Over 8 years ago: 😮
People sometimes ask if I think it's risky for everyone to have access to AI. I think it's MORE risky for an exclusive & homogeneous group alone to develop tech that impacts us all.
Thomas confirmed her stance remains unchanged eight years later.
Over 8 years ago: 😮
People sometimes ask if I think it's risky for everyone to have access to AI. I think it's MORE risky for an exclusive & homogeneous group alone to develop tech that impacts us all.
Positive users agree that broad AI access reduces risks via mutual deterrence like Oppenheimer's view, while negative users suspect the warnings mask profit motives and corporate double standards.

@jeremyphoward it's going to be the same as always. "It's too dangerous for you to use! ... Unless, you would be willing to pay us, then we might reconsider. "
even on gov level and selling GPUs to China they did the same. Threat to humanity, unless there's profit to be made.

@jeremyphoward mythos with no guardrails will kill us all (25% chance of human extinction, their words ) but hand it to amazon, apple, microsoft, google, jpmorgan and 150 other corporate? Suddenly it's fine. turns out the apocalypse is only dangerous when poor people have access to it

@jeremyphoward The woes we feared are fallen from the skies, and weigh in earnest on each mortal head.

@jeremyphoward The only constant is how much changes in 8 years, yet some takes from back then still hit different now. 🧠

@jeremyphoward but the examples missed the on the high stakes on today's world.

@math_rachel Quite the prophecy.

@jeremyphoward Eight years on, the empirical record backs this up: misuse studies keep finding closed models in real attack chains too — access restrictions changed the threat model less than predicted. The homogeneity risk got less attention but aged even better.

yes i also think n agree with Oppenheimer - if everyone has it it’s a great stabilizer as mutual destruction is guaranteed - u give n get respect n treated fairly in the world order
same goes for AI
surprisingly ppl who compare ai with nukes (i don’t) get ai policy wrong n as well as nuclear policy

@jeremyphoward You both were always ahead of the rest … Thanks to our conversations back then, @EDAI_Ltd was always build with AI safety and governance in mind.

@jeremyphoward Crazy how fast things move now.

@math_rachel @jeremyphoward If everyone builds it, nobody dies

@jeremyphoward Wild to see how far things have come
Eight years feels like a lifetime in this space
Thomas confirmed her stance remains unchanged eight years later.
Over 8 years ago: 😮
People sometimes ask if I think it's risky for everyone to have access to AI. I think it's MORE risky for an exclusive & homogeneous group alone to develop tech that impacts us all.