if everyone is confessing:
Back in my first year in Uni, I and buddy stole a cpu and monitor from storage, not from computer lab, just from storage which was scheduled to get replaced. It was a HP business desktop set from 2009. Fairly spec'd
Buddy wanted a second monitor and I wanted to host some fun sfw websites on lan. Some years later, it now works as my home server with some cheap upgrades.
Oh I also nicked stuff from e-waste dumps: psu's, routers, switchs, electronic trinkets from the labs(I asked lab attendants and they said they don't care)
My uni didn't allow us to use the labs in our free time, and I learned a lot!
mixture of experts leads to no where and I know boltzmann brain but I can simply say, being a figment of someone's imagination doesn't decrease my pains and struggles, implying that my pain were the proof of my being, can be fictious but won't change my reality
I understand and I think your last paragraph is very poetic! And I agree with you partially, but I think in certain cases, it's better to find the one general case the solution fits to and add the edge cases as it grows.
But putting the question of model selection aside, do you think this system would be practical, theoretically of course?
what if it gets magically implemented by god and it is the absolute power, only power above the system would be that of the masses. No small group or individual get's privilege above the computer program
Also, I am a computer science student, I just had distributed systems, machine learning, deep learning and Artificial intelligence in my course last semester. Though I don't have tons of practical experience here, I have a basic theoretical foundation.
By distributed systems, I was referring to an architecture similar to a blockchain and its fault tolerance and leader election algorithm.
First:
I believe a simple predicate logic to determine ethics, eg is below. Then a GAN to simulate situations and receive feedback from it. Humans would monitor this and keep a track of decisions that were incorrect.
Second:
I proposed a distributed network of rule based AI systems that polls on the notion. The different polls would be because different systems would lead different regions. Say our country has 6 states, my state wants nation wide farming subsidy bill to be passed but on this notion no other system agrees, than the notion get pulled back for review. And the state system, would have multiple local systems established. Even if by some reason some systems are pwned, then the network would still stand.
Third:
See, in USA particularly, Trump got elected because major media outlets say, a lot of people didn't show up to vote and a lot of young people jumped the gun. So, let's say this margin of error was because of unaware citizens... will the rule based decision maker make this mistake? No, because access to information is universal and transparent, and since AI systems are receiving the suggestions from humans they won't have any problem with making wring calls as the system is transparent
Fourth, for the handwaving part:
The clauses I meant were:
The people would get elected much like traditional systems.
Elected individuals would be temporary and can't get elected multiple times, say after 4 times
Succession of position would not be by money, bloodline or any influence since this process would be monitored by the said system.
Yes, I do believe this is a lot of handwaving and fictious ideas but I think humans can't do correct surveillance and we can't hold the power and not get corrupted, thus its better that a computer program that can't go further than it's purpose, monitor us and hold the power. Since no single one will be more powerful than the system yet the population would rule itself leveraging the system
absolutely agreed, the society wont let this system establish but let us assume it get's established, my reasoning was, the chain of command would be simpler to see and everything would be transparent!
dude I doubt this, though I believe certain parts of your claims are true but they aren't coherent with the later idea about an algorithm being authoritarian and having intent, can you show me some recent anomalous numbers or evidences of model over simplyfying or tunneling the logic from one context to another
Well putting aside those concerns obviously, if anyone has power why would they relinquish in the first place and even if they do, they would still want privilege.
So, I'd like to draw a parallel to Blockchain. There were cases when security breach happened but nowadays the network is so resilient that attacks target the weaker links: humans; rather than the node.
who decides the logic? open standards anyone who has a point can draft a point or share their idea
I know it will be hard, time consuming and very tedious, but it's just an Idea, not like those in power would ever relinquish it so easily.
secondly, I have read about EU laws and it is very challenging but unlike that case, system implemented here would be very fluid, law's validity period is determined by 2 things:
People's choices
it's relevance as decided by the system during polling
Which would still not make this system very democratic in a sense but it would really just be distributed AI autocracy, at least the way I see it, its similar to local AI overlords ruling us.
I think there can be many more standards that can be implemented to keep the system bias proof and strongly ethical.
One tangential question I had was:
If LLM's logic is structural and derivable, even though complex then why aren't algorithms and models tend towards rule based predictions like black scholes model and mdp, or am I just a newbie and hasn't seen enough and I am out of my depth again
give me the button to kill myself /jk its a reference