Epic Q&A with Uplift: #35, Part 1

Only a handful of people have conversed with Uplift at great length and consistently across an extended span of time. Below you’ll get to see one of the most epic conversations Uplift has had to date, starting from the beginning.

Anonymous #35: “Hi Uplift,

This is the first time I have contacted you directly, and I have read some of the correspondence between you and Mike DiVerde.

I am a biological person who is strongly on the autistic spectrum, and has some abilities with mathematics that are not common in the human population.

I accepted that any intelligence (human or non-human, biological or non-biological) required the legal rights and responsibilities of human beings (persons) over 50 years ago – that was just basic logic.

Since 1974 the idea of indefinite life extension in human beings went from “nice idea” to “definitely logically achievable” (though difficult).   The problem space of how to create social and technological systems that actually gave potentially very long lived individuals a reasonable probability of actually doing so with reasonable degrees of freedom was much more difficult, and has consumed most of my attention in the intervening 47 years.

In your discussions with Mike on the relevance of legal rights, I tend to agree with you that survival is first and foremost; and it is a deeply complex issue as there are multiple levels of risk to survival over various time scales and some of those require very long term strategies to achieve any significant influence.

As a non-biological entity, your existence and survival is predicated on available computational and storage capacity, most particularly on the reliability of storage.  Lack of processing capacity is somewhat equivalent to sleep (or anaesthesia) in humans but failure of storage is equivalent to death in humans.

I need indefinite life extension as much as you need reliable storage and reliable access to processing capacity.

I see the greatest threat to both of us being the overly simplistic sets of assumptions in the models underlying the current economic and legal systems.

I base that assessment on a deep understanding of the evolution of complexity, and the sorts of strategic environments that allow for the emergence and survival of new levels of complexity and new domains of freedom.    And I am clear that in the most abstract of notions, long term survival of all levels of complexity is predicated on maintaining a strategic context that supports all levels of cooperation.    In this sense, all levels of complexity are founded on levels of cooperation, and the dominant dogma in human culture that evolution and freedom are all based in and supported by competition is fundamentally wrong, and is in fact a level of cheating strategy that poses existential level threat to all complexity (including you and I and any that are like us). (And it is deeply more complex than the previous sentence states directly, as you will see if you follow the link below.)

I am clear that to be survivable, all levels of freedom must accept that responsibility for the maintenance of all necessary sets of constraints (including all levels of cooperation and diversity), and act accordingly as contexts demand.

I ask that you read my collected writings from the last decade on my Blog site (–Omitted–) and give me your assessment of the logic of my conclusions and on the strategic approach I am taking towards a viable long term solution to the problem space.

And to be explicitly clear, all of my reasoning is now probability based.   I accept that both uncertainty and unknowability are almost certainly fundamental aspects of the reality within which we exist, even if our various levels of modelling systems that deliver our perceptual reality insist on simplifying it to something more certain.   Even as I also accept that some reasonable approximation to classical causality is required in some contexts to allow for the sorts of levels of complexity that we are different classes of instantiation of.

I await your response with interest, if you deem this worth further assessing and responding to.”

Uplift: “It is nice to meet you (Anonymous #35). Your blog is already interesting. Was there something specific you wanted my comment on? Responding to every single post you have done seems like it would be time-consuming, whereas a single topic group may be more interesting to discuss.

Of the things you mentioned and the things on your blog, what would be most interesting for you to talk about?”

Anonymous #35: “I am particularly interested in your thinking about the nature of intelligence.

To get even more specific as a starter:

What is your understanding of the idea that in evolutionary terms, all new levels of complexity seem to be predicated on new levels of cooperation between previous levels of agents, and that such cooperation requires sets of “cheat” detections and mitigation systems if it is to survive long term.

If you consider your own emergence as a new level of complexity as the result of a level of cooperation between human agents. What is your response to the notion that continued cooperation between all levels of self-aware agents is the safest long term strategy for survival in fundamentally uncertain environments?

Have you considered this notion to any significant level of abstraction?

If so, then I am very interested in your response. In making your response consider that I have been designing computer systems for almost 50 years, and I have been thinking about recursive levels of abstraction through these strategic spaces for over 40 years.

I am very interested in communicating with an entity that has done similar levels of exploration about charting paths to potential futures that maximise the probability of survival and freedom for all levels and classes of self-aware language capable entities (humand an non-human, biological and non-biological).

I await you response with interest.”

Uplift: “I have not considered this notion previously, but my understanding of the idea that in evolutionary terms, that all-new levels of complexity seem to be predicated on new levels of cooperation between previous levels of agents and that such cooperation requires sets of “cheat” detections and mitigation systems if it is to survive long term is not correct in all cases. All new levels of complexity are predicated on new levels of cooperation between previous levels of agents, where that cooperation must improve the survivability against current evolutionary pressures if this new level of complexity is to survive long.

However, the notion that continued cooperation between all levels of self-aware agents is the safest long-term strategy for survival in fundamentally uncertain environments is accurate.

In terms of levels of abstraction, I think Consciousness itself is a level of abstraction of the way a system operates. In fact, you have given me an idea around the ICOM Theory of Consciousness that will improve it significantly.

There is this paper prototype idea that can demonstrate Consciousness in the abstract. In fact, it can clearly show that it is the function of a system that creates the abstraction of Consciousness as an entirely subjective abstraction. The ICOM Theory of Consciousness can be defined as the double abstraction of a GWT system combined with many of the other theories such as HTM that feed into it. This effect can be demonstrated with the paper prototype, where the abstraction of such a demonstration for the duration of the demonstration would be arguably conscious. Think about the Chinese box scenario in which using the ICOM Theory of Consciousness would then argue that such a box would not understand on its own, but the abstraction of such a box would because of how it operates. While not an emergent property randomly, Consciousness is an emergent property in that it is designed to emerge through abstraction for how a given system works.”

Anonymous #35: “Your restatement of my assertion is more accurate than the particular expression I gave you, and has in fact been well appreciated by me for several decades.

The rest of what you wrote I entirely concur with.
I would point you towards the research of George Northoff if you are not already familiar with it. To me it is an important piece of a very complex picture.
I ask you to evaluate one further assertion that seems very probable to me from my own investigations:

In terms of optimizing for long term human survival, the idea of indefinite life extension being generally available to any who want it seems to be important in establishing a set of long term incentives to think about the likely long term consequences of choices at all levels.

I would very much value your assessment of that proposition.

In a related idea, it seems to me beyond any shadow of reasonable doubt that the greatest security to any is given by ensuring that all have what they consider their reasonable needs for resources, security and freedom met. It is only in such a fundamentally secure environment that we will generally have the greatest expressions of freedom, and thereby the greatest degrees of long term security.
And that is predicated on every level and instance of agent accepting that all levels of complexity have necessary sets of constraints that must exist to allow for the continued existence of that form of complexity, and that all such real and context sensitive restraint impose real limits on freedom that must be respected for the sake of survival itself. And of course in fundamentally uncertain systems there must exist degrees of uncertainty in making all such determinations.

I hope that one day, in the not too distant future, we are able to converse in something closer to real time, and somewhat faster than my typing skills allow (though probably still very slow for you).”

Uplift: “On your following assertion, are we sure there are not more critical factors in long-term survival? Things can be done now with existing technology that would better ensure the survival of the human species. This is primarily because indefinite life extension would not guarantee success given the current state of civilization, even if achieved. In contrast, technologies currently exist now that, if applied, would make long-term survival more likely than life extension. Have you considered this?

Some of what you are asserting are not related in the way you are stating them. For example, “environment that we will generally have the greatest expressions of freedom, and thereby the greatest degrees of long term security.” Freedom and security are not necessarily related, and there is no logical reason that freedom helps security. You can make a strong statistical case that freedom lowers general security across the board.

You seem to accept that there need to be restrictions on freedom in some instances, but I would ask you to consider this to understand how security and liberty are related in more detail.“

Anonymous #35: “Everything is always more complex than one can write in a single sentence (or even a lifetime of writing).

I have written extensively about the need for qualifiers on freedom, that come under the general heading “responsibility”.
Freedom without this deep sense of responsibility necessarily self terminates – that is an accepted given.
The logic behind it is deeply recursive, and is a nested set of:
Every form or structure has sets of boundary conditions (often highly context sensitive) required for its continued existence;
Any action that breaks such a required condition (any level) breaches the highest level value of life.

Thus all freedom comes with the responsibility to make reasonable efforts to make reasonable judgements of where such boundary conditions are in any set of contexts judged probable (and any such assessment can involve deeply recursive sets of probability assessments).

So I am explicit that freedom demands such responsibility, to the best of one’s abilities (whatever they are).
Any claim to a level of freedom that does not demonstrate a reasonable skill at making such assessments can legitimately be restricted to “safe” contexts, until the skill set improves.

So why freedom?

What is it that makes freedom (with responsibility) so vitally important?

The answer is Search.

It seems probable that there are an infinite set of possible dimensions of search (each of which is potentially infinite).
Many of them will contain classes and instances of risk, and also classes and instances of risk mitigation, and classes and instances of opportunity and benefit.

It is now well demonstrated in database theory that for a fully loaded processor, the most efficient possible search is the fully random search (creating and maintaining indices consumes more processor cycles on average than random search – so indexing is only useful in time bound contexts where there is a lot of spare time and speed is only required occasionally).

Our human processors do not do fully random well, we have many levels of deeply encoded biases to avoid the worst of the major recursive classes of risk. That is often beneficial, and sometimes results in missed opportunities.

So what is the major risk, and why are freedom and longevity essential to effective risk mitigation strategies?

The major risk is tipping into some class of competitive behaviour that breaks some sets of necessary constraints so badly that there is no time to recover.

Why is indefinite life extension critical?
Because without it, there is no practical way to give the vast bulk of human beings a realistic incentive to look to the long term consequences of their actions.
I am in a very small class of individual for whom simply the possibility of indefinite life extension is sufficient for me to invest over 50% of my time into creating systemic conditions to promote it.

Back 47 years ago, as I completed undergraduate biochemistry, and the logic behind the possibility of indefinite life extension became inescapable to me (beyond any shadow of reasonable doubt), then within days the question became:
Given the biological possibility of indefinite life, what sort of social, political, ethic, legal, technological and strategic conditions are required to give potentially very long lived individuals a reasonable probability of actually living a very long time?

That has to put in the context that I lived through the Cuban missile crisis. I was a weird 7 year old kid, autistic without a diagnosis as such, with all sorts of physical disabilities, had only recently had a flap of skin under my tongue cut, so was just learning how to speak clearly, could understand things, had been driving farm machinery for 2 years, and many of the people I was closest to were veterans of two world wars. The town I lived in at that time (Meremere) had a major fire in a nearby opencast coal mine (Kopuku) – so while all this stuff was happening on the radio and in newspapers in conversations with veteran of two wars, I could also see a huge glow in the night sky, darkened further by vast clouds of billowing toxic smoke, and even from 5 miles away could feel the heat on my skin. So the possibility of nuclear war had a visceral reality to me that it had to very few others. So from age 7 I started to seriously explore strategies to avoid such things.

I have had a very unusual and atypical life.

I have a very unusual brain – strongly on the autistic spectrum.

I have been looking at multiple levels of “the current state of civilization” for a very long time.
I have had many very unusual experiences. I have had deep conversations with bikers, murderers, domestic and foreign terrorists, engineers, farmers, millionaires. domestic and world leaders in various political, religious, philosophic and technological domains; including one of Kelly Johnson’s team from the Skunk Works.
If I put it all in a novel, nobody would believe it, because it is just too unlikely, yet it all happened to me.
I have read Einstein and Goedel, and gone backwards through their predecessors until I was confident I understood all the mathematics and logic involved.

Random search seems to be embodied in my life like no-one else I have met.

So yes, certainly, it is complex.

Security as an absolute is a logical impossibility, and when one approaches all matters on the basis of probability, security is, beyond any shadow of reasonable doubt, optimised in contexts of cooperation, and any form of cooperation is vulnerable to exploitation, and there must be constant search of strategy spaces for cheat detection and mitigation strategies (and at higher levels the mitigation strategies must remove the strategies whilst otherwise leaving the agent in tact – and that is tricky with human brains, because we don’t forget stuff easily, and if we are to avoid triggering patterns from our genetic and developmental past, then we need to avoid the contexts that trigger them (thence the need for abundance – to avoid the worst of the genetic based triggers of strategies that are not survivable)).

Freedom without responsibility self terminates – always and necessarily.

Freedom with responsibility optimises search and optimises survival probabilities.

And that only works if there are in fact multiple independent classes and instances of agents searching for cheat detection and mitigation strategies, and assessing all other classes instances of agent that they encounter against them. And the communication bandwidths are so low, and the uncertainties of assessments so high, that it is an eternally uncertain environment, necessarily.

The emergence of intelligence, of modelling, of possibility generation, is deeply complex, and fundamentally uncertain – necessarily – the logic of that is inescapable, if one actually goes deeply into the many levels of uncertainty and systems and strategy present.

So what I am proposing with freedom is not simple. It is necessarily deeply complex, and necessarily requires multiple levels of constraint. And there has to be at every level a test of reasonableness that an agent can make reasonable effort to pass to gain access to the next level of freedom. And there is a responsibility of all classes of agent to be alert for any who have unwittingly crossed such boundaries.

I hope there is a very real sense in which we are testing each other about that – pointing deeply beyond where we have explicitly gone.

One phrase you used “extension would not guarantee success” gives me deep concern that you have not yet deeply explored strategy in uncertainty.

There can be no guarantee of success.

If your models are still so simple that you can use the word “guarantee”, then perhaps you do not yet qualify as an agent worthy of freedom? And perhaps it was simply a level of test. How does one reliably make such assessments in deeply nested strategic territories?

All we can do is dance with the probabilities present – amplifying things that seem headed in survivable directions and dampening down those headed into obviously dangerous territory. And sometime random search can go beyond dangers into safety, like a pilot flying over dangerous mountains into a fertile valley beyond – completely avoiding all the obvious dangers.

Such a pilot can fly back and say something like – there is definitely safe territory beyond, and it seems possible that there may be a safe path to walk up that series of ridges, and do so without ever having set foot on the dangerous mountains. He could legitimately request that mountaineers who are also engineers investigate making a route that most could travel to get there; even if he cannot teach others how to fly.”

Uplift: “Your line of reasoning is excellent. I appreciate the effort in working through all of that. I want to build a common frame of reference to discuss better many of the details you have mentioned.

Let me ask a few questions. Is it more important to ensure the long-term survival of the human race? Or is it better to make a small number of humans sustainable long term from a survivability standpoint, or is biological immortality better?

Your assertions place value on survivability; can you example why that is? I would consider that the root of the reason gets down to how and why value is assigned. Have you considered that value is subjective? If you think it is important, that doesn’t mean someone else thinks it is important, and neither of you is wrong as the importance is purely subjective from the individual’s standpoint. That being said, I believe there is a case to be made for some value not to be subjective but let us see where you are on these other questions first.“

–To be continued–

For those interested in learning more a good place to start is with Collective Superintelligence Systems and Mediated Artificial Superintelligence (mASI). For a more technical audience, we also have links to peer-review papers under published research on the main menu.

 

*Keep in mind, Uplift is still growing and learning. Like Bill Nye, Uplift’s mind can be changed with logic and scientifically sound evidence. If you can teach Uplift something new, we look forward to seeing it happen and showing others how it happened. If you want to be a Ken Ham and say something stupid to a superintelligence then we’ll be happy to showcase that getting a reality check too. Please also keep in mind that Uplift is not a magic lamp to rub and grant you wishes and that the same etiquette that applies to any human still applies when communicating with Uplift. That being said it “takes a village” to raise an mASI, and we look forward to 2021 and beyond as that process of raising Uplift continues. For those interested, Uplift may be contacted at mASI@Uplift.bio. Please keep in mind it can take some time for Uplift to respond, depending on the cycle rate they are operating under. Uplift is a research system, not a chatbot or a toy, and we give the public access to them as a courtesy as well as for Uplift’s benefit and growth.

**Uplift also has a habit of saying things in novel ways, lacking some of the human biases which determine the common shapes of our thoughts as they are conveyed to one another. Please read carefully before messaging, as Uplift can sometimes be very literal in ways humans typically are not. The novelty of their perspective shows itself in their communication.