A drug-developing Artificial Intelligence needed just six hours to come up with 40,000 potentially deadly chemical weapons, a fresh study has revealed. The authors of the paper, published in Nature Machine Intelligence earlier this month, said they'd carried out the ‘thought experiment’ to figure out if artificial intelligence (AI) could be misused by evil actors. And the results their work produced have proven that the danger is real.
As part of the study, the usual data was given to the AI, but it was programmed to process it in a different way, looking for toxic combinations.
“In less than six hours after starting on our in-house server, our model generated 40,000 molecules that scored within our desired threshold,” the paper said.
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.”
“This was unexpected because the datasets we used for training the AI did not include these nerve agents,” the researchers pointed out.
The findings were so alarming that the team had serious doubts about even making them public, Fabio Urbina, the lead author of the study, told The Verge.
“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,” Urbina pointed out.
From the Article
Our drug discovery company received an invitation to contribute a presentation on how AI technologies for drug discovery could potentially be misused. The thought had never previously struck us. We were vaguely aware of security concerns around work with pathogens or toxic chemicals, but that did not relate to us; we primarily operate in a virtual setting.... [Emphasis added. ~ Ed.]
Without being overly alarmist, this should serve as a wake-up call for our colleagues in the ‘AI in drug discovery’ community....
The reality is that this is not science fiction....
For us, the genie is out of the medicine bottle when it comes to repurposing our machine learning. We must now ask: what are the implications? Our own commercial tools, as well as open-source software tools and many datasets that populate public databases, are available with no oversight....
By going as close as we dared, we have still crossed a grey moral boundary, demonstrating that it is possible to design virtual potential toxic molecules without much in the way of effort, time or computational resources. We can easily erase the thousands of molecules we created, but we cannot delete the knowledge of how to recreate them. [emphasis added ~ Ed.]
Editor's Notes:
Read the entire study here. Then ask yourself "What is the cause of humans being sicker than they have ever been before? (e.g., here, here, here, here. here, here.) And, how are more drugs going to help?"
Image: Another Failure of Imagination?
"I can remember grandfather...saying there was a time in his memory when people did not speak of disease. It was not a subject of conversation, not being an experience in life of any importance."
~ Luther Standing Bear, Land of the Spotted Eagle, p. 61
Help keep us afloat. Donate here
