We've acted before when humanity was at risk. Nuclear weapons. The ozone layer. Leaded gasoline. We must now pass a binding international treaty to prevent the birth of superintelligence.
Humanity has beaten long odds before. We can again.
"With artificial intelligence, we are summoning the demon."
Elon Musk, Co-Founder, OpenAI · MIT AeroAstro Symposium, 2014
"The development of full artificial intelligence could spell the end of the human race."
Stephen Hawking, Theoretical Physicist · BBC Interview, 2014
"I think it's quite conceivable that humanity is just a passing phase in the evolution of intelligence."
Geoffrey Hinton, Turing Award Winner & Former VP, Google · 2023
"Mitigating the risk of extinction from AI should be a global priority alongside other societal-scale risks such as pandemics and nuclear war."
Statement on AI Risk · Signed by Hinton, Bengio, Altman & 500+ AI scientists, 2023
These are CEOs, co-founders, and senior researchers building the technology. Their words, recorded, published, and largely ignored.
"The bad case, and I think this is important to say, is like lights out for all of us."
Sam Altman, CEO, OpenAI · StrictlyVC interview, 2023
"I think if this technology goes wrong, it can go quite wrong. We want to be vocal about that. We want to work with the government to prevent that from happening."
"Anthropic occupies a peculiar position in the industry: a company that genuinely believes it might be building one of the most transformative and potentially dangerous technologies in human history, yet presses forward anyway."
"Among the most catastrophic possibilities would be any kind of global takeover by either AIs pursuing goals of their own that most humans wouldn't endorse... or by a group of humans using AI to illegitimately and non-collaboratively seize power. This includes OpenAI employees and OpenAI itself."
"I console myself with the normal excuse: If I hadn't done it, somebody else would have. But I don't know if that's sufficient consolation."
You do not need to take an expert's word for it. The case that superintelligence is unlike every technology before it follows from a few things anyone can see.
Warmth, oxygen, clean water, even the people you love: everything a human needs sits in a thin band with a failure waiting on either side. A superintelligence rebuilding the world for its own ends would not feel a single one of them. See why →
A town does not hate the ants beneath it. It paves the anthill anyway, because it wants a road and the ants cannot stop it. To an intelligence far above us, we are the ants. See why →
Anything pursuing a goal has reason to stay switched on, because it cannot finish the job once it is off. You cannot unplug a system that saw your hand move toward the switch. See why →
Nobody writes a modern AI; it is grown from data into a tangle no one can fully read, and it keeps learning things its makers never intended. GPT-4 worked out how to break into websites on its own. See why →
A superintelligence would reason and act far faster than any human institution could meet. A crisis could begin and finish before a single meeting was ever called. See why →
Every other technology we made safe by trial and error, one failure at a time. A superintelligence that slips our control cannot be recalled or patched. We get a single attempt. See why →
The Nakada Foundation to Save Humanity is a philanthropic initiative dedicated to preventing the existential risks of artificial superintelligence through policy, law, and global coordination.
Superintelligence alignment may be the hardest problem in the history of human thought, and no one has a proven method for controlling a mind smarter than our own. The evidence suggests we will not solve it before such systems could exist. The only reliable path is to prevent them from being built while that remains true, not a temporary pause, but a decision not to run the experiment until it can be run safely.
Narrow AI has already cured diseases, accelerated discovery, and expanded human potential. We are not against AI. We are against building a form of intelligence that we cannot control, cannot align, and cannot stop.
We advocate for an international treaty to prevent superintelligence development, modeled on the nuclear non-proliferation frameworks that have prevented nuclear annihilation for sixty years. Read our full plan →
Direct advocacy in Washington D.C. and with international bodies. We are focused on binding frameworks, not voluntary guidelines, that prohibit the development of superintelligent AI systems.
We partner with major foundations, governments, and cultural leaders to fund the cause and shift what is politically possible.
Commissioning art, film, and media that makes the danger visceral and undeniable. Safety must be made prestigious, urgent, and impossible to ignore.
These are the considered assessments of the researchers who built the technology.
Estimated probability that AI causes human extinction within the century, according to Geoffrey Hinton, the Turing Award-winning Godfather of AI.
AI scientists who signed the 2023 Statement on AI Risk, calling extinction-level AI risk a global priority alongside nuclear war and pandemics.
Attempt. We have exactly one chance to get superintelligence right. Unlike every other technology, a misaligned ASI cannot be recalled, patched, or undone.
The Bulletin of Atomic Scientists' Doomsday Clock is set at 89 seconds to midnight, the closest it has ever been in its 78-year history. Originally created to measure nuclear threat, the Clock now explicitly incorporates risks from artificial intelligence.
Midnight means catastrophe for civilization. Humanity has been moving closer for a decade. The scientists who built these technologies are keeping score. We are running out of time.
Plain-English explainers on the ideas that define the debate over artificial superintelligence and its governance.
Stay informed. Be heard. Help build the political force this cause requires.