Nick Clegg has defended the release of an open source artificial intelligence model by Mark Zuckerberg’s Meta, as he claimed that the “hype” about the dangers of AI was getting ahead of the development of the technology.
Meta’s global affairs chairman and former UK deputy prime minister spoke on Wednesday after the company said it was opening up access to its new large language model (LLM), Llama 2, which will be free for research and commercial use. LLMs are trained in big data and endorse generative artificial intelligence products like the ChatGPT chatbot.
Some experts have warned that making AI models open source, or freely available to use and adapt for unique purposes, could lead to the technology being used for malicious purposes.
Speaking on BBC Radio 4’s Today programme, Clegg said: “My view is that the hype has gotten a bit ahead of the technology. I think a lot of the existential caveats relate to models that don’t currently exist, the so-called super-intelligent, super-powerful AI models: the view that AI develops autonomy and agency for itself, where it can think for itself. . and reproduce.
“The models that we are opening are very, very, very inferior to that. In fact, in many ways they are quite stupid.”
Clegg said that Meta had 350 people “testing” its models over several months for potential problems, and Llama 2 was more secure than any other large language model available, such as OpenAI’s ChatGPT.
He said that “it’s not like we’re at a T-junction” where companies can “choose to open source or not,” adding: “Models are already being open sourced all the time.”
While Clegg said he strongly believed AI should be regulated, he added that that “doesn’t mean all AI models need to be open source.”
In May, one of its engineers warned Google that it could be beaten in the AI race by commonly available open source technology. “The inconvenient truth is that we are not positioned to win this arms race and neither is OpenAI. While we were fighting, a third faction was quietly eating our lunch,” the engineer wrote, referring to the open source AI.