AI security activist group PauseAI coordinated global protests to demand a pause in the event of AI models more powerful than GPT-4.
Protesters gathered in 14 cities around the globe on Monday, including New York, London, Sydney and Sao Paulo, to attract attention to potential AI risks.
The declared aim of the protest in keeping with PauseAI website The aim is to “persuade the few powerful people (ministers) who will likely be attending the following AI Safety Summit (on May twenty second) to be the adults within the room.” It is as much as us to make it clear to them that the “The risks are too big to disregard and that they’re the one ones with the ability to resolve the issue.”
It's happening immediately. Our second international protest.https://t.co/FlF1y5jpO2 pic.twitter.com/RmOcRtbKiC
—PauseAI ⏸ (@PauseAI) May 12, 2024
The AI Seoul Summit, going down on May 21 and 22, is the six-month follow-up to the UK AI Safety Summit held last November. The initial enthusiasm for international cooperation on AI security appears to be waning as several participants have withdrawn from the upcoming summit.
PauseAI pragmatically admits that “we cannot expect countries or firms to risk their competitive advantage by pausing AI training runs for long periods of time when other countries or firms will not be doing the identical. That is why we want a worldwide pause.”
They want the foremost goal of the upcoming summit to be the creation of a global AI safety agency much like the IAEA. Ironically, that is something that OpenAI CEO Sam Altman also suggested.
On the identical day, the PauseAI protests took place, demanding a halt to more advanced models OpenAI released GPT-4o, the faster and more powerful version of its GPT-4 model.
While AI fans marveled at GPT-4o's latest capabilities, PauseAI demonstrators were less passionate about the exponential growth of recent AI models.
Will sit-ins, posters and leafleting be enough to influence world leaders to take the meaningful motion that PauseAI says humanity needs? Can the inexorable march of AI advances be halted long enough to make sure AI safety guardrails are in place?
PauseAI thinks it's possible. In a post on X, the group said: “AGI is just not inevitable. This requires hordes of engineers with salaries value tens of millions. It requires a completely functional and unrestricted supply chain of probably the most complex hardware. We must all allow these firms to jeopardize our future.”
The protests have launched effective motion against genetically modified foods, nuclear weapons and climate change. However, the inherent dangers of those problems are widely accepted by world leaders.
Conflicts over future AI security issues must compete with the growing excitement about tangible AI advantages that humanity is currently experiencing.
Will the PauseAI protesters achieve their goal and earn humanity's gratitude? Will their fears ultimately prove unfounded? Or will they be those to say sooner or later, “We told you so.”
