AI invents 40,000 chemical weapons in just 6 hours

A drug-developing Artificial Intelligence needed just six hours to come up with 40,000 potentially deadly chemical weapons

The authors of the paper . . . carried out the ‘thought experiment’ to figure out if artificial intelligence (AI) could be misused by evil actors.

It came up not just with the VX compound, which is one of the most dangerous nerve agents ever created, but also with some unknown molecules, “predicted to be more toxic.”

The dataset they used on the AI could be downloaded for free and they worry that all it takes is some coding knowledge to turn a good AI into a chemical weapon-making machine,”

Source
Source 2 in case Source 1 is blocked
Journal Article

We've been hearing more and more about both the potential benefits and dangers of artificial intelligence. While artificial intelligence could have unimaginable benefits, it could also prove to be extremely dangerous, as this article points out.
Researchers set up a study to use artificial intelligence to look for toxic chemical molecules. The AI came up with 40,000 in just 6 hours - some of which were more toxic than VX nerve agent. What's alarming is that this ability to use artificial intelligence to find chemicals such as these is hypothetically available to the general public.

We're at a very interesting point in technological development. We've opened Pandora's box with artificial intelligence. Soon there will be no way to prevent any random person from figuring out how to develop any number of deadly chemical weapons from the comfort of their own home. As the article says, all it takes is some coding knowledge. How do you protect against something like that?

a reply to: trollz

How do you protect against something like that?

Upload your consciousness to the cloud

a reply to: trollz

Imagine how many new, incurable diseases it could invent... plus lab viruses!

Oh wait... Chinese AI already did that!

Damn and I can only come up with a few thousand in an hour.

Meh, six hours? I've done it accidentally dozens of times. And came up with ways to combat it. Lol.

Nobody was ever hurt but we did have to evacuate many times.

a reply to: sine.nomine

Your forays into Taco Bell are a national security issue.

a reply to: trollz

So we're going to need a 8390457913804581903 character alphabet for the new Rona variants? That's what this means?

Stupid robots...

Interesting.... how come it couldn't find 40.000 cures for cancer or hiv?

Why the hell are they even looking into things like this?

Is it a sports game to them?

Unreal...

"Hey guise, lets see how many potent deadly things we can make... aaaaaaannnd GO!"

There have been many many warnings over the years about AI's , Warnings from people and Warning signs from the AI's them selves .

James Cameron has made several Blockbuster hits subversively warning us about AI's , He has made it very clear .

The Problems with AI's is the lack of Conscience the lack of Emotions , and normally I'm not one to lean on emotions but it is after all what makes us human.

No matter how much good AI can do, there's always the other side- the one that world leaders will use for power, control and financial gain.

I can imagine a day when the 1% and a select few 'cream of the crop' are vaccinated just prior to the release of a new, deadly virus for which there is claimed to be 'no vaccine available'.

If a nuke won't do us this will be next , but of course we can fear ourselves to death or just live...

Uhm... why couldn't someone run a program to have the AI figure out how to enable a freer, kinder society where everyone lives a long time, has lovely necessities and as complete a sense of autonomy as is possible in a society?

AI is a wonderful tool and at some level we're simply wet AI, but the initial programming is important!

originally posted by: flice
Interesting.... how come it couldn't find 40.000 cures for cancer or hiv?

It was specifically looking for toxic chemical molecules. It could presumably also be used to search for cures for things.

originally posted by: trollz

originally posted by: flice
Interesting.... how come it couldn't find 40.000 cures for cancer or hiv?

It was specifically looking for toxic chemical molecules. It could presumably also be used to search for cures for things.

And there folks, is more ongoing Proof of Death.
It would be a shoo-in for AI to devise and compound a
cure for everything wrong with us...
if the creatures running the planet were noble instead of evil.
Contortass outa that one, I'll wait.
^^^ That first handle isn't copyrighted BTW

a reply to: trollz

Up Next - A.I. Invents a Planet Killing Human Virus that gets Rid of All of Us Except the Elite............Hmm...

Do people want Skynet?

Cuz that's how you get SKynet.

a reply to: trollz

Wouldn't this 'speculative' research be prosecutable under the Geneva Convention?

I seriously doubt this was a simple “thought experiment”

I imagine someone in the arms industry is very, very pleased with the results, especially thanks to the previously unknown compounds,