1,100 signatories call for a pause in training ‘giant AI systems’

29 Mar 2023

Image: © Chanakon/Stock.adobe.com

The open letter is calling for a six month pause to the training of AI models more powerful than GPT-4, in order to develop new safety protocols for AI design.

A growing list of people, including CEOs and researchers, are calling for action to address the “profound risks” surrounding AI systems.

These 1,125 people (at time of writing) have signed an open letter calling for a six-month pause to the training of powerful AI models.

Some notable names in the open letter include SpaceX and Twitter CEO Elon Musk, Skype co-founder Jaan Tallinn, Apple co-founder Steve Wozniak and MIT researchers.

The open letter, created by the non-profit Future of Life Institute, wants this training pause to apply to AI systems that are “more powerful than GPT-4”, which is the latest model released by ChatGPT creator, OpenAI.

“This pause should be public and verifiable, and include all key actors,” the open letter states. “If such a pause cannot be enacted quickly, governments should step in and institute a moratorium.”

The letter calls for AI labs and independent experts to “jointly develop” a set of safety protocols for advanced AI design, which would be “rigorously audited and overseen by independent outside experts”.

The open letter also wants AI developers to work with policymakers to “dramatically accelerate development of robust AI governance systems”.

“AI research and development should be refocused on making today’s powerful, state-of-the-art systems more accurate, safe, interpretable, transparent, robust, aligned, trustworthy and loyal,” the letter reads.

Some of the signatories come from companies developing their own AI systems, such as Meta and Google researchers, and the CEO of Stability AI. Nobody from OpenAI has signed the letter at time of writing.

The rise of AI systems

In recent months, many companies have raced to integrate AI systems into their services, as certain developments have shaken up the sector.

The most significant of these was the launch of ChatGPT, the advanced chatbot that gathered a lot of global attention since it launched last November.

Microsoft has been working to integrate OpenAI’s technology into a variety of its products, such as an AI-powered Bing to challenge Google’s search engine dominance.

Google has responded with its own AI chatbot – Bard – to try to level the playing field, while other companies have been releasing their own customised contenders.

A recent report by Europol experts claimed that AI chatbots such as ChatGPT can exacerbate problems of disinformation, fraud and cybercrime.

10 things you need to know direct to your inbox every weekday. Sign up for the Daily Brief, Silicon Republic’s digest of essential sci-tech news.

Leigh Mc Gowran is a journalist with Silicon Republic