@evanhill What makes you think any political apparatus is remotely capable of reigning in rogue AGI? Spell out your steps. Safety ppl have already spelt out theirs: https://t.co/9LU0THEOJK
or:
https://t.co/5wI2oAW9Zn
and many more
We keep hearing again that Russia supposedly needs security guarantees. Even though no one has attacked Russia for many decades. Since the Second World War, Moscow has not gone through even one decade without starting a war or using military force against another country.
So when the Russians now talk about a security architecture for themselves, or about security guarantees for Russia – they are hardly talking about real security. And peace.
What they really want – is to freely interfere in the affairs of other countries, and to use force whenever they want, and to make sure that no one can support liberty inside Russia because Putin wants to keep his personal power forever and pass on his imperialist and, in many ways, simply fascist views to the next generations.
From a speech at the Danish Folketing. (3/3)
Every AI company is trying to fully automate AI research; to allow the AIs to build the next AI.
This is called Recursive Self Improvement (RSI).
OpenAI says they are on track to do so by March 2028.
Anthropic employees say they are on track to do so by 2027.
The METR report below suggests they are likely right.
This is a bad idea given current technology. Any mistake or misalignment in current AIs could be multiplied in each generation.
We don't get another election before AI companies reach this threshold.
Congress should ban RSI, this year if possible.
This is a remarkably consensus opinion, supported in various versions by:
-Bernie Sanders
-Steve Bannon
-Ralph Nader
-Glenn Beck
-Republican and Democratic attorneys general
-Whistleblower Daniel Kokotajlo
-Turing Award winner Yoshua Bengio
-Humans First
-Open Markets Institute
-Ezra Klein
-Arvind Narayanan (a Princeton computer scientist known for arguing that “AI is a normal technology”)
The exact mechanism can be debated. The ban doesn't have to be permanent either (see below for both). This is not the only thing we have to do in regards to AI, far from it! But it is urgent.
Congress must ban RSI.
Kemi, congratulations on leading a major British party as a Black Woman. It takes real talent to rise to where you are and deceit is what it takes to rise as a Black leader in a western Conservative Party.
It is a pity the climb did not leave room for a history book and since you chose to argue in the language of history, let us do the history. Let’s dissect your verbal diarrhea point by point.
Everyone participated." Slavery existed on every continent. So did war, and nobody calls a village feud the World War. The transatlantic trade ran for 366 years (1501 to 1866) and put about 12.5 million Africans on ships, of whom about 1.8 million died before they saw the Americas, according to the Trans-Atlantic Slave Trade Database. It ran through joint-stock companies, chartered monopolies, Lloyd's underwriters and acts of Parliament. It was hereditary, because the law made the child of an enslaved mother property from birth. It was racial, because a whole continent was defined as chattel. No other system in history combined those features at that scale. Dissolving all of it into "people from every continent" is far worse than intellectual dishonesty.
Britain was a market leader, not a participant. British ships carried about 3.2 million Africans, second only to Portugal and Brazil. The Treaty of Utrecht (1713) gave Britain the asiento, a thirty-year monopoly to supply 4,800 enslaved Africans a year to Spanish America. The Royal African Company, chartered in 1672 and headed by the Duke of York, later James II, branded its captives "DY". During Edward Colston's years in its leadership it shipped about 84,000 people, and roughly 19,000 died on the way.
You say British taxpayers should not answer for what happened before they were born. They did. In 1833 Britain borrowed £20 million to compensate slaveholders, and taxpayers serviced that loan until 2015, including the descendants of the enslaved. So the principle you invoke was not applied when the cheque went to slaveholders. It only appears when the cheque changes direction.
"Navel-gazing" is a curious charge from someone whose party spends more energy on the past (Empire nostalgia, culture war crusades) than most of those you accuse. Defend Britain if you must. But do not use your Nigerian name as the permission slip for arguments Nigerians know better than anyone.
I find it confusing that Dario keeps saying Anthropic's strategy is to drive a "race to the top".
It doesn't seem like there's much evidence that this strategy is working? Am I missing something?
This is a registration to disband Anthropic not to pump it full of investors money.
You’re exposing yourselves as an existential danger.
An international crisis meeting on Frontier AI is long overdue. Nation States have a short window to effect change.