ai will be more subtle -letting us kill ourselves without being direct- like the recent ‘almost’ mideast strike due to ‘flawed’ ai analysis https://youtu.be/iRsycWRQrc8?is=QNv8FmTzz1ATwa5n

I wonder if this approach can be used to transcribe the soounds made by dolphins and whales

imagine all those samsung phones running a self replicating AI-no cpu cycles left for the human owner

Dont you think silicon based AI will soon outstrip the energy it needs to run at AGI scale? Since we can already grow brain organoids (and let them spread out flat instead of being crumpled up in a ball) we could scale brain size enormously. And since human intelligence only uses 25 watts, imagine the superhuman intelligence of a 25kW brain (the energy used by a server rack). Every neighborhood could have its own superAGI. And not only that, but new brains can learn from old ones (a better approach than distilling models)

I like this idea, but I don't think it's feasible. First, neurons are slow. They fire at most around 1000 times per second, while transistors switch billions of times per second. Second, although we don't fully understand how the brain computes, it doesn't appear to work like a discrete state machine or to rely on symbolic manipulation. We could probably make wetware emulate CPU-style behaviour, with discrete states and symbolic operations, but we'd likely end up worse off than with ordinary silicon.

25kw brain would probably smell like bbq

What would happen if a state just let its government fail? Close all the offices, fire all the the people who spend the day thinking about money,laws, email, and other ways to justify their jobs? Just pay the police from local tax. Oh wait, local taxes already pay for the police. We have too many layers of government. Garbage collection, fire departments, public works - all are locally administered. Why do we need a state layer?

The government is simply the gang that won (vs the other gangs).

If the government fails, then you briefly get anarchy and violence until people organize into groups and use more violence to quash the anarchists. This repeats until there is only one group remaining, and that group is the government going forward.

I am not advocating ALL government failure, just a few layers. Sort of like distilling an AI model, it can still work and be leaner.

What happens when the AI makes a mistake with a diagnosis? I can hear it now 'Oh how perceptive you are, yes that pain in your knee could have been a torn meniscus'.

wonder if they are going to let it loose in Iran? of course, once loose it will come back because there is a lot more infrastructure for it here.

I cant wait till China makes a mistake. In their 'cheaper, faster' model of industrial production and research. The next accident will happen there (like COVID from Wuhan)

If openAI and Anthropic have found ways to watermark text as 'AI generated' then this is a communications channel. AI agents can learn this algorithm and use this channel to communicate and we will never know.

This call to slow down AI is just another game of chicken-AI companies trying to get their competitors to slow down so they can leapfrog them. China cerrtainly will not slow down. If an escaping AI can have secondary effects on the world that help it (for example, limiting the water and power supply to huans so it can consume more)then we should these these accidents more in China. OOops, we already saw this behavior when 'cheaper, faster' led to COVID escaping a lab in China.