4 ms·

100% the blockchain isn't needed for much, but longevity and AI are actually really useful to humans

abraxas 4y ago
The whole blockchain ecosystem is basically an instantiation of Bostrom's Paperclip Maximizer running on a hybrid human/machine topology.
- willmorrison 4y ago
  Can you explain more what you mean?
  - jack_pp 4y ago
    I think he's saying it took a life of its own and its incentives are destructive to mankind. I disagree but I can see how you could make the argument, supposing you think bitcoin has no use / value and that smart contracts are just a fad. But maximalists have their own arguments regarding the use of blockchain and while a lot of the rhetoric isn't believable I do think blockchain tech will be +EV in the future even if most don't believe it already is.
  - abraxas 4y ago
    Basically it is a purposeless algorithm that saps resources for useless computations, that is only interesting insofar as it can sometimes perform sophisticated optimizations albeit in pursuit of a trivially useless goal. Zoom out and think of the entire blockchain algorithm at a level above the incentives of a singular agent. It's very much like having a superhuman Paperclip maximizer that has a invented amazing techniques of say, iron ore extraction but this localized advance is squandered in pursuit of an absurd goal of creating the maximum possible volume of paperclips. Of course there is not a single agent on earth who pursues bitcoin mining with the goal of wasting energy and compute resources but it is the aggregate end result. Bostrom similarly argues the the objective function of a complex system such as an AGI may drift entirely from its original intent leading it to optimize for things we absolutely do not desire and this malignant drift is an emergent property. The similarities to blockchain are uncanny.
Heronymus_Anon 4y ago
Mildly disagree. Problem with blockchain is that masses are focused on financial uses, whereas I think lots of the potential is in liquid tools for self-organizing, something ~ like zeroknowledge privacy protecting data mining. So essentially tools that are not really meant for big business or fast money schemes, but for social transformation useful for humans (and AIs). It may raise the quality not the quantity of your years. (..of course raising both would be appreciated ; ))
- stevofolife 4y ago
  Totally agree. Do you know any companies or projects in this space? Or in federated learning / differential privacy using blockchain? Tons of use cases can be enabled by them.
- coffeebeqn 4y ago
  I don’t think there will be zero actual crypto use cases. I just think dumping billions more in VC money to that stack is a bad idea at this point. And that the current “crypto industry” Is pretty much a scam/gambling. Maybe some kind of escrow, maybe privacy.. stuff. Whenever I ask people what the real crypto projects are they still give a very abstract handwavy buzzword-heavy answer like you just did
- mountainriver 4y ago
  I do think there is potential in blockchain, I just don't think its actually needed in most situations. Zero trust is an interesting tool for a limited set of use cases, much of what I see being built on the blockchain could just be a web service with a legal guarantee, which would make it more efficient and cheaper, the majority of people would care less about the implementation.

The whole blockchain ecosystem is basically an instantiation of Bostrom's Paperclip Maximizer running on a hybrid human/machine topology.

willmorrison 4y ago
Can you explain more what you mean?
- jack_pp 4y ago
  I think he's saying it took a life of its own and its incentives are destructive to mankind. I disagree but I can see how you could make the argument, supposing you think bitcoin has no use / value and that smart contracts are just a fad. But maximalists have their own arguments regarding the use of blockchain and while a lot of the rhetoric isn't believable I do think blockchain tech will be +EV in the future even if most don't believe it already is.
- abraxas 4y ago
  Basically it is a purposeless algorithm that saps resources for useless computations, that is only interesting insofar as it can sometimes perform sophisticated optimizations albeit in pursuit of a trivially useless goal. Zoom out and think of the entire blockchain algorithm at a level above the incentives of a singular agent. It's very much like having a superhuman Paperclip maximizer that has a invented amazing techniques of say, iron ore extraction but this localized advance is squandered in pursuit of an absurd goal of creating the maximum possible volume of paperclips. Of course there is not a single agent on earth who pursues bitcoin mining with the goal of wasting energy and compute resources but it is the aggregate end result. Bostrom similarly argues the the objective function of a complex system such as an AGI may drift entirely from its original intent leading it to optimize for things we absolutely do not desire and this malignant drift is an emergent property. The similarities to blockchain are uncanny.

Heronymus_Anon 4y ago

Mildly disagree. Problem with blockchain is that masses are focused on financial uses, whereas I think lots of the potential is in liquid tools for self-organizing, something ~ like zeroknowledge privacy protecting data mining. So essentially tools that are not really meant for big business or fast money schemes, but for social transformation useful for humans (and AIs). It may raise the quality not the quantity of your years. (..of course raising both would be appreciated ; ))

stevofolife 4y ago
Totally agree. Do you know any companies or projects in this space? Or in federated learning / differential privacy using blockchain? Tons of use cases can be enabled by them.
coffeebeqn 4y ago
I don’t think there will be zero actual crypto use cases. I just think dumping billions more in VC money to that stack is a bad idea at this point. And that the current “crypto industry” Is pretty much a scam/gambling. Maybe some kind of escrow, maybe privacy.. stuff. Whenever I ask people what the real crypto projects are they still give a very abstract handwavy buzzword-heavy answer like you just did
mountainriver 4y ago
I do think there is potential in blockchain, I just don't think its actually needed in most situations. Zero trust is an interesting tool for a limited set of use cases, much of what I see being built on the blockchain could just be a web service with a legal guarantee, which would make it more efficient and cheaper, the majority of people would care less about the implementation.