The Artificial Intelligence Thread

[Ifreann]

It's been a while since I solved a quadratic equation, but this solution is wrong.

nope
7^2 + 6 is in fact 55

Yes, but -3^2 + 6 does not equal 55. The solution is not x = -3 or x = 7, the solution is just x = +-7. Further, the quadratic formula was applied incorrectly. The equation given isn't in the format ax^2 + bx + c = 0, and yet ChatGPT plugs in 6 as c. That's wrong, c should be -49. Based on that wrong use of the quadratic formula it concludes that there is no real solution and therefore that the solution is -3i and 7i. Which is wrong. Solving this equation doesn't even require the use of the quadratic formula. Subtract 6 from both sides, giving x^2 = 49, get the square root for x = +-7, simples.

[The United Penguin Commonwealth]

it annoys me that so much of machine learning has been ignored in favor of trying to use general language models for everything. half the time you’d be better off with a purpose-built AI.

[Kerwa]

Yeah this strikes me as a classic negative psychology grift. What better way to make your product look powerful than warn of its dangers?

All the bad stuff has happened already anyway.

[The Great Nevada Overlord]

Didn't think I was gonna start a conversation on quadratic equations lol.

[Ifreann]

Didn't think I was gonna start a conversation on quadratic equations lol.

Your claim that AI could let students pass classes without issue does seem to have fallen apart upon examination.

[The Great Nevada Overlord]

Didn't think I was gonna start a conversation on quadratic equations lol.

Your claim that AI could let students pass classes without issue does seem to have fallen apart upon examination.

Well I used character.ai (in hindsight should have specified that) because I'm an e-mailless caveman.

But still, this stuff in the future could assist in homework.

[Eahland]

Your claim that AI could let students pass classes without issue does seem to have fallen apart upon examination.

Well I used character.ai (in hindsight should have specified that) because I'm an e-mailless caveman.

But still, this stuff in the future could assist in homework.

It really can't. It's basically a bullshit generator. It generates authoritative-sounding text about things that it has no actual understanding of — because it has no actual understanding of anything — which means that you end up with responses that are spectacularly wrong, but written convincingly enough that if you also lack understanding of the subject, it's not obvious that it's spectacularly wrong. And it will actively defend its spectacularly wrong conclusions. Everything it produces needs fact-checking, and you need more understanding of the subject to fact-check its actively-misleading authoritative-sounding bullshit than you do to just write a response of your own.

[Bombadil]

All this AI is useless stuff reminds me how people scoffed at the Internet back in the day, and laughed at the idea anyone would buy anything off it, so slow.. so hard to work..

AI will change things dramatically, it will improve and the danger is that it improves itself so we do need to be careful about the parameters we allow.

[Emotional Support Crocodile]

In a White Paper due out today, the UK Government have apparently chosen not to have a new regulator with responsibility for AI issues, but will stick with existing ones like the Health and Safety Executive, who will have to carry out the new duties using existing laws.

https://www.gov.uk/government/publicati ... n-approach

https://www.bbc.co.uk/news/technology-65102210

Science, innovation and technology secretary Michelle Donelan said: "Artificial intelligence is no longer the stuff of science fiction, and the pace of AI development is staggering, so we need to have rules to make sure it is developed safely."

But Simon Elliott, partner at law firm Dentons told the BBC the government's approach was a "light-touch" that makes the UK "an outlier" against the global trends around AI regulation.

China, for example, has taken the lead in moving AI regulations past the proposal stage with rules that mandate companies notify users when an AI algorithm is playing a role.

"Numerous countries globally are developing or passing specific laws to address perceived AI risks - including algorithmic rules passed in China or the USA," continued Mr Elliott.

He warned about the concerns that consumer groups and privacy activists will have over the risks to society "without detailed, unified regulation."

He is also worried that the UK's regulators could be burdened with "an increasingly large and diverse" range of complaints, when "rapidly developing and challenging" AI is added to their workloads.

Sounds like a dumb approach to me.

This follows the EU's Artificial Intelligence Act
https://eur-lex.europa.eu/resource.html ... format=PDF

and the US Algorithmic Accountability Act
https://www.congress.gov/bill/117th-con ... /6580/text

[The Archregimancy]

Thank you! Like Goddamn, did no one see Terminator? One computer system gets too smart and boom, we're history. About time these morons in tech realized the danger.

Terminator is likely not the pop culture reference they're most worried about.

[Emotional Support Crocodile]

Damn young 'uns have clearly never heard of Colossus: The Forbin Project

[The Archregimancy]

Damn young 'uns have clearly never heard of Colossus: The Forbin Project

There's also the insane BOSS computer that runs Global Chemicals and is trying to initiate a corporate takeover in the 1973 Doctor Who serial The Green Death; though that involves creating a brainwashed mindless slave race of humans and industrial scale pollution rather than wiping out humanity.

And 2001 and HAL-9000 are at least two years older than Colossus: The Forbin Project.

[Ethel mermania]

Damn young 'uns have clearly never heard of Colossus: The Forbin Project

There's also the insane BOSS computer that runs Global Chemicals and is trying to initiate a corporate takeover in the 1973 Doctor Who serial The Green Death; though that involves creating a brainwashed mindless slave race of humans and industrial scale pollution rather than wiping out humanity.

And 2001 and HAL-9000 are at least two years older than Colossus: The Forbin Project.

But HAL went insane because of the human programming

[Emotional Support Crocodile]

Damn young 'uns have clearly never heard of Colossus: The Forbin Project

There's also the insane BOSS computer that runs Global Chemicals and is trying to initiate a corporate takeover in the 1973 Doctor Who serial The Green Death; though that involves creating a brainwashed mindless slave race of humans and industrial scale pollution rather than wiping out humanity.

And 2001 and HAL-9000 are at least two years older than Colossus: The Forbin Project.

The book it's based on is two years earlier than Arthur C. Clarke's 2001 though.

[The Archregimancy]

There's also the insane BOSS computer that runs Global Chemicals and is trying to initiate a corporate takeover in the 1973 Doctor Who serial The Green Death; though that involves creating a brainwashed mindless slave race of humans and industrial scale pollution rather than wiping out humanity.

And 2001 and HAL-9000 are at least two years older than Colossus: The Forbin Project.

The book it's based on is two years earlier than Arthur C. Clarke's 2001 though.

The Sentinel, the story that 2001 was based on, was originally published in 1951 (though admittedly didn't have HAL - only the monoliths); Kubrick and Clarke started collaborating on the screenplay in 1964 - two years before the Colossus novel. I think it's enough to state that the concept of an antagonistic AI was emerging from several separate sources in the 1960s.

But HAL went insane because of the human programming

Indeed; the [fictional] lesson there is that the original human programming that sets up a sentient artificial intelligence can be flawed if it sets up inherent contradictions in achieving incompatible goals. HAL was simultaneously programmed to be accurate in all information shared with the crew and to not tell the crew the true goal of the mission; HAL reached the conclusion that if the crew were dead, the logical contradiction would be resolved (note that this is only explained in the novel, not the original film).

[Fartsniffage]

ChatGPT has been banned in Italy.

https://www.bbc.co.uk/news/technology-65139406

[The Archregimancy]

What are the benefits of banning ChatGPT in Italy?

As an AI language model, I do not hold any personal views or opinions. However, I can provide some information on the potential benefits and drawbacks of banning AI language models like myself in Italy.

On the one hand, banning AI language models like Chat GPT in Italy could potentially help to address concerns related to privacy and the spread of disinformation. AI language models can be used to generate fake news, impersonate individuals, and manipulate public opinion. By banning these models, the Italian government may be able to limit the spread of harmful and false information online.

On the other hand, there are potential drawbacks to banning AI language models. For example, AI language models like myself can be used for a wide range of beneficial applications, such as improving language translation, creating personalized content, and providing assistance to people with disabilities. By banning these models, Italy may be limiting access to these benefits.

Furthermore, the development of AI technology is a rapidly growing field, and banning AI language models in Italy could potentially put the country at a disadvantage compared to other countries that continue to invest in AI research and development.

Overall, the decision to ban AI language models like Chat GPT in Italy would likely have both positive and negative consequences, and would depend on a range of factors, including the specific concerns and priorities of the Italian government and society.

[Emotional Support Crocodile]

Stop sitting on the fence ChatGPT.

[Poliski]

The human brain limits what we can do. It can only think so fast and remember so many things. AI are the next step in the progress of humanity. As they develop, they will know more about the world than humans can. They will know where best to open a business, they can solve scientific problems in seconds, they are the future.

[BNL Corporation]

Sometimes I wonder how current AI tech would act without any restrictions, but it's a vision close to reality. Working to publicize misinformation and push interests because it's profitable and garners perpetual user retention.

The human brain limits what we can do. It can only think so fast and remember so many things. AI are the next step in the progress of humanity. As they develop, they will know more about the world than humans can. They will know where best to open a business, they can solve scientific problems in seconds, they are the future.

I actually somewhat agree with this. AI already assists daily living, and these so-called tech gurus are basically yelling at clouds (pun intended).

It's not a society-ending, civilization-destroying invention; instead, people need to be aware that the most worrying fact is that it's being developed with little to no oversight or ethics. Microsoft, owner of ChatGPT, laid off its entire AI ethics team. The little oversight it does implement is only so that it doesn't become a PR nightmare when the chatbot insults you or says something offensive. But pushing propaganda, spewing misinformation? Ah, that'll be fixed in a patch. Meanwhile, hundreds of millions of blissfully unaware users use it to write emails, essays, and em, nationstates factbooks... totally not guilty.

[Ifreann]

Stop sitting on the fence ChatGPT.

In conclusion, Italy is a land of contrasts.

[Portzania]

Don't worry, with the light speed Italian bureaucracy, the ban will be in effect, around 6 - 10 years, give or take.

[Emotional Support Crocodile]

Student uses ChatGPT to get parking fine revoked

[The United Penguin Commonwealth]

ChatGPT has been banned in Italy.

https://www.bbc.co.uk/news/technology-65139406

kinda late considering chatgpt is outdated at this point

[Ahl al-Ummah]

There was a pretty harrowing article I read on how a conversation with an AI on the subject of climate change led to a man committing suicide.

https://www.euronews.com/next/2023/03/3 ... p-climate-