Rambling Ideas of a Dumpling Maker

on

Automation, robotics and AI are projected to take over 1 in 3 jobs within 20 years (read this article for a full list of jobs, each with probability attached to it). Many low skilled repetitive work will be replaced by a more efficient, cost effective robots and the ensuing unemployment crisis is a serious issue facing developed countries. The main solution (and in many minds the only solution) is to provide Universal Basic Income – or UBI – an unconditional payment of living wage to all adult citizens. First proposed in the middle of 19th century, it seems that automation may finally make this idea economically viable.

UBI has the potential to eradicate poverty once and for all, free individuals from anxieties of the daily grind and provide the freedom to choose what they really want to do. If the basic income covers our rent, bills, food, child care and other costs, our priorities would inevitably change dramatically. We may choose to study further, create art, volunteer or take part in projects that have higher risk factors like starting a new business. We may choose to invest the free time in creating and sustaining a stable environment in which to raise our children, or choose a profession purely from the prospective of how much fulfilment we gain from the experience.

There are arguments against UBI, such as the possibility of people becoming lazy. I would assume that there will be a small percentage of people who will choose to do absolutely nothing, but the percentage would probably be quite similar to what it is currently. There will likely be a reduction in work hours across the whole spectrum of professions, but if people are well rested, enthusiastic about work, have a healthy work life balance, free of stress and stress related illnesses, the productivity will go up. I’d like to think that we’re an incredibly curious species incapable of sitting still for too long and boredom will eventually compel most people to seek engaging and rewarding activities.

The main question is how do we fund it. It will replace the current welfare system, which could partially fund the UBI, but if it is going to be a national scale unconditional payment, a large portion of it will have to be sourced through other forms of taxation. Who do we tax? and how do we make sure that the tax will be paid? Robot tax is an obvious one. Perhaps measured in how much human work hours each robot displaces and taxed accordingly. Bill Gates, Elon Musk and Stephen Hawking were proponents of this taxation system. Data tax is also an interesting option. As data becomes more and more valuable, data giants such as Facebook and Google could fund parts of the UBI in return for use of customer’s data.

But if gigantic multinational corporations are evading tax payments by shifting it’s assets to offshore tax havens right now, it’s hard to believe that they will pay out a far bigger tax, internationally, in the coming decades. I can’t think of a way to enforce taxation on global companies, when we are painfully limited by our national borders. Implementing strict laws will just see these companies flee to another country.

Perhaps, just perhaps, we may see a first glimpse at the possibility of a global governance. A single entity that rules over all humanity on Earth. From impacts of climate change to water scarcity, poverty and  soil degradation and explosive increase in migration, we have a plethora of issues that cannot be resolved unless we act as one overriding entity with a long term vision.

This idea rightfully instils fear in anyone with a basic understanding of history. It’s littered with examples of human corruption and misuse of power, even at a puny scale of one nation, often with devastating consequences. How then, could we ensure a rightful, empathetic planetary system devoid of corruption, protected from skewing into an authoritarian regime and act with the long term well being of all sentient beings in mind?

I think Artificial Intelligence will hold a key component in realising this system. A global governance system has to have a safeguard against human errors, misjudgement and pure evil. Democratic governance with short term-time was a way to ensure that, and it’s worked relatively well in the 20th century. But it’s completely insufficient in a world that requires fast decision making based on long term visions, with hundreds of self interested nations involved.

What if the AI governance was guaranteed to be free of human errors? Unaffected by wants, needs and feelings that often clouds the judgement of us mere beings? What if it was able to compute enormous numbers of possible scenarios and work out the most efficient way to ensure the overall well being of every sentient being? What if it was able to make key decisions, map a detailed plan spanning hundreds of years and put it into action based on the survival and well being of our species for the next 10,000 years? What would that plan look like? Would it look similar to permaculture? Quite likely.

The next question that presents itself along this line of thought is, how do we ensure that an AI capable of this much power, far superior in intelligence than us, do not go rogue, remain benevolent and unhack-able? How can we possibly create defences against it from turning malevolent when the subject has super-intelligence capable of easily outwitting us? What stops the AI from manipulating and changing it’s own algorithms or designing and building hyper-intelligent AI’s that are beyond our scope of understanding, let alone control, with its own set of principles with complete disregard for human interests?

These are still very much open questions but I’d imagine some form of distributed AIs with a democratic governance model might work as an intermediary step. They could vote for key decisions at a speed of light, coming to some form of consensus with each action, thus eliminating the chances of one rogue AI taking control. It will inherently be protected from hacking as no central power exists and a single hacked AI cannot rule over the majority.

One thing’s for sure, we’re alive at an incredible time in history. At the door step of a new era that promises to be full of volatility and unimaginable possibilities. Let’s hope we live to see it flourish as something that benefits all living things.

5 Comments Add yours

  1. Ben Roth says:

    i’m with you. it almost seems like the only option really.
    let’s come up with positive future vision instead of repeating the same old dystopian narrative. maybe AI can really be a benevolent compassionate guide, sort of a machine buddha.

    “If you stand back you can see we have the money, the power, the scientific know-how, the love and the community to produce a kind of human paradise.”

    — Terence McKenna

    Liked by 1 person

  2. freisprung says:

    i’m with you. it almost seems like the only option really.
    let’s come up with positive future vision instead of repeating the same old dystopian narrative. maybe AI can really be a benevolent compassionate guide, sort of a machine buddha. i once read a great article about this notion. maybe it was this one here, not sure:

    Click to access 2011-hughes-selflessrobots.pdf

    “If you stand back you can see we have the money, the power, the scientific know-how, the love and the community to produce a kind of human paradise.”

    — Terence McKenna

    Like

    1. cshimasaki says:

      Thanks for that interesting article Ben. Grounding in embodied sense data is an interesting idea I’ve never thought about in AI development. But I totally understand and relate to the notion that compassion requires being in touch with one’s own suffering and that a machine can be taught moral virtues upon it. I’m still unconvinced, though, if what we know of our bio-chemical algorithms will translate into digital algorithms in the same way. Perhaps the first generations of AIs maybe based on our consciousness development, but once those AIs begin to code the next generation AIs I would assume that they will design it in ways that we can’t even imagine.

      Like

Leave a reply to cshimasaki Cancel reply