Amodei warns: exponential development of AI requires urgent security measures

Artificial intelligence technology is developing at an "exponential" pace, Anthropic's CEO told CBS, calling the situation a "warning signal" for the industry that requires serious security measures

Artificial intelligence is developing at an "exponential" pace: Dario Amodei, head of Anthropic, defining the situation as a "warning signal" for the sector that requires a series of security measures because, as already anticipated, without a slowdown, within 6-12 months, AI could be able to lead a swarm of agents capable of taking control of the entire Internet. "I don't think I fully understood what such rapid progress would really mean," Amodei admitted in an interview with CBS News, which will air on 'CBS News Sunday Morning' and of which a preview has been released.

Amodei: "Development is accelerating, we must slow down"

Amodei described the development of AI as a progression of "one, then two, then four, then eight, then 16, then 32", stating that we are currently "on the curve" and that this "is starting to get steep. It doesn't mean we have to panic today. It doesn't mean we have to block everything", continued the Italian-American manager, adding however that it is "an alarm signal: we have to slow down". Amodei noted that technology requires security measures that can keep up "with the speed at which technology itself evolves."

Read also

OpenAI ready to slow down development. Anthropic slows down virus research

The support of Musk and Altman

The interview came just hours after he published a blog post in which he urged AI companies to proceed more slowly and with greater caution in developing the technology, stressing that risk prevention is key. The proposal immediately found support from Elon Musk (SpaceX) and Sam Altman (OpenAI). Amodei outlined a three-point plan with the aim, he stressed, of "managing the pace of advancement of the technological frontier". The first concerns internal evaluators ("embedded evaluators"), the second democratic coordination and the third global coordination. If handled carefully, “AI may be the latest in a long line of technological miracles that have elevated and ennobled humanity,” he wrote in the post. “But, like many previous technologies, it carries risks, and because it is such a powerful technology, those risks are serious.”

Read also

Coxon and the others, those who evoke the Terminator scenario for AI