Meta's Nick Clegg Debunks AI Election Threat: A Sword and Shield for Democracy
According The Guardian's coverage: in the realm of major elections worldwide this year, there seems to be a reassuring lack of systematic malevolent interference, as indicated by Meta's global affairs chief, Nick Clegg.
Clegg dismisses the notion that generative AI poses a significant threat to elections, asserting instead that such technology can be more beneficial for safeguarding democracy than undermining it.
Speaking at the Meta AI Day event in London, Clegg pointed out that in major elections held across the globe this year, tools like large language models, image and video generators, and speech synthesis tools have not been extensively utilized to undermine democracy.
"While it's important to remain vigilant," Clegg stated, "the evidence from recent major elections, including those in Taiwan, Pakistan, Bangladesh, and Indonesia, suggests that these AI tools have not been systematically employed to disrupt the democratic process."
Clegg urged a perspective shift on AI, likening it to a double-edged sword rather than just a shield against harmful content. He credited AI as the primary reason for the continuous improvement in reducing unwanted content across Meta's platforms.
Furthermore, Clegg noted that Meta is collaborating with industry peers to enhance these systems further, especially given the unprecedented number of elections this year.
However, the landscape might undergo changes soon due to Meta's own initiatives in this domain. The imminent launch of Llama 3, Meta's most advanced GPT-style large language model, is expected to occur in the following weeks, with a full rollout anticipated by summer.
Unlike its counterparts, Meta has traditionally released AI models as open source, allowing for broader usage but also raising concerns about misuse by malicious actors. Nevertheless, this approach facilitates scrutiny by external observers, enabling them to assess the accuracy and bias of these systems more effectively.
Yann LeCun, Meta's chief AI scientist, cautioned against another potential threat to democracy posed by AI: the risk of a few closed models dominating the landscape. LeCun emphasized the importance of diversity in AI systems to ensure they cater to various languages, cultures, and value systems worldwide.
"In the near future, every interaction with the digital world will be mediated by AI assistants," LeCun predicted. "It's imperative that these systems are diverse, akin to the necessity of a free and varied press. Given that every AI system carries inherent biases, it's crucial to have a multitude of perspectives represented, which cannot be achieved by a handful of companies on the west coast of the US alone."