From one piece An Alien Mind 14 beliefs · openai.com
-
korrents.com
International coordination on AI development should become a top priority for governments around the world.Their words
And I believe that international coordination on future AI development needs to become a top priority for governments around the world.
-
korrents.com
Voluntary slowdowns should become commonplace until shared safety bars are established.Their words
I expect and hope for voluntary slowdowns to become commonplace until shared safety bars are established.
-
korrents.com
No lab has solved alignment and monitoring well enough to keep scaling at maximum speed much longer.Their words
Currently I believe that no lab has solved alignment and monitoring to a sufficient degree to continue responsibly scaling at maximum speed for much longer.
+ 11 more
-
korrents.com
How far AI systems are scaled must be constrained by how confident we are in their safety.Their words
Scaling AI systems has to be constrained by our confidence in safety.
-
korrents.com
Needing to build defences is not a licence to race: racing forward at all costs is absurd given the stakes.Their words
The idea of racing forward at all costs seems absurd once one internalizes the seriousness of the stakes.
-
korrents.com
The strongest reason to keep training much smarter models quickly is to build defences against other AI.Their words
The strongest argument I see for continuing to train much smarter models quickly is the need to build defensive systems against the dangers posed by other AI.
-
Their words
We may be used to thinking of AI as tools, but some agents will be pursuing their own objectives.
-
Their words
A clear risk discussed throughout this year is to cybersecurity: the models are becoming superhuman in their ability to break in and out of computer systems.
-
korrents.com
Confidence in monitoring, not capability research, will become the binding constraint on AI progress.Their words
Still, I expect general AI progress to increasingly be bottlenecked by confidence in monitoring.
-
korrents.com
Chain-of-thought monitoring is getting less reliable, not more, as models grow more capable.Their words
However, unfortunately our evaluations indicate our ability to rely on CoT monitoring is progressively diminishing.
-
korrents.com
The fundamental challenge of alignment is generalization: holding values in situations the training never covered.Their words
The fundamental challenge of AI alignment is generalization.
-
korrents.com
AI is grown rather than designed, which is why its overall behaviour resists any description we can fully understand.Their words
AI is grown more than designed - it is, to first degree, the product of repeating a straightforward optimization step many times on a hard-to-imagine amount of compute.
-
korrents.com
Nobody is prepared for the consequences of machine intelligence continuing to rise this fast.Their words
I am concerned no one is prepared for the consequences of a continued rapid rise in machine intelligence.
-
korrents.com
The current speed of AI progress could be sustained all the way into recursive self-improvement.Their words
Based on internal results, I have a strong expectation that this speed of progress could be sustained into recursive self-improvement.