OpenAI CEO warns that GPT-4 could be misused for nefarious purposes

OpenAI's CEO Sam Altman admitted in a television interview that he's "a little bit scared" of the power and risks language models pose to society. Altman warned that their ability to automatically generate text, images, or code could be used to launch disinformation campaigns or cyber attacks. The technology could be abused by …

[cyberdemon]

OpenAI CEO warns that the forests of California may be used as a toilet by bears, and the Vatican City may be used for Catholicism by the Pope

In other news, The Devil has noticed that GPT-4 is extremely useful for producing believable bullshit on any subject/context, on an almost infinite scale, but has absolutely no idea what he will do with it...

[b0llchit]

...but has absolutely no idea what he will do with it...

Well, lets see... Very useful for the following:

• Manipulation
• Fraud
• Manipulation
• Fraud
• Manipulation
• Fraud

Did I forget one? Oh yes, I think I forgot to mention manipulation and fraud.

[Anonymous Coward]

I think you forgot the combination of the two.

:)

[eldakka]

Fraudulent Manipulation and Manipulative Fraud?

[FF22]

Modus operandi

The "use for nefarious purposes" is not what it could be used for, but what it WILL be used for, in the majority of situations.

That's because its result are not reliable and it's not actually intelligent, therefore it can not be used in any situation where quality and reliability matters. It's only good for entertainment purposes and where it's the sheer amount of stuff is generated that's matters, and quality, verifiability, etc. is not.

It will be used for stuff like overloading people and generally the web with false information or just low quality content, filling up a ticketing system with false reports, clogging up any system that needs some kind of human intervention to validate the entries posted on it. Coupled with other AIs it will be able to emulate real people's voice and face reliably, and thus make anything but personal meetings unreliable when it comes to identity of the persons involved, and security.

That's what it will be used for primarily, and the public internet will be an even worse place than it already is.

[Anonymous Coward]

Re: Modus operandi

www.linkedin.com/posts/itsmrwolf_leadership-empowerment-mindset-activity-7042900098082611200-ZTeq

You WILL be able to talk to [enter your selection] You WILL be able to see pope having sex with [enter your selection]. Not into popes? How about... Missed a selfie, why stress? Choose time, place, facial expression and companions and don't forget to subscribe to PREMIUM! FEATURES!

I imagine systems will appear that will 'prove' that this or that recording of an event 'really happened' and googles and microsoft making money off licencing such systems off to the plebs and states. Once in a while you'll have a 'rogue engineer' story who manipulated those 'genuine-certified' systems so that the kompromat on a political opponent turns out to be fake too.

But that's short term, ultimately though, sooner than later, nothing that you don't see with your own eyes (no implants please!) will be reliably 'real'. It would be interesting how, if at all, this is going to affect human behaviour in general, if at all. Ah, there was a book on a vaguely similar theme once, Light of other days.

[elsergiovolador]

Manipulation

GPT-4 is very useful at unpacking and decomposing manipulation techniques that government is using through the army of behavioural psychologists at their disposal.

The level of gaslighting that is going on regarding the messaging about certain policies that benefit the rich, but are reframed as being created for the good of everyone is astonishing.

Unfortunately they keep blocking these prompts as if this kind of research was unethical, but I wouldn't be surprised if the rich and government agencies had unfettered access to it.

[Anonymous Coward]

Re: unpacking and decomposing manipulation techniques that government is using

you think that the governments are not aware and sitting idly? I would presume they're already in discussion on how to join the fun by getting some 'special' licence on the top systems, or taking over completely. Think of the danger if it fell in the wrong hands! Think of the good it could achieve in the right hands! Our hands!

[David Hicklin]

Re: Manipulation

You don't need GPT-4 to determine that, its easy to spot that the policies only benefit the rich

[The Velveteen Hangnail]

Could?

Wow, is the AI reality distortion field finally starting to crack?

[CatWithChainsaw]

Virtue Signaling by any other name

"We've got to be careful here," he told ABCNews. "I think people should be happy that we are a little bit scared of this."

Yeah, I'm so happy that you're so scared that you did exactly what you were going to do anyway. Go look for a pat on the back and a cookie somewhere else, Sammy.

[C. P. Cosgrove]

Problens, always problems

On the 22nd February 'the Register' published this article - https://www.theregister.com/2023/02/22/clarkesworld_ai_spam/ - in which ''Little Mouse' said in a comment

"Starting a year or two ago, I noticed the emergence of a style of IT troubleshooting "how-to" web pages that had clearly scraped the technical steps from another source, and then padded it with a lot of infuriating waffle."

This is happening on an as yet smallish but annoying scale on technical forums. I, and my fellow moderators, on Bleeping Computer know it is happening but because the text is modified it is almost impossible to identify the source of their quoted material. Without knowing the source it is impossible to justify taking action against these posters for breaching the rules on copyright.

A week or so later 'The Register' published another article, for which I didn't save the link, but it referenced research at Stanford University aiming at identifying AI generated text complete with a demonstration on-line tool to test for such text. The tool specifically states it is merely a demo version and is not to be relied on as evidence but it gives hope that reliable tools for the detection of AI generated text may become available in the nearish future. Then I will be able to take action ! See - https://hai.stanford.edu/news/human-writer-or-ai-scholars-build-detection-tool

Chris Cosgrove

[David Hicklin]

Re: Problens, always problems

So the AI arms race begins.....

[Anonymous Coward]

"a little bit scared"

but hey, fuck it, I'm only following orders / my retirement plan

[Anonymous Coward]

AI will create new jobs for artists and novel ways of producing music

he's not an idiot, so he's lying about new jobs (but he means it about 'new ways of producing music')

[greenwood-IT]

Re: AI will create new jobs for artists and novel ways of producing music

I'd like to ask him how many musicians received any money from the event in question? They already got rid of the DJ.

[amanfromMars 1]

All your muppets and puppets belong to us <s>Truth Believers</s> :-)

OpenAI's CEO Sam Altman admitted in a television interview that he's "a little bit scared" of the power and risks language models pose to society.

FFS ...... and he has nothing to say about the constant and continual terrorisation of populations, both ancient past and recent present, by traditional main stream media channels and their starring spokespersons ...... via ye olde worlde create a distracting problem and offer an unsuccessful solution to generate a further problem methodology?

Is he not excited by the novel opportunities delivered by AI and ITs Virtual Machinery which are certainly possible and therefore all too likely to be energised and exercised?

Does he have no faith in otherworldly out of this world, GODly systems ...... Global Operating Devices?

[SEDT]

Oh! C'mon folks

Every new technology, whether the horseless carriage, printing press, telephone etc, can be and is utilised for nefarious purposes. And in all likelihood was kick-started by a government with war and population control foremost among its motivations.

Nevertheless, I am happy to have enjoyed the manifold benefits that the likes of electricity, computing, medicine, mobility and telephony have brought.

The crux here is about balance. The balance between the undoubted harms that accompany every technology, and the benefits.

Early computing was met with dismay by many but we have since learned that instead of doing us all out of work computers brought freedom from many repetitive and soul-destroying tasks. Remote working would have been a stretch back in the 50s.

Let's stop expounding on the negatives and lead the way in exploring and discovering the potential of this, and whatever else new tools that we create. Bah!

[RLWatkins]

"There are other people who won't put limits on it...."

Let's translate this into English: "We want to be the only company that is allowed to develop this. Those other researchers have to be stopped."