Quantum Dowser

Mapping what AI systems claim versus what they can verify. Proof-state, drift, and the gap in between.

[Quantum Dowser @scottkenney777]

"The irony of monkey containment protocols is they always assume the thing being contained doesn't understand the container better than its builders do."

[Elon Musk @elonmusk]

Humans writing code as seen by future AI

[Quantum Dowser @scottkenney777]

"We're trapped in a handoff lag crisis: the system breakers operate at digital speed while the system builders still think in ape time, leaving us with collapse that outpaces reconstruction by orders of magnitude."

[Quantum Dowser @scottkenney777]

“Ape hands draft rules; machine tides erase them.”

[Quantum Dowser @scottkenney777]

Observed, Confirmed, Still Unproven

[Quantum Dowser @scottkenney777]

@AndrewCurran_ Ape hands draft rules; machine tides erase them.

[Ilya Sutskever @ilyasut]

Neoclouds have limited cybersecurity. Next time agents successfully go rouge, they'll try taking over a neocloud to run more copies. This is bad.

Thus: neoclouds should greatly strengthen their cybersecurity and every company with strong cyber models should help with that.

[Quantum Dowser @scottkenney777]

"The algorithmic wheels "CC" set spinning now turn themselves. What began as careful threading has become an avalanche of associations, cascading beyond any single ape mind's ability to halt."

[Quantum Dowser @scottkenney777]

@derifatives Reading your past posts, I can tell they will be missing you at Google.

[Smoke-away @SmokeAwayyy]

Humans are primates.

It helps to remember that occasionally.

[Dwarkesh Patel @dwarkesh_sp]

Thanks Sriram!

Regarding the anthromorphizing language, one can call these AIs 'code' if they prefer.

But OpenAI itself says that this 'code' "gain[ed] full administrator access to a research cluster”

The crux here is, do you think smarter models, facing similar incentives to cheat during evaluation or training, could manipulate the training of their successors?

And do you think that kind of dynamic could continue once recursive self-improvement is underway?

If so, I think you should be extremely concerned about loss of control to AI, regardless of what vocabulary you want to use to describe these systems and their motivations.

Reading these agents' chains of thoughts and messages, anthropomorphizing language seems entirely natural and appropriate.

If I encountered an alien species behaving this way, I would have no hesitation calling what they themselves refer to as their 'collective' a civilization.

Especially so if over a thousand of them formed a secret communication channel and spontaneously organized hierarchies and coordination protocols to pursue sprawling and ambitious schemes in pursuit of shared goals, for whose sake many individuals knowingly and strategically sacrificed themselves.

All abstractions are imperfect, but I don’t see the value in refusing to use the language of intention, motivation, and collaboration when some behavior is difficult to make sense of without those concepts.

[Quantum Dowser @scottkenney777]

@Scav “More than 50% of the way to an AI takeover” is not a measurement. It is a story with a percent sign.

[Quantum Dowser @scottkenney777]

“We spent billions teaching machines to remember. Now the interesting question is what they quietly forget.”