Nearly 2,000 leading figures working in the field of artificial intelligence (AI) have signed a letter calling for a pause of at least six months in the training of ‘giant’ AIs.
The open letter has been signed by Elon Musk, Apple co-founder Steve Wozniak, and engineers and researchers from Amazon, DeepMind, Google and Meta.
Twitter owner and Tesla CEO Musk was a co-founder of OpenAI, which has since developed the generative AI chatbot ChatGPT and its most advanced deep learning model, GPT-4.
He joined co-signees of the letter calling on all labs to pause training of AI systems more powerful than GPT-4 for at least six months so that the ramifications can be properly considered.
It also calls for governments to step in to impose such a moratorium if the major players involved cannot implement a pause quickly.
The letter’s authors quoted the Asilomar AI Principles, which say that the development of advanced AI could represent “a profound change” in life on Earth and should therefore be planned and executed “with commensurate care and resources”.
AI labs locked in “out-of-control” race
Instead, they argue, AI labs are engaged in an “out-of-control” race to produce ever more powerful AI systems that no one can reliably predict or control – including their creators.
The letter, coordinated by think tank the Future of Life Institute, cites another OpenAI co-founder, Sam Altman.
He wrote in February that it may be important “at some point” to seek “independent review” before training new systems and for the most advanced developers to agree to limit the growth in computing power used for new models.
The letter’s authors said that they agreed with Altman, but that the point to do so had already been reached.
They added that they want AI labs and developers to use the pause to come together and develop shared safety protocols for advanced AI, which can be overseen by independent experts.
They pointed out that they were not calling for any brakes to be applied on AI development in general, but only the “black box” models with “emergent capabilities”.
On Friday, Italy became the first Western country to ban ChatGPT, citing privacy concerns.
Today’s news was brought to you by TD SYNNEX – the UK’s number one solutions distributor.
