ChatGPT incoming

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Chuck Taper on 18/04/2023 18:21:47:

Just to labour a point.

I assume that photography must have caused some concern when it arrived – any idiot could now independently produce clear, detailed images without the need for skilled ….etc, etc, etc.

[…]

.

No need to assume, Frank … it is very well documented

Here is a verbatim quote from the article that I referenced:

”One hundred and seventy years ago, Baudelaire pronounced the nascent medium of photography to be the “mortal enemy of painting, (&hellip refuge of all failed painters, the untalented and the lazy”. Today, AI artists are berated in much the same way by photographers. Back then, photographers took the likeness, the realistic representation, away from painters. It proved a liberating blow, the prerequisite for the development of modern painting.”

MichaelG.

.

Edit: __ verbatim except for the stupid #####ng smiley thing introduced by this [soon to be forgotten?] forum software

Edited By Michael Gilligan on 18/04/2023 20:31:27

[Chuck TaperParticipant @chucktaper]

An interesting TED talk – https://www.youtube.com/watch?v=C_78DM8fG6E

Sort of the end of computings steam age [If you will pardon the mixed metaphor]

Regards.

FrankC.

[Chuck TaperParticipant @chucktaper]

Messed up the link. Missed the edit window.

Oh well.

[Bill PhinnParticipant @billphinn90025]

Posted by Bill Phinn on 10/04/2023 14:22:24:

Posted by SillyOldDuffer on 10/04/2023 11:14:14:It, and other software, are now capable of answering questions like Bill Phinn's "In Flaubert's Madame Bovary Charles is largely to blame for Emma's death". Discuss. Refer closely to the text in your answer." well enough to fool a college lecturer!

I think you're right, Dave, though maybe that says more about the capabilities of some college lecturers than the capabilities of ChatGPT.

I would genuinely be interested in seeing an attempt by ChatGPT at this essay question. I think the answer could be superficially plausible, but, along with people like these, suspect that on closer analysis it would not pass muster.

To update this part of the discussion, I've just put the Flaubert question to Bing's AI chat bot. I expected largely evasive waffle and a re-telling of the plot, but instead got a very impressive 300-word answer.

I think this spells the end of homework and coursework assessments in virtually every subject. From this point on, surely only exams [from which all digital devices are banned] will result in fair grades.

[Ady1Participant @ady1]

I'm going to wait until these things are openly available and not being used as a personal data gathering vehicle

[Nigel Graham 2Participant @nigelgraham2]

I've a sinking feeling that open availability and personal data gathering go hand in hand.

[S KParticipant @sk20060]

These systems are impressive and are growing better exponentially. They will change the world, for better and for worse. Prepare to be spammed with auto-generated fake news and believable personalized calls from your "friends" who are in trouble and need you to send money fast. (And, as always, war and porn will be amongst the first and biggest applications.)

I haven't tried one myself yet, partially because of the privacy issue. But also because I don't want to use it for shopping and other stuff Microsoft, et. al., will want you to use it for (it's all about money in the end, isn't it?).

Instead, I'd like to use it to explore the zeitgeist of human knowledge and culture. But one "problem" with the current crop of publicly-accessible generative AI's is that they are deliberately hamstrung. For example, they will refuse to converse on a given topic for more than say a dozen prompts to stop them from meandering. And, of course, they are heavily constrained because of the fear that they will go off into unfortunate directions that may hurt people's sensibilities.

Now we see the likes of Google actually promoting government restrictions on AI. I wonder if that's really just "rent seeking" – the desire to remain the nearly-sole controller and provider of access? Probably, but it's probably also doomed to failure.

There are efforts to develop public domain LLM's (and photo generation, etc.) that will run locally on a reasonably powerful personal computer. For example, Meta's model was made open source, and their "weights" (the trained data) leaked. These are interesting to me, and one day I'll see if I can install one.

The scariest question of all is when they will achieve superior intelligence to us (whatever "intelligence" means), and what they will grow capable of. Don't go thinking that humans will remain on top. At the rate they are improving, it could be just a year or two. And sooner rather than later, they will be handed abilities to interact with the world autonomously. It's inevitable.

Pandora's box has been opened.

Edited By S K on 25/05/2023 16:05:10

[Ady1Participant @ady1]

We already have mass cheating in popular videogames, forum chat spambots and computer viruses have been about for years

Now we'll have AI developing and running DDoS exercises and military AI hacker computers having a go at each others networks

In 20 years the global internet may be toast and split along global political boundaries

(but As long as the ME site is still up and my Alibre key still works I'll be fine)

Edited By Ady1 on 25/05/2023 16:58:38

[Neil WyattModerator @neilwyatt]

I've found ChatGPT very useful for researching topics. It saves consulting multiple sources when looking at a subject where there is no obvious authority. But you do need to do some poking around of your own to fact check.

Neil

[John HaineParticipant @johnhaine32865]

Posted by Neil Wyatt on 25/05/2023 17:32:56:

I've found ChatGPT very useful for researching topics. ….

Neil

Having done some horology searches using it I have no faith whatever in its answers. The whole point about doing research on something you don't know about is that you need reliable information. If you don't know in the first place you can't judge. If you then have to go off and find other sources you might as well not bother.

What worries me is that people will just use it to search information and rely on the answers – for example there is documented evidence of someone having a criminal conviction ascribed to them which wasn't in fact the case.

[S KParticipant @sk20060]

Automatic fact-checking is a top priority for all LLM's, and is in place in most already. Or you can say "fact check that for me, please" and it should provide links, etc.

[SillyOldDufferModerator @sillyoldduffer]

Posted by John Haine on 25/05/2023 17:50:17:

Posted by Neil Wyatt on 25/05/2023 17:32:56:

I've found ChatGPT very useful for researching topics. ….

Neil

Having done some horology searches using it I have no faith whatever in its answers. The whole point about doing research on something you don't know about is that you need reliable information. …

Humans are worse though. So far no AI has been deliberately economical with the truth or told a lie. Unlike everybody who is reading this! We are all sinners.

AI are less likely to misread stuff too. For some strange reason I thought John was into horoscopy!

Dave

[S KParticipant @sk20060]

I have friends who are professors, and they are in two camps. One camp is ready to give up on any hope that their students will actually learn anything from here on out. The other is "this is just another tool."

When I was young, it was "why should I use tables of logarithms when I have a calculator now?"

Next it was "why should I memorize those formulas when I can just google it?"

Now it's "why should I bother working hard on an assignment when, in seconds, a machine can produce a much better-written essay, etc., than I ever could if my life depended on it!?" (Sadly, in most cases, they are right!)

Little story about the first one: When I was young, we were taught to use tables of logarithms, but we were allowed to use new-fangled scientific calculators if we wanted – we just had to get the same answers as given by the tables. Alas, I was too poor to afford an actually-accurate HP calculator, but I didn't recognize the danger, so I blithely used my junk-quality one during an important test. BAD IDEA!

[Ady1Participant @ady1]

A bit more info on the AI "technology" today

It looks like "AI" is simply a case of running standard data sets really fast on amazingly high end chips so there is in fact no real jump in anything at all except the data computation rate which any computer programmer knows has been evolving over the decades anyway

So a chess computation that took 12 months 20 years ago now takes 10 seconds

The technology has got faster and can do much bigger sums, it has not got smarter

It's exactly the same technology that kicked off the Bitcoin era, GPU parallel processing

The real evolution comes through the coding system,  if they can program these things with a simple but powerful language like BASIC which anyone can comprehend then an awful lot of solutions to knotty problems will suddenly start appearing

Edited By Ady1 on 25/05/2023 20:21:10

[Ady1Participant @ady1]

This level of computation means that you can put together a program which allows the computer to produce and run its own programs

This makes the birth of the AI bogyman computer possible, but it won't know its the bogyman to start with

[S KParticipant @sk20060]

Posted by Ady1 on 25/05/2023 20:12:38:

It looks like "AI" is simply a case of running standard data sets really fast on amazingly high end chips so there is in fact no real jump in anything at all except the data computation rate which any computer programmer knows has been evolving over the decades anyway

With due respect, this is not correct anymore. That used to be the case, back when Deep Blue beat Kasparov in chess. That was accomplished "simply" by trying millions of moves vastly faster than a human could. Faster chips certainly help that strategy: Now instead of needing a supercomputer, a phone can beat the strongest player on earth.

It's the strategies that are qualitatively different now. It's essentially about creating a mind; a vast network of neural connections. Training that mind, e.g. by having it play chess against itself, or by reading millions of documents, results in the reinforcement or weakening of those millions of connections until it's imbued with knowledge or understanding. Once that training is complete, you can present the AI with a chess-board, and it immediately "sees" the top moves with very little effort involved, just like a top chess master can immediately see the top moves too.

So it's almost backwards from what it once was:

Old AI: Extremely little knowledge, e.g. just the legal moves, but tons and tons of computations stumbling about almost at random trying to find a move that works.

New AI: Tons and tons of learning (e.g. playing games against itself, or reading the whole internet), followed by the near-instantaneous evaluation of solutions based on what it learned.

Yes, faster chips still helps. Learning language skills to the level of GPT4 still takes enormous amounts of effort. It takes people a decade or more to learn to write, too. But after that, the knowledge is known by the computer, and can be used almost for free.

(Nvidia just got lucky to have the right hardware at the right time, and they have many others breathing very heavily down their necks.)

Edited By S K on 25/05/2023 22:20:36

[Chris MateParticipant @chrismate31303]

Its based from data, theres still a human mind behind it, it can be steered, so certain humans will be very happy with it, as data change the results will change, it cannot think like a slow human brain, it however can compute, some human brains confuse intelligence with computing speed, memory capability, data gathering, it still does not have the flexability in the hardware like a human brain(Flesh & blood microbes etc), so we are back to Square-One where one human man has find a way to control others till they eventually figured out how its happenning. In between reality is going on.

Most humans just believe everything too easily.

Edited By Chris Mate on 26/05/2023 05:00:19

[HopperParticipant @hopper]

I am seeing increasing numbers of posts on internet forums, Twitter and other social media that are obviously generated by ChatGPT or similar AI. They are just too well researched for someone who is not a professional in the field to have the knowledge on hand, and yet they contain some obvious clangers if you have the real-world expertise in that field to spot them.

It think it is possible that in a relatively short period of time, internet social media will become such a melee of these warring AI-generated rubbish posts it will all be rendered a meaningless echo chamber, the human content lost and maybe we will all go back to talking in real time to real people at the local pub, club, community group, football etc. So, it might actually do us some good.

[Ches Green UKParticipant @chesgreenuk]

….internet social media will become such a melee of these warring AI-generated rubbish posts it will all be rendered a meaningless echo chamber….

I think you are correct. At the moment there are dumb bots directing the conversation on a lot of media platforms, and they can usually be spotted. But once AI bots enter the playing field then we may as well give up.

If there was a way for the platform to weed out the AI bots then fine, but I can't see how that is possible. They will always be one step ahead of any policing.

Ches

[HopperParticipant @hopper]

I don't think it is necessarily AI bots posting. More like humans wanting to appear more knowledgeable than they really are but not wanting to do even the usual armchair "research" via Google. Just plug the question into ChatGPT then cut and paste the answer into your Tweet, forum post etc, add a few insulting comments on the end, and bask in the accolades of being the resident expert on the topic in hand.

I don't think AI has yet developed an artificial ego to rival the human version.

[Ches Green UKParticipant @chesgreenuk]

Hopper,

Yes, I was off on a slight tangent…there are humans no doubt using AI generated info to punch above their weight on media platforms.

I think the longer term issue will be when AI bots (rather than humans) use media platforms to exercise 'thought control' over those that use media.

Ches

[Ady1Participant @ady1]

That beautiful fluffy baby computer from 1980 turned into a 2030 Gremlin apocalypse

[HopperParticipant @hopper]

Apparently the Gremlin apocalypse is already upon the hallowed halls of academia. Just saw this article about an academic paper about how AI makes student plagiarism harder to detect — written by a chatbot well enough to fool the academic experts (which may or may not be hard to do!) LINK

Writing a forum post is child's play by comparison. A certain American motorcycle forum I sometimes look at has an uncensored political and current affairs discussion section. Predictably it is a cesspool of misinformation, ideological biases and personal abuse. There are also many seemingly erudite, well researched and well argued lengthy posts that are so out of character with the poster's previous not-very-smart posts indicative of a not-very-high level of education that one has to wonder. Google searching seems to indicate they were not cut and pasted from existing websites that Google trawls so I seriously wonder if they are using ChatGPT etc to formulate their arguments and "one-up" their forum opponents.

Earlier in this thread, Neil posted an AI response to the perennial question: Which is a better lathe, Myford or Sieg? The response was a well-balanced reasoned argument. But what if you asked ChatGTP a biased/leading question such as "What are the arguments in favour of a Myford lathe being better than a Chinese lathe?" Or "What are the shortcomings of Chinese lathes compared with Myfords?" Would be a different result I reckon.

I don't seem to be able to log into or create a new account at ChatGPT for reasons beyond my technical skill level (maybe I need to pay for an account?) so can't test this out myself.

But I get the strong feeling this type of AI is going to change the social media world much faster than the real world in the next year or two. (Except perhaps for academia!)

Edited By Hopper on 26/05/2023 10:53:57

[BazyleParticipant @bazyle]

Posted by Hopper on 26/05/2023 10:51:41:

I don't seem to be able to log into or create a new account at ChatGPT for reasons beyond my technical skill level (maybe I need to pay for an account?) so can't test this out myself.

Please can someone ask Chatgpt how to get Hopper a free account.

[HopperParticipant @hopper]

Posted by Bazyle on 26/05/2023 11:09:18:

Posted by Hopper on 26/05/2023 10:51:41:

I don't seem to be able to log into or create a new account at ChatGPT for reasons beyond my technical skill level (maybe I need to pay for an account?) so can't test this out myself.

Please can someone ask Chatgpt how to get Hopper a free account.

[Author]

Posts