It’s hard to imagine textbooks are going to thrive in an AI-dominated future, so suspend your disbelief as you imagine schoolchildren turning to this page full of key dates:
- March 29, 2023: More than 1k tech leaders urge a pause on AI development, signing a letter about it presenting “profound risks to society and humanity.”
- April 27, 2023: Tech pioneer Geoffrey Hinton, AKA “The Godfather of AI,” resigns from Google, saying he regrets his life’s work.
- Nov. 17, 2023: A power struggle over safety temporarily deposes OpenAI CEO Sam Altman. Speed wins.
- Feb. 12, 2024: AI safety researcher Dr. Roman Yampolskiy says there’s no proof AI can be controlled. One month later, he gives humanity a “99.999999%” probability of doom.
- March 11, 2024: A report commissioned by the US State Department warns AI could lead to human extinction.
- May 17, 2024: OpenAI dissolves its safety team.
Many other largely ignored, experts-sounding-alarms moments could’ve fit here, and this week offered one more candidate: a group of AI workers signed another open letter.
But this one felt different
It sang a more resigned tune than past letters, not harping on AI’s “serious risks” like its predecessors, instead focusing on protections for whistleblowers who want to speak out about those risks.
The letter’s 13 signatories seem to already know the counterpoint to any direct appeal for safety: money. Trillions of dollars worth of it.
- For every signature on an open letter, tens of billions of dollars fly into AI from Big Tech, VC firms, and government subsidies.
Safety schmafety
We’re too deep into an all-out arms race for warnings of doom to land, and as the stakes rise, the rate of development will only accelerate, per Axios.
- Google was moving slowly on AI until OpenAI got too far ahead; OpenAI, in turn, will move faster as Google quickens its pace.
- China and the US are engaged in a similar dance.
Those who want the boulder to stop rolling downhill should continue speaking out, and AI employees who keep signing letters should be commended for risking retribution to do so.
But the monied interests driving the AI revolution are unlikely to pay much attention at this point; they are far beyond open letters.
