Last year’s inaugural global AI Safety Summit in Britain convened world leaders and tech titans like Elon Musk and Sam Altman to grapple with the potential dangers of artificial intelligence. Now, as the second summit nears, the conversation has transformed.
Early anxieties centred on AI posing an existential threat to humanity. Today, the spotlight shines on the technology’s current limitations and resource constraints. Experts like Martha Bennett emphasize the immense data and energy demands of AI development. OpenAI’s CEO even posits that breakthroughs in energy and chip production are essential for AI’s advancement. This aligns with concerns regarding the environmental impact of data centres, a growing factor in the policy discourse around AI.
Professor Stilgoe urges caution against overhyping AI’s capabilities. He anticipates that practical applications will likely diverge from the grand visions of tech giants like Musk and Altman. This is evident in the recent stock market response to Meta’s ramped-up AI investment, which suggests a growing scepticism towards overly ambitious promises.
The upcoming Seoul summit, a virtual precursor to the next in-person gathering in Paris, is expected to be a more modest affair. Key figures from the EU and several governments, including Canada and the Netherlands, will be absent. This might be due to the postponement of the Paris event and the shift in focus towards more practical concerns, such as data access, copyright issues, and the environmental impact of AI development.
While the initial excitement has waned, international cooperation remains critical. Experts like Linda Griffin acknowledge that these summits need time to establish a productive dialogue and find their footing. Organizations like Mozilla prioritize attending the Paris event, which may take centre stage again after the Seoul summit tackles practical challenges. The future of AI safety hinges on addressing current limitations and resource needs, alongside responsible development and international collaboration. The upcoming summits offer a platform for these discussions to mature, ensuring that AI fulfils its potential for positive societal impact while mitigating potential risks.
