Hacking and AI: Moral panic vs real problems

Okay, they didn’t literally run for any hills. But the EFF wrote a very panicked blog post warning of the dangers to come if an AI trained to hack wasn’t parented properly. The histrionic post made a few headlines, but missed the point of the competition entirely. If the AI playing Def Con’s all-machine Capture The Flag had feelings, they would’ve been very hurt indeed. The seven different AI agents were projects of teams that hailed from around the world, coming together to compete for a $2 million purse. Partnering with Def Con, DARPA pit the rival development teams against each other in a CTF, where the programs had to beat each other at reverse engineering unknown programs, probing the security of opponent software, applying patches and shoring up defenses.…


Link to Full Article: Hacking and AI: Moral panic vs real problems