Please or to access all these features

AIBU?

Share your dilemmas and get honest opinions from other Mumsnetters.

Why are some people anti Ai

37 replies

NedLuddlite · 21/01/2024 01:33

The comparison between modern-day concerns about artificial intelligence (AI) and historical Luddites provides a fascinating lens through which to examine societal reactions to technological advancements. The Luddites were a group of English textile workers in the early 19th century who protested against the industrialization of their craft, fearing that new machinery would replace their jobs and disrupt their way of life. Drawing parallels between their concerns and contemporary apprehensions about AI reveals both similarities and differences.

Similarities:

Job Displacement:

Luddites: The primary concern of the Luddites was the displacement of skilled workers by automated machinery. They feared that the introduction of power looms and other mechanical devices would render their craftsmanship obsolete, leading to unemployment.

Modern-day AI: There are parallel concerns today regarding the impact of AI on employment. Automation and AI technologies have the potential to replace certain jobs, especially those that involve repetitive tasks. The fear is that as AI systems become more sophisticated, they might encroach upon a wide range of professions, leading to job displacement.

Economic Inequality:

Luddites: The Luddites were also motivated by concerns about economic inequality. They believed that the benefits of industrialization would be concentrated in the hands of factory owners, leading to the impoverishment of the working class.

Modern-day AI: Similar concerns exist in the context of AI. Questions about who controls and benefits from AI technologies, as well as the potential for increased economic inequality, are prominent in discussions. Issues such as the concentration of AI development in certain corporations or countries and the potential for job losses among the less-skilled workforce contribute to these concerns.

Resistance to Change:

Luddites: The Luddites resisted technological change, viewing it as a threat to their established way of life and livelihoods.
Modern-day AI: There is a parallel resistance to the widespread adoption of AI technologies. This resistance can be rooted in concerns about job displacement, ethical implications, and the broader societal impacts of AI. Some argue that the rapid pace of AI development may outstrip society's ability to adapt responsibly.
Differences:

Nature of Technology:

Luddites: The Luddites were primarily concerned with the introduction of specific machinery in the textile industry.

Modern-day AI: The concerns about AI are more diverse, encompassing a broad range of applications from autonomous vehicles and natural language processing to machine learning algorithms used in finance, healthcare, and more. The scale and scope of AI's potential impact on various aspects of society set it apart from the specific technological advancements faced by the Luddites.
Global Impact:

Luddites: The Luddite movement was localized to specific regions in England and was largely a response to the changing dynamics of the textile industry in that era.

Modern-day AI: Concerns about AI have a global dimension. AI technologies are being developed and implemented worldwide, raising questions about international collaboration, regulation, and ethical standards on a global scale. The implications of AI extend far beyond the borders of any single country.

Acknowledgment of Benefits:

Luddites: The Luddites were skeptical of the benefits of industrialization, seeing it primarily as a threat.

Modern-day AI: While concerns exist, there is a broader recognition of the potential benefits of AI, including increased efficiency, advancements in healthcare, and solutions to complex problems. The discourse around AI often involves a nuanced evaluation of both its risks and rewards.

In conclusion, the comparison between modern-day concerns about AI and historical Luddites reveals shared anxieties about job displacement, economic inequality, and resistance to technological change. However, the differences in the nature and global impact of the technologies highlight the complexity of the AI discourse. It is crucial for society to engage in thoughtful and informed discussions, considering the ethical, economic, and social implications of AI while also recognizing its potential contributions to human progress.

OP posts:
New posts on this thread. Refresh page
MobilityCat · 18/03/2024 06:02

Vegetus · 11/03/2024 17:27

It's just a language bot it doesn't really think for itself. It doesn't have a 'mind' of it's own like the coming true AGI will, it can only tell you information it's been trained on in chatGPT's case that's the internet.

Don't get me wrong it's very clever and getting better with each iteration but it isn't true AGI for the reason stated above. The founders and engineers who work on it will tell you it's not AGI and it won't be until it thinks for itself without human input or training but in the future I'm sure it will be.

I hope that your real name isn't Sarah Connor. It may be bad news for all of us but it will be particularly bad for you.

Vegetus · 11/03/2024 17:27

BassoContinuo · 11/03/2024 16:13

ChatGPT seems to think it is AI - I just asked it and got the response

Yes, I am an AI, specifically a language model developed by OpenAI called ChatGPT.

It's just a language bot it doesn't really think for itself. It doesn't have a 'mind' of it's own like the coming true AGI will, it can only tell you information it's been trained on in chatGPT's case that's the internet.

Don't get me wrong it's very clever and getting better with each iteration but it isn't true AGI for the reason stated above. The founders and engineers who work on it will tell you it's not AGI and it won't be until it thinks for itself without human input or training but in the future I'm sure it will be.

BassoContinuo · 11/03/2024 16:13

Vegetus · 11/03/2024 13:43

It's not AI. You don't know what you're talking about if you state it is.

ChatGPT seems to think it is AI - I just asked it and got the response

Yes, I am an AI, specifically a language model developed by OpenAI called ChatGPT.

Vegetus · 11/03/2024 13:43

MobilityCat · 11/03/2024 12:06

I tried Chatgpd and I can assure you that it is AI. "Compose an article on..." Write a poem about.. " Compile a training program for..." Play Dungeons and Dragons with me."

It's not AI. You don't know what you're talking about if you state it is.

TheGreatGherkin · 11/03/2024 12:26

SkyNet 😉

ancienticecream · 11/03/2024 12:22

I once asked ChatGPT about a nearby town. It started banging on about how well connected it was, boasting not one, but two train stations! However, the train stations were subject of the Beeching cuts, and closed in the early 1960s.

I think AI is dangerous for us a society. From the introduction of 'face-smoothing' via Samsung phones in 2017 (goodbye to many people's self-esteem) to now not being able to tell the difference between what's a real video and what's a fake video, I dread to think what is possible by the time my kids are teenagers. I work in tech, but I do not like AI at all.

GasPanic · 11/03/2024 12:16

AI is going to be a wealth concentrator, much like a lot of technology that has been developed recently.

The many get replaced by AI in order to enrich the few. It would be great if AI was used to enrich society generally, but I am not sure historically that is how it has panned out.

So yes, people who will be replaced by AI have significant amounts to fear.

MobilityCat · 11/03/2024 12:06

Vegetus · 21/01/2024 04:38

Chats bots are not really AI, They are more like a fancy search engine.

I tried Chatgpd and I can assure you that it is AI. "Compose an article on..." Write a poem about.. " Compile a training program for..." Play Dungeons and Dragons with me."

EBearhug · 21/01/2024 18:30

True @Jovacknockowitch it is interesting that they were presented (and used in popular language) as bad. Maybe not surprising as unsuccessful labour movements for the people aren't well liked by the establishment.

It was (1812) also the time of the Napoleon's Wars, and ruling classes everywhere were unsettled by the French Revolution (started 1789.)

We're in a very unsettled world at the moment, too. AI is a tool and should mostly only be used to do what humans tell it to - a lot of that use currently is military.

BassoContinuo · 21/01/2024 17:40

I’m anti-AI on MN. We don’t need even more OPs that are lengthy and difficult to read.

NomenNudum · 21/01/2024 17:30

very true. I asked ChatGPT to suggest some books set where I live in Europe. It came up with four titles - one set in America, one in Russia, one in space, and one completely made up title.

ThatWriterInTheCorner · 21/01/2024 16:49

Here's why I am worried about AI. I'm a copywriter, and one of the clients I work for uses an AI model to generate basic product descriptions for their website, which are then edited by a team of humans (including me).

The AI model is excellent at spelling, punctuation and grammar. It is pretty good at tone of voice. And it is utterly dreadful at accuracy. It regularly (as in between 80% and 100% of the time) produces complex, convincing hallucinations, describing features of the product that simply aren't there. These aren't based on any inputs the AI has been supplied with. It's simply freestyling based on what it's been trained on, which is previous website descriptions.

This doesn't matter much, because it's just descriptions of products. A human is checking them, and even if we fuck up, the human buying the product can check against the image on the website and see what's true.

However, AI is increasingly writing content for news sites, where hallucinations could be much more scary. If a news story told you Joe Biden had grown arms like a lobster, or that Donald Trump had exploded, you probably wouldn't believe it. But if it said Joe Biden had fallen asleep in a meeting with Putin, or Donald Trump hsd been diagnosed with dementia, you might.

Also, AIs are often used to pull content from other sites and repackage it. They can do this in seconds. Not hours. Seconds. So the hallucinated Sleeping Biden or Demented Trump story could suddenly be on multiple sites, gaining more apparent credibility with every appearance.

It still wouldn't be true. It would just be a hallucination from some AI somewhere. But how would you know? And if it tallied with your existing political position, how much effort would you put into checking? How long before you decided all the news was potentially made up, and stopped listening altgorthet?

That's why I'm scared by AI.

MrsTerryPratchett · 21/01/2024 16:29

True @Jovacknockowitch it is interesting that they were presented (and used in popular language) as bad. Maybe not surprising as unsuccessful labour movements for the people aren't well liked by the establishment.

Machines, then computers, now AI are supposed to give us a utopia of ease and happiness. But they haven't. Because of resource hoarding.

Jovacknockowitch · 21/01/2024 16:17

The Terminator when it comes is going to be a whiny, racist incel.

So true.

Jovacknockowitch · 21/01/2024 16:16

Sick of the Luddites being misrepresented - they were rioting because they faced fucking starvation.

"But whilst these outrages must be admitted to exist to an alarming extent, it cannot be denied that they have arisen from circumstances of the most unparalelled distress. The perseverance of these miserable men in their proceedings, tends to prove that nothing but absolute want could have driven a large and once honest and industrious body of the people into the commission of excesses so hazardous to themselves, their families, and the community. At the time to which I allude, the town and county were burdened with large detachments of the military; the police was in motion, the magistrates assembled, yet all these movements, civil and military had led to—nothing. Not a single instance had occurred of the apprehension of any real delinquent actually taken in the fact, against whom there existed legal evidence sufficient for conviction. But the police, however useless, were by no means idle: several notorious delinquents had been detected; men liable to conviction, on the clearest evidence, of the capital crime of poverty; men, who had been nefariously guilty of lawfully begetting several children, whom, thanks to the times!—they were unable to maintain."

MrsTerryPratchett · 21/01/2024 16:05

brrrrrrrrrrrrrrrrrrrrrrrrrrrrr · 21/01/2024 13:08

The most sinister things about AI are IMHO:

a) Making it impossible to know what is real and true.

b) Destroying the utility of the internet by flooding it with low quality articles / information which other AIs re-digest as input and eventually use as a basis to post yet more guff, meaning the signal to noise ratio crosses the threshold and makes the internet unusable.

I think the most sinister thing is that it 'learns' from the internet. Which is a cesspit of racism, sexism, misogyny and misinformation. Considering the make-up of the tech industry, just look at the gaming scandals, it's going to be the truest case of 'garbage in-garbage out' in history.

The Terminator when it comes is going to be a whiny, racist incel.

NomenNudum · 21/01/2024 13:39

I have a number of issues with AI.

  • Philosophically, art and creativity is what makes us human, so why TF would you outsource that.
  • When you test AI on something you know a lot about, it clearly talks bollocks, so you can safely presume it's also talking bollocks about stuff you know less about.
  • It will create a skills shortage as people will no longer see the value in training in certain areas, and by the time we realise AI is not up to the job those valuable skills will be lost
  • and perhaps most importantly, it's an environmental disaster: the LLM data sets are so large they use up vast quantities of energy and water. I don't think humanity can afford to waste those resources right now.
QueenMegan · 21/01/2024 13:14

Define "the people". You can't

brrrrrrrrrrrrrrrrrrrrrrrrrrrrr · 21/01/2024 13:08

The most sinister things about AI are IMHO:

a) Making it impossible to know what is real and true.

b) Destroying the utility of the internet by flooding it with low quality articles / information which other AIs re-digest as input and eventually use as a basis to post yet more guff, meaning the signal to noise ratio crosses the threshold and makes the internet unusable.

Weefreetiffany · 21/01/2024 13:03

SunflowerSeeds123 · 21/01/2024 05:49

My DD wants to go into a creative industry either photography/media or some kind of creative writing role. She uses PS for some of her images but generally tries to keep manipulation to a minimum. She can spend HOURS on research, travel, lighting, processing and presentation of a piece. It is an endeavour and I fully support her learning her craft.

AI is learning VERY fast and although if you ask ChatGPT for an image it might still make mistakes it's not far from being accurate. There's already been huge debate for years over how images are produced (one that was heavily improved by PS won a national prize for nature photography) and where the line is. DD will have AI challenging her to prove her work is untouched and genuine. It takes DD up to a week to produce one good image, AI can do it instantly. It's like buying something mass-produced vs buying something handmade.

It's interesting that the science of photography challenged how we produced images in the Victorian age, which was by hand. This is another evolution. I hope in the end the two ways can coexist, but DD is hyper-aware that there may be big challenges to come in her future.

This will be an important skill. I’ve read that some criminals are using 6 fingered gloves so they can say cctv of their crimes is doctored by AI, photographers who can prove the veracity of images through knowledge of digital construction will be in high demand. Good luck to your daughter, sounds really interesting and worthwhile

EBearhug · 21/01/2024 08:16

get the tech bros fired up

Yes, they're one of the main issues with AI.

Images are interesting- I went to a talk by Marcus du Sautoyand he was talking about how AI-generated images of gym weights - the sort you hold in your hand for bicep curls or whatever - often had a disembodied arm included, because most of the images it learnt from had someone using the weights, so it thought an arm was an integral part of it, and then built in that error.

SunflowerSeeds123 · 21/01/2024 05:49

My DD wants to go into a creative industry either photography/media or some kind of creative writing role. She uses PS for some of her images but generally tries to keep manipulation to a minimum. She can spend HOURS on research, travel, lighting, processing and presentation of a piece. It is an endeavour and I fully support her learning her craft.

AI is learning VERY fast and although if you ask ChatGPT for an image it might still make mistakes it's not far from being accurate. There's already been huge debate for years over how images are produced (one that was heavily improved by PS won a national prize for nature photography) and where the line is. DD will have AI challenging her to prove her work is untouched and genuine. It takes DD up to a week to produce one good image, AI can do it instantly. It's like buying something mass-produced vs buying something handmade.

It's interesting that the science of photography challenged how we produced images in the Victorian age, which was by hand. This is another evolution. I hope in the end the two ways can coexist, but DD is hyper-aware that there may be big challenges to come in her future.

Weefreetiffany · 21/01/2024 05:41

Definitely an essay. Pre gcse? Better off posting on Reddit and saying something goady to get the tech bros fired up and firing off opinions you can pinch and sort and pass off as your own thinking… a kind of aggregation, a bit like how an ai chat bot works? Be the interface op! You can do it!

Isthisexpected · 21/01/2024 05:28

steff13 · 21/01/2024 02:17

Is this a paper you've written? Did you want suggestions on editing or...?

Yeah we're just being used for a critique before hand in I think.