The people's AI on a blockchain consensus?
I really hope that steemians who have the technological understanding of how to build a blockchain will express their opinion on the ideas presented here. Because I have just a basic understanding of the technology and its applications and would love to hear relevant feedback.
The context of intelligence
There are things in the event horizon of humankind that are inevitable. For as long as we can remember as a collective, we have built on the understanding that came before us. With time, things started to become clearer and clearer, man being able to understand the elements surrounding him started to use them to his favor. And every so often, an understanding of something was so powerful that it had the possibility to reshape human interaction and evolution. Starting with understanding fire, the barter system, agriculture, money, the printing press, the internet and much more. These are technologies build on the understanding of different principles.
Some of these revolutionary technologies come silently, like bitcoin. Others came with a massive library of stories on how it can and will go wrong. In not a historian, but I believe that nothing in our history ever had the possibility of reshaping society to the degree artificial intelligence has (if you take over the world, I apologize for not writing your name with capital letters).
Intelligence is the real reason we were able to get to this point. Intelligence is getting us into space, and that same intelligence is going into the atom. There is nothing more versatile, except for the automation of intelligence.
So instead of having a team that needs to understand a problem and arrives at a solution, you now have an automated intelligence. Which has access to all the public digital information in the world and is able to understand problems from every point of view it has access to. It can assimilate the information of any science needed to analyze the issue assigned and it can think from every point of view simultaneously. This would not be a single intelligence, but multiple ones that have been developed differently and put to work, The capacity of solution generation is unimaginable, innovation would happen probably every second.
But we need to be realistic and understand that intelligence also created monstrous things. From bombs to capitalism, we have countless examples of innovation that only helped a few and concentrated power. Such is also true with artificial intelligence, it will be used by people that only have their interest in mind.
The battle of AI's
Artificial intelligence development is not happening in just one place. There are many actors in this field, with different values and goals. It would be naive of us to think that artificial intelligence is not being developed for unpleasant purposes. But what will happen when two AI's have opposing commands and in order to resolve the problem, they need to face each other? Would the "best" AI win?
The world we live in has different power structures, which use any mechanism at hand to hold and extend the power they have. The development of artificial intelligence will give them another tool to use in furthering their agenda. I would assume that some power structures and different alliances between them, will develop their AI, in the form best serving them.
In this whole game of power, there is only one structure of interest to me: the people.
The people's AI
I believe that magical things can be done in the world is by developing a collective AI. An artificial intelligence that is tailored to real human needs and is oriented towards a sustainable future. A benevolent AI as @wildsunlove calls it in one of his posts.
The key difference between the people's power structure and the other structures is that we are distributed and decentralized. Which is really good. Because in a centralized environment innovation is limited by guidelines and objectives that were predetermined. But, a decentralized and distributed system can generate solutions from everywhere. There is no one place of research, a team or a team of teams. You have a global connection and almost anyone that wants to bring innovation is free to do so. And those who are not able to connect, are restrained by political reasons. Not because the internet does not let them in.
So the people's AI would be a global collaborative project that would involve innovation coming from any part of the planet. All that information needs to be structured in such a way that would grant safe, free and open access to all that need it.
The global AI blockchain
Decentralised innovation is fantastic, but its very hard to structure in a coherent and manageable way. Because new ideas and technologies are coming from all around, there needs to be a way in which their value is assessed. If the goal is to make a better world, we need to integrate the innovations that work towards this goal. Given the stake at play, there needs to be a system that cannot be controlled and is collectively managed.
Up until now, blockchain technology has seemed to be the best way to agree upon and store information. Of course, other technologies need to be integrated in order to make this function. There needs to be different levels and players involved. This is what I have envisioned so far:
All the persons and groups involved in this effort would need to play two roles exactly as with steemit. One would be the of the generator of innovation and the other of a curator. The people that are working on developing AI technology need to be part of a collective effort of judging the potential and benefits of innovations submitted to the blockchain. So there needs to be a sort of consensus between the developers. After arriving at a common view of what a benevolent AI is and what are the properties of this AI.
This is only a theoretical structure that needs to be discussed and built upon:
1. The "nodes"
In order to become a "node" in the network, there needs to be some sort of "proof of intention" that only allows participation to those who have deemed trustworthy. In this particular case, I believe that we need a certain threshold of nodes before the blockchain starts running, in order to assure decentralization.
2. The submission process
The submission of new ideas to the blockchain should only be possible for participating nodes. And in order for those ideas to be accepted on the blockchain, a consensus between the other nodes needs to occur.
3. The "miners"
The blockchain AI will need a lot of computational power in order to run. This is where the other people, that are not innovators, come in. People can contribute their processing power to different aspects of this AI. I don't know if there can be a monetary incentive. But because different people are impacted by different problems, the solution to that problem can be the incentive. Because of this, people should be able to choose where to contribute their processing power.
4. The users.
Access to this AI solution generating entity should be free, open and borderless, just like the internet. People that need to find the solutions to their problems would need to have access to it in order to find solutions. But because all people cannot be trusted, it is imperative that the three other groups presented above need to execute an AI that can analyze requests and only work on those considered to be constructive.
Most probably the architecture of this blockchain AI will be different than was is conceived here. But the need for an open and global network that works for the betterment of humankind is ever growing! Let's build it together.
I eagerly await your opinions and ideas on what is presented here.
Other posts:
https://steemit.com/internet/@lishu/introduction-post-the-future-is-ours-let-s-build-it-together
https://steemit.com/blockchain/@lishu/can-blockchain-be-used-to-write-history
Interesting article. You make some good point & raise some interesting questions. I look forward to reading the replies.
Thanks, I really hope this will spark a conversation...
It is refreshing @lishu to find a perspective like yours and a worthy presentation that up-voting deserves!
In my view nature beat us in that AI already exists, in the form of your minds and mine; that the internet serves as a neural system of sorts, giving us simultaneous access to a platform like this as a communal blackboard on which to share the products of our abilities at large.
Complete emancipation of tech based AI to me is a nightmarish thought. Where for us fulfillment might have the form of lasting happiness derived from perfecting a secure niche within which we can productively thrive in peace for as long as we can imagine, what would it be from 'their' point of view? What would the motive for continued existence for such a being be? Will it possess empathy for any creatures other than itself? Or will we have created the perfect psychopath?
Should we succeed, as I think you imply, in building in prohibitive mechanisms against bad behavior toward us, will it prevent 'them' from removing those somehow in the end?
Sorry, that it took so much time for a reply. I left steemit alone for a while.
I see human level AI happen, but I have my doubts regarding an emerging consciousness. What I believe is more possible, is that the AI will work as an additional layer of processing for individual human brains. So if bad thing come, they come from people's intentions, rather than from an independently acting AI.
But, this is still not a good scenario, because we still need to limit the damage done by bad actors. Hopefully we find a way.
Carrying a DEMP blaster in your belt might not be a bad idea though. Criminals will have them - Directional ElectroMagnetic Pulse pumps that can sense the operating frequency of malignant AI and zap it's circuits with a resonant pulse!!
But coming back to your post - maybe benevolently designed vigilante AI constantly scanning for bad or suspicious behavior of AI setups and their potential creators all over could be a good starter for keeping the playing field clean.