5 ms·

I think the most likely way to achieve “AI alignment” is the peasants with crossbows. Everyone having a powerful AI at their fingertips provides a balancing fac

I think the most likely way to achieve “AI alignment” is the peasants with crossbows. Everyone having a powerful AI at their fingertips provides a balancing factor that is only overcome if those tools are in the hands of very few. Trusting Google and various governments to “do the right thing” is absurd - even assuming good behavior and intention on their part, they just aren’t nearly as competent as people seem to assume. Everyone having access everywhere all at once provides the transparency and space to innovate, but also to counterbalance the use of AI in dark ways. I fundamentally don’t like the crossbow and guns and second amendment argument. AI isn’t a tool to murder with, having no other function than the death of your neighbor. I feel like it’s more like the “should you be required to license your seeds from a megacorp” type thing. Food is existentially important, and it could be weaponized. But it’s not fundamentally a weapon, it’s fundamentally a thing of utility.

hackernewds 3y ago
If you replace "peasants with arrows", with "good guy with a gun" I think the whole argument rings a different tone..
- iudqnolq 3y ago
  It's less "good guy with a gun" and more "some random country with a small to medium army". Access to violence doesn't democratize on the individual level, but as the parent comment alludes to weapons changes allowed peasant armies to defeat trained knights, changing the structure of society. You personally might not be able to exploit open AI fully, but there are lots of countries and orgs that could. Remember millions of dollars is loose change in the defense context.
coffeebeqn 3y ago
A finite resource fairly allocated to all? Which civilization are we talking about again? Certainly not humans. Also a billion or trillion or infinite ants can’t outsmart a single human. So unless some hard cap is enforced (how?) then mini AIs will probably not do anything
- czl 3y ago
  > Also a billion or trillion or infinite ants can’t outsmart a single human. Cells are simpler than ants so can trillion or infinite cells outsmart a single human? Why not? How many cells is a single human?
  - kig 3y ago
    Around 37 trillion cells, of which 25 trillion are red blood cells. Plus a 100 trillion gut bacteria. The brain is about 0.1 trillion cells, of which 0.02 trillion cells are the cerebral cortex, which is occasionally consulted by the rest of the brain on matters such as what words to enter into this textarea.
trident5000 3y ago
We definitely got a small taste of what people do with their time when they are not at work during Covid...and it wasnt pretty.
- Cthulhu_ 3y ago
  Not pretty for who? Can you explain?
  - trident5000 3y ago
    The summer riots that were fueled by people not needing to be anywhere.
TheRoque 3y ago
Well somehow I trust google more on having my data, than anyone in the world. I prefer the knowledge to be in a known identifiable entity that can be accounted for, than run in the wild and be used by anyone with bad intention (once again, regarding my data). So this is not about data, but AI, but I think it still applies.
- hackernewds 3y ago
  this is how I schedule my daily life. it is inevitable that I use software services. I'd like to keep them limited to one megalith, and that is the Google ecosystem. at least their biz model is not predatory, where using my individual data exceeds the risk of doing so.
- zelphirkalt 3y ago
  > I prefer the knowledge to be in a known identifiable entity that can be accounted for […] Google can be accounted for? As in "I know the entity that has the data." or "can be held accountable" or "will be held accountable"?
otabdeveloper4 3y ago
> powerful AI AI is not powerful. AI doesn't increase information complexity or create new information. It actually does the opposite - it averages out the complexity peaks. In effect, AI isn't intelligence, it dumbs everything down and reduces information to a common-denominator baseline. (In the future the real difference between the "haves" and the "have nots" will be those who can understand and deal with information complexity and those who can't. Creating information complexity is the only thing humans are naturally good at, it's our biological imperative.)
- kris_lujanovic 3y ago
  AI is not here to dumb down anything but automate tasks and find and reveal complexity esp hidden complexity. Humans are limited by their short and long term memory and bandwidth esp in modern digital environment. I would still rate opportunity gap as higher priority over understanding complexity. its still funny when people try to be digital nomads just to discover geopolitics or time zones
- throwbmw 3y ago
  Any further reading recommendations on this topic?
marcus_holmes 3y ago
I would prefer a "peasants and longbows" analogy; any country that trained its people in AI would become super effective but would have to keep its peasants very happy.
hollerith 3y ago
>I think the most likely way to achieve “AI alignment” is the peasants with crossbows. Everyone having a powerful AI at their fingertips provides a balancing factor that is only overcome if those tools are in the hands of very few. I think neither elite actors like Sam Altman nor the peasants will manage to get their AI to do what they want it to do once the AI becomes knowledgeable enough and good enough at planning to make plans resistant to determined human opposition. I.e., I think inequality is a distraction from the main problem with frontier AI research, namely, we are not going to figure out how to aim one such that it stays aimed after its cognitive capabilities have become dangerous.

hackernewds 3y ago

If you replace "peasants with arrows", with "good guy with a gun" I think the whole argument rings a different tone..

iudqnolq 3y ago
It's less "good guy with a gun" and more "some random country with a small to medium army". Access to violence doesn't democratize on the individual level, but as the parent comment alludes to weapons changes allowed peasant armies to defeat trained knights, changing the structure of society. You personally might not be able to exploit open AI fully, but there are lots of countries and orgs that could. Remember millions of dollars is loose change in the defense context.

coffeebeqn 3y ago

A finite resource fairly allocated to all? Which civilization are we talking about again? Certainly not humans. Also a billion or trillion or infinite ants can’t outsmart a single human. So unless some hard cap is enforced (how?) then mini AIs will probably not do anything

czl 3y ago
> Also a billion or trillion or infinite ants can’t outsmart a single human. Cells are simpler than ants so can trillion or infinite cells outsmart a single human? Why not? How many cells is a single human?
- kig 3y ago
  Around 37 trillion cells, of which 25 trillion are red blood cells. Plus a 100 trillion gut bacteria. The brain is about 0.1 trillion cells, of which 0.02 trillion cells are the cerebral cortex, which is occasionally consulted by the rest of the brain on matters such as what words to enter into this textarea.

trident5000 3y ago

We definitely got a small taste of what people do with their time when they are not at work during Covid...and it wasnt pretty.

Cthulhu_ 3y ago
Not pretty for who? Can you explain?
- trident5000 3y ago
  The summer riots that were fueled by people not needing to be anywhere.

TheRoque 3y ago

Well somehow I trust google more on having my data, than anyone in the world. I prefer the knowledge to be in a known identifiable entity that can be accounted for, than run in the wild and be used by anyone with bad intention (once again, regarding my data). So this is not about data, but AI, but I think it still applies.

hackernewds 3y ago
this is how I schedule my daily life. it is inevitable that I use software services. I'd like to keep them limited to one megalith, and that is the Google ecosystem. at least their biz model is not predatory, where using my individual data exceeds the risk of doing so.
zelphirkalt 3y ago
> I prefer the knowledge to be in a known identifiable entity that can be accounted for […] Google can be accounted for? As in "I know the entity that has the data." or "can be held accountable" or "will be held accountable"?

otabdeveloper4 3y ago

> powerful AI AI is not powerful. AI doesn't increase information complexity or create new information. It actually does the opposite - it averages out the complexity peaks. In effect, AI isn't intelligence, it dumbs everything down and reduces information to a common-denominator baseline. (In the future the real difference between the "haves" and the "have nots" will be those who can understand and deal with information complexity and those who can't. Creating information complexity is the only thing humans are naturally good at, it's our biological imperative.)

kris_lujanovic 3y ago
AI is not here to dumb down anything but automate tasks and find and reveal complexity esp hidden complexity. Humans are limited by their short and long term memory and bandwidth esp in modern digital environment. I would still rate opportunity gap as higher priority over understanding complexity. its still funny when people try to be digital nomads just to discover geopolitics or time zones
throwbmw 3y ago
Any further reading recommendations on this topic?

marcus_holmes 3y ago

I would prefer a "peasants and longbows" analogy; any country that trained its people in AI would become super effective but would have to keep its peasants very happy.

hollerith 3y ago

>I think the most likely way to achieve “AI alignment” is the peasants with crossbows. Everyone having a powerful AI at their fingertips provides a balancing factor that is only overcome if those tools are in the hands of very few. I think neither elite actors like Sam Altman nor the peasants will manage to get their AI to do what they want it to do once the AI becomes knowledgeable enough and good enough at planning to make plans resistant to determined human opposition. I.e., I think inequality is a distraction from the main problem with frontier AI research, namely, we are not going to figure out how to aim one such that it stays aimed after its cognitive capabilities have become dangerous.