look at the research.

 

ANDY00 That's not a 30 second sound-bite from marketing folks, but from the people actually involved in the design and/or training of the various systems. As Unfocussed Mike points out, there is a lot of pop-culture coverage that's simply incorrect. False-to-fact. Wrong. And that's from many mainstream sites and publications in addition to the many people parroting what they think they heard from sources that were oversimplifying things to make it easier to "understand". (See "Lies to Children"--watered down overviews that aren't actually correct but might make something vaguely close to the truth sound kind of understandable by those without the background to understand the more complex reality.)

Whether it's "real AI", machine learning, large language models, generative AI, etc, there seems to be more incorrect statements posing as facts than there is accurate information: simplified or not.

Look, I've been as honest as I can be from a personal point of view. I think these are intelligent systems (not sentient, but intelligent). It's not human or animal; it's artificial. So artificial intelligence is what it is, or the official term is ANI i believe (Artificial narrow intelligence) with the next stage in its evolution being AGI (Artificial general intelligence) Maybe that's not good enough for you guys—tough. That's my own personal opinion. You lot can call it what you like. It's allowing systems to find new antibiotics—fact. It's helping cure diseases—fact. It's deciphering animal language—fact. It can also be used for much more, from telling someone a recipe to controlling a billion-dollar F18, but this is just the beginning. Personally, I don't care what you call it. Call it Bob for all I care. I think it's an amazing step forward and will change life for generations in the future. i list and quote a load of experts in the field and im told well theyre all lying us lot on a photography site know better than them, now ive been very clear from the start im no expert, like most i follow what i read or come across on the internet including Google Russ, and its all saying this is AI and its developing fast, the only place where i see folk saying its not ai is here and if this site was called AI expert i would be much more receptive but its a photography site where people say dont listen to the rest of the planet its not really artificial intelligence its really dumb. well that dumb thing is doing some pretty intelligent stuff we couldn't before now :-) and i recon in future it will probably be doing a lot lot more in everything but remember when that system or robot takes your job its not AI its just a dumb fake system :-D

You know, just looking at what it can bring in simpler terms, like ChatGPT making assistants for the elderly or self-driving cars for those who can no longer drive, would make a huge difference in someone's life. That's not a gimmick to those people. The smart systems that can detect if your breathing is off or if your spending has changed in your bank—all AI, all pretty great steps forward

These new intelligent systems can make a massive difference to everyone from the huge corporation to the elderly person sitting alone, they can also be used for bad  purposes just like everything else but that the same with everything,

Edited by ANDY00

While this is a photography / modelling site, as has been pointed out on a number of occasions by several people, a significant number of those here are no Professional photographers, but people for whom photography is a hobby, and who do something else as their 'day job'.

Personally, I work as a software engineer, and have done so for over 30 years, and while I've not worked on AI type systems, I have kept a general interest in them, and the possibilities they offer.

'AI' has the potential to be enormously useful - making 'smarter' machine that can help people in their daily lives, to remove the need for people to work in dangerous situations (EG dealing with hazardous materials, or robot firemen), and to aid in discovery of new medicines, etc.

As for 'automation' taking jobs - yes, it does make (financial) sense for some jobs to be replaced by automated systems - we've had robotic assembly machines for a long time in the car industry for example, and there are plenty of cases of firms using AI for 'promotional' art, etc. (some of which are scams where AI imagery is being used to sell something that doesn't exist, but that's a separate topic).

Some of this is good - dangerous or menial, low pay, work getting replaced is a good thing (in my opinion), using 'AI' to con people, on the other hand, is not.

Kevin Connery said

ANDY00 said

Look, I've been as honest as I can be from a personal point of view. I think these are intelligent systems (not sentient, but intelligent). It's not human or animal; it's artificial. So artificial intelligence is what it is, or the official term is ANI i believe (Artificial narrow intelligence) with the next stage in its evolution being AGI (Artificial general intelligence) Maybe that's not good enough for you guys—tough. That's my own personal opinion.

 

I certainly agree with the part in bold.

I've not been keeping as abreast of all the changes since the early days of neural networks (frequently touted as AI), machine learning (ditto) and fuzzy logic (ditto), than when I was working on them as part of my job, but I do know a number of folks in the field, and they're even more irked by the widespread-but-incorrect use of the AI label.

Language is fluid, though, and it's possible we'll see 'intelligence' and 'artificial intelligence' change definition through sheer repetition. (Much like 'organic food'--as if there were any other type--for example.)

ANDY00 said

It's allowing systems to find new antibiotics—fact. It's helping cure diseases—fact. It's deciphering animal language—fact. It can also be used for much more, from telling someone a recipe to controlling a billion-dollar F18, but this is just the beginning. 

Agreed.

I will note that the F18 was released in 1983, and their control systems were nowhere near AI at that time, nor were similarly- and more-advanced control systems through the early 2000's. But that label did get some minimal traction even then. (My stint with control systems and software was in the 1990s through 2002, where AI was slapped onto ML, neural nets, and fuzzy logic; I'm woefully out of date on the deep details, but do still try to keep track of the various offshoots.)

ANDY00 said

Personally, I don't care what you call it. Call it Bob for all I care.

If you're so passionate about something you don't care about, I'm impressed. :) But Bob is a perfectly fine label.

Seriously, though, none of the mass-market tools labeled AI today are what anyone in the field would call AI. That doesn't diminish in the slightest the advantages those current tools offer; it just recognizes that the AI buzz-word escaped the zoo, and is being applied to virtually anything with an algorithm more complex than 1+1=2. (I realize 1+1=2 isn't an algorithm; it's just being used as an analogy here.)

ANDY00 said

People have a hate thing going on about it being called AI. Not much I can do about that. It's artificially intelligent, so artificial intelligence it is. Plus, literally the whole world's calling it that, so it's kind of weird to call it that non-artificial intelligence thing.

Only if you use a unique definition of intelligence. But if you do choose such an unusual usage, sure, calling it AI would be correct. With the current usages and definitions of the word, though, we're not there yet.

Useful? Absolutely; the various tools are showing great promise and usage already. Intelligent, however, is not a term I'd use for any of the current systems. 'Bob' would be a more honest label.

Edited by Kevin Connery

it was f16 i apologise but still 100% autonomous flight control, no less impressive or scary and now according to many reports cannot be beaten by a human pilot over 100s of test flights, obviously humans are effected by G-Forces - AI is not, Humans have fear no matter how brave they are - AI does not. humans get fatigued - AI is like the Duracell bunny, flying a fighter jet requires intense thought processes and skills in mathematics and physics, a humans just cannot compete with an AI at this kind of fast paced thought also this is what we publicly see, if we can see this then you know the militaries of the world have moved way way beyond this by now, We only get to see it when its considered obsolete tech by the powers that be but if this is what they consider old and it cannot be beaten by human actors its scary to think what they have now.

Sidenote, notice news actors and military sources all stating - AI, all videos and quotes I've added have all publicly called this tech AI 

 

 

Edited by ANDY00

I would be interested to know how the new ChatGPT-4o performs in the Turing Test. It's trained on human emotion and facial expressions and can see, hear, and speak. It's uncannily real when conversing with it. Would it pass the Turing Test? How long before it's impossible to distinguish it from humans, aside from the fact that it possesses more knowledge than us as individuals? You have to admit, maybe it's a fancy algorithm or computer program, but it's incredibly lifelike and intelligent already, and it's essentially just a baby. What can we expect in a year, two years, or even a decade?

I'm sure this software will be flooding the call centre market already, so we will probably be talking to these things on a regular basis very soon as it removes all call centre jobs from the marketplace. And with Google just releasing its newest one also, there's a real race going on right now to be the leader in the marketplace, not just the consumer market but the job market as well.

I think you're absolutely right ANDY00 that we will soon be (or we already are) speaking with AI on some call centres.  And there will many, many workers who will soon find themselves replaced with the technology.

As someone in the tail end of my working life, I'm somewhat cynical about workers demanding that companies refrain from modernising with the aim of reducing the workforce.

Sure, it is galling to see the job you have proudly performed for decades be replaced by a machine or someone fresh out of school or a piece of software. Galling, shocking and potentially very worrying from a "how am I going to put food on the table and pay my mortgage"-perspective. But at the end of the day, companies have a responsibility to their shareholders, and if they do not move "with the time", they often find themselves out of business which certainly doesn't help workers.

This is not to say that companies should just ignore the human perspective. I think thriving businesses thrive because they master the art of looking after shareholders as well as their "human resources".

YorVikIng said

I think you're absolutely right ANDY00 that we will soon be (or we already are) speaking with AI on some call centres.  And there will many, many workers who will soon find themselves replaced with the technology.

As someone in the tail end of my working life, I'm somewhat cynical about workers demanding that companies refrain from modernising with the aim of reducing the workforce.

Sure, it is galling to see the job you have proudly performed for decades be replaced by a machine or someone fresh out of school or a piece of software. Galling, shocking and potentially very worrying from a "how am I going to put food on the table and pay my mortgage"-perspective. But at the end of the day, companies have a responsibility to their shareholders, and if they do not move "with the time", they often find themselves out of business which certainly doesn't help workers.

This is not to say that companies should just ignore the human perspective. I think thriving businesses thrive because they master the art of looking after shareholders as well as their "human resources".

This is why countries are now talking about the need for a universal basic income. AI and robotics both threaten to, and are moving into, the workforce at great pace now, so more and more job fields will be taken over by AI. In some ways, that will be great and improve lives, but like with anything, there will be downsides until the technology finds an even foothold—or rather, interestingly enough, maybe until humans find a new foothold for themselves in a new world where work is an option in the future.

Just no toaster's please lol

If you follow the thoughts of ai+robots taking over all human work, then where does it end?

Is it a utopia where humans get fat from doing nothing and soon become incapable of everything?

Or is it an end to the megacorps, who need money to keep going, but if no one is earning money, then who pays? And if no one is earning money, who keeps the governments in the manner they have become accustomed to?

My bet is just more government focus on making sick consumers and cannon fodder, both of which keep the economies of healthcare and war going, both of which prop up sick societies.