Hostile AI: You're soaking in it
omniorthogonal.blogspot.com
omniorthogonal.blogspot.com
The mistake that many people of the developed world seem to make is to think that the Page's of the world see us as a distinct class apart from the billion global poor who live on less than $1 a day. It's a straight line from their suffering being just the business cost of sending rockets to Mars, to our $60,000 a year lives being the cost of doing it at scale.
Right now, we can be comfortable in sharing Page's and Elop's abstractions. People we know aren't effected. When it comes our time, the best we can hope for is to be the Sixth Army on the Volga and get some lip service that our suffering was honorable and our death glorious and will go down in history a thousand years hence.
Look up your street to the left, then down it to the right. If you don't see a native American, it's probably you.
Meanwhile, getting everyone in the US to use electric vehicles instead of gas would make a huge difference in the world. A society where transportation costs are much lower will be a very different society. Throw in Google self driving cars and we may not be mostly jammed into crime ridden cities any more, for example, since commuting would be a few pennies of electricity sitting in a cabin sleeping or working on a laptop instead of driving.
A million children a year die from gastro-intestinal problems and diarrhea http://www.gatesfoundation.org/What-We-Do/Global-Health/Ente...
1.3 million children died from pneumonia in 2011: http://www.gatesfoundation.org/What-We-Do/Global-Health/Pneu...
200 million people were afflicted with malaria in 2010: http://www.gatesfoundation.org/What-We-Do/Global-Health/Mala...
Sure, you probably don't know any of those people. They don't have laptops or cars or spare cabins in the woods. And they certainly don't hang out at libertarian circle jerks.
In truth, there's a lot of injustice in this world, and that's not something that can be legislated or marketed away or even corrected except in some small way by charitable people doing charitable things on their own or in small groups. We shouldn't feel guilty for good fortune, but we can still help lift the lost, weak, innocent, dying.
Corporations, charities, and governments cannot, and will not, do that: they are an expensive middleman; more than that, they are an abstraction layer that insulates us from the pain of others.
This is humanity. This is us.
There's a cost and sometimes we display a weakness of character for not paying it.
So what? How does this justify his conclusion:
> But they have chosen a fantasy version of the problem, when human interests are being fucked over by actual existing systems right now. All that brain-power is being wasted on silly hypotheticals, because those are fun to think about, whereas trying to fix industrial capitalism so it doesn’t wreck the human life-support system is hard, frustrating, and almost certainly doomed to failure.
Even if the finance industry / corporations are a parasite, so what? That doesn't wish away AI: AI will still remain a threat (or not a threat) regardless of whether Goldman Sachs is dissolved. Threats are not a zero-sum game: if you show that finance is a parasite, that doesn't suddenly alter the universe to guarantee AI is safe and easy.
Even if we interpret him charitably as making a nuanced cost-effectiveness argument that 'we face many threats, have very limited resources, and corporations are more cost-effective to fight than working on AI', that still seems pretty darn unlikely. Corporations are a major political issue which attract the attention and effort of hundreds of thousands of people and inspire real world action like Occupy Wall Street; why does this fight need a few more people in it? Diminishing returns set in a very long time ago. And fundamentally, financial types have (or have not) been parasites throughout history. Their impact is bounded. They are not anything new under the sun.
Whereas with AI, we don't know whether AI will ever exist; whether the impact will be huge or close to zero; hardly anyone takes it remotely seriously (quick question: have you ever read a mainstream article which did not make either a _Termintor_ or 'Rapture of the Nerds' allusion?); and the sum total of all efforts on thinking about the topic is, what, maybe $7m a year? (Google's safety board + MIRI + FHI + the new existential risk org whose name escapes me). You have to assume some pretty extreme values to make cost-effectiveness criterion spit out 'nobody should be thinking about AI! Everyone should be swotting up their Marx!'
Conflating socially-harmful corporations with 'hostile AI' does both topics a disservice.
First, I don't think that corporations are nothing but parasites, and said as much in the article. I did say that finance was parasitical. That was probably too strong. Allocation of capital is a useful function; the problem is that the allocators are helping themselves to a far larger cut of the proceedings than the value they produce (and I don't particularly want to argue that here, just wanted to clarify what I meant).
Second, the more central point about whether the Singularity Institute types are wasting their time and should refocus their thinking. It is true that not many people are thinking about the effects of superhuman AIs, and there's no harm and possibly some good in a few smart weirdos focusing on that, especially since corporations have been around and critiqued for so long.
However, we are so far away from producing anything remotely close to that level of AI, that I fear their work is ungrounded. It strikes me as fun quasi-SF rather than serious engineering.
But I take believe they are seriously interested in world improvement, so my suggestion to them in their work on "benevolent goal architectures" is to study how existing goal architectures work, that is to look at existing economic and social systems where disparate individual goals are coordinated into cooperative or conflicting action, and how the results are related to human goals and human flourishing.
For my part, I am in the process of trying to understand their friendly AI work better so I critique it from a more informed standpoint.
At what point would you allow research to go forth? And you know, there's a lot of work that can be done which is not 'directly coding an AI'.
> But I take believe they are seriously interested in world improvement, so my suggestion to them in their work on "benevolent goal architectures" is to study how existing goal architectures work, that is to look at existing economic and social systems where disparate individual goals are coordinated into cooperative or conflicting action, and how the results are related to human goals and human flourishing.
But why would one expect any of that to transfer to AI? What do the internal dynamics of a Board of Directors have to do with, say, the Code Red worm? What lessons can we extract from 501(c)3 nonprofits which will tell us anything about deep-learning-based architectures? Do CEO salaries really inform our understanding of Moore's Law? Or can study of Congressional lobbying seriously help us better understand progress on brain scanning and connectomes? Do Marxian dialectics truly help us improve forecasts for when feasible investments in neuromorphic chips will match human brains?
The closer I look at corporations and modern economies, the more worthless they seem for understanding the possibilities of AI, much less engineering safe or powerful ones. Modern economies are based on large assemblages of human brains, acculturated in very specific ways (remember Adam Smith's other major work: _The Theory of Moral Sentiments_), which are limited in many ways, and which are fragile and nonrobust: consider psychopathy, or consider economic's focus on 'institutions'. Why do some economies explode like South Korea, and others go nowhere at all? Even with millennia of human history and almost identical genomes and brains, outcomes are staggeringly different. (You complain about corporations; well, how 'friendly' is North Korea?)
And this is supposed to be so useful for understanding the issues that we should be focused on your favored political goals instead of directly tackling the issues?
Studying human systems is one of the best ways of studying complex systems and systems engineering, which are already crucial for complex engineering projects, like developing a complex AI. Before we can even talk about our future binary overlords being hostile or friendly, we will have to study how basic, but constantly developing, AI integrates and plays off of human social systems. We have to gather quantified data about how two distinct forms of intelligence interact and what, if any, conclusions can be generalized to a future where humans are no longer the species with the highest intelligence.
You have no data about how real AI's would behave in our society except for fiction, which contains no more guidance now than the Bible did for 16th century astrophysics. We have no consistent models that explain our own intelligence, let alone an artificial one that has yet to exist. You can pontificate about Plato's ideal Terminator but it won't make a bit of difference until we get our telescope.
What does it mean to study a generic 'complex system' and 'systems engineering' and what does this have to do with estimating the potential risks and dangers?
> we will have to study how basic, but constantly developing, AI integrates and plays off of human social systems.
This presumes you already know all about what the AI will be and is putting the cart before the horse.
> We have to gather quantified data about how two distinct forms of intelligence interact and what, if any, conclusions can be generalized to a future where humans are no longer the species with the highest intelligence.
Consider an aborigine making this argument: 'we have observed their firearms and firewater, and know there are many unknowns about these white men in their large canoes; if we look at their capabilities, our best analyses and research and extrapolations certainly suggest they could be a serious threat to us, but we must reserve judgement and quantify data about how our forms of intelligence will interact with theirs'.
> You have no data about how real AI's would behave in our society except for fiction, which contains no more guidance now than the Bible did for 16th century astrophysics
Really? We know nothing about AI and our best guesses are literally as good as random tribal superstitions?
> We have no consistent models that explain our own intelligence, let alone an artificial one that has yet to exist.
Someone tell the psychologists and the field of AI they have learned nothing at all which could possibly inform our attempts to understand these issues.
I don't think it's any more arrogant than, say, Djikstra pointing out that submarines move in a completely different manner than a human swimmer. How arrogant were the Wright brothers in looking at birds and deciding to try to achieve some of the same goals by a completely different mechanism?
If computers thought the same way humans did, if they came with builtin moral sentiments, the product of a very unusual evolutionary history and social structure, if they had access to no novel capabilities, then humans and forms of cooperation developed over the last few eyeblinks (centuries) might be relevant. But then, no one would care about the issue in the first place...
Sure, the corporations will fight back through lobbying etc. but at the end of the day we are still talking about the human intelligence of the corporate masters versus the collective intelligence of the citizenry. So the strategy of the corporation is not a form of artificial intelligence. Most importantly, the corporate masters fully realize that they live in a symbiosis with the rest of the society so they will not pursue paths that threaten the human civilization, for example they will not engage in nuclear war.
Unfriendly AI on the other hand, is an artificial life form capable of superhuman intelligence that has no necessary symbiosis with humans. It will view humans in the same way we look at apes: an inferior life form from our past evolution that can be preserved in reservations, studied and certainly destroyed if it grows to the point of competing for our resources. Bonus, they will likely be imune to high doses of radiations and have a fault tolerant distributed architecture specifically designed to survive nuclear war.
Yeah, I find this (that financial actors are acting in a hostile manner) more tenable than the impression I took from your blog post, which was that the financial system itself was acting with intentionality.
> look at existing economic and social systems where disparate individual goals are coordinated into cooperative or conflicting action
I hope you can bring that perspective to the table; I've noticed that a lot of the public work is about preserving goals while modifying structure, while actually mediating between conflicting goals/values is a bit handwavey.
And yeah, I hope we're far away from human level AI, but we're talking about our uncertainty on conceptual distance to an undiscovered theoretical framework, here. It's not quite as bad as trying to guess when the Riemann Hypothesis will be resolved, but it has that quality.
Why do you think this? The idea behind the Singularity is that even an AI that starts far below human-level, but is capable of recursive self-modification, can bootstrap itself to superintelligence. We don't need to understand the human brain to create such an AI; we need only to:
A. create a the simplest of "digital animals", and
B. give it the capability to push itself into self-directed Lamarckian evolution to satisfy its preferences;
C. and have this all happen on electronic time-scales (e.g. successive "generations" being created in milliseconds or less),
and we'll likely see a Singularity.
> look at existing economic and social systems where disparate individual goals are coordinated into cooperative or conflicting action, and how the results are related to human goals and human flourishing.
"Existing economic and social systems" allow humans to cooperate with humans. The whole point of FAI research is to figure out how to make an AI that is "human"-enough that it would have cause to participate in such systems with us.
A paperclip-maximizer has nothing to gain from participating in a stock market, or an election. People do not want to be turned into paperclips, and it does not want paperclip-resources turned to human ends. Its goals are not disparate, but rather incompatible with our own.
A non-Friendly AI should not be conceptualized by your intuition as a certain kind of human that we can possibly learn to "live-and-let-live." It can, at best, be conceptualized as something like a Baby-Eating Alien[1]. Usually, though, it would seem to us something more akin to a supermassive black hole: an unlimited force that consumes many things we want and outputs nothing useful to us, that we must keep away from, and especially must avoid creating in the laboratory.
Also, evolution optimizes for survival and reproduction, not intelligence. While it may be possible to design an artificial evolutionary system that optimizes for superintelligence, I don't think doing so is trivial (or it would have been done).
Let's posit that your image of post-singularity non-friendly AIs is accurate. They are an immense and irresistible force, like a black hole. I can imagine something like that, but I have a great deal of difficulty imagining something like that that is also constrained by initial design decisions made by humans before the singularity, which is what the friendly AI people hope to achieve. To be fair I haven't read all their stuff yet and I'm sure they address this point.
[1] I think I read it in Trevor Kletz' book "What Went Wrong?: Case Histories of Process Plant Disasters".
That's the tough nut to crack. Everything is predicated on that capability. It's not about AI overcoming current challenges. It's also about it overcoming future ones - and evolving fast enough so that it's smart enough at the time it meets those challenges.
All these are non-trivial issues.
I also like the alternative problem that you posit: what can we do about existing semi-autonomous systems whose functioning is harmful to humanity? However, I don't feel very hopeful about any sort of solutions we might have to these problems. Semi-autonomous systems at societal level have existed for probably nearly as long as as humanity itself. Are there some good examples of such systems that were initially harmful but have been carefully cultivated by humanity to make their effects largely beneficial?
1 - The researchers can't program those institutions; 2 - The researchers intend to use AI as a tool for that goal.
Really we have no idea. True nothing around today is remotely close to human intelligence, but the question is; will AI be a slow gradual process, or will someone make a breakthrough and suddenly make a lot of progress very quickly? In any case, it won't happen at all if people aren't working on it.
1) Our current economic system is already a form of AI. Using the wikipedia definition, it is a system that perceives its environment and takes actions that maximize its chances of success. Humans that are in this system are inputs in the environment. They may have individual agency, but cannot direct the system as a whole.
2) The "success" that this AI is maximizing is not currently in line with optimizing human interest.
3) Instead of theorizing about how to make sure a future AI system is in line with human interest, we should look at the current dominant AI and attempt to improve that.
All three of those premises can be argued, and I've come to reasonable arguments on both sides just on my own with a few minutes of thought. But I think those are the premises to argue.
It's interesting to think of corporations as being Unfriendly in this sense. The analogy isn't perfect, though: humans make up a corporation's computing substrate, so they're forced to value humans more than the classic Unfriendly AI would.
Money is a highly inefficient way of improving the world.
True vis a vis itself - as money in and of itself does not have an ethical "direction." So I would say you are throwing the bullion out with the bathwater.
Money is a very efficient way of focusing effort - as your example indicates. How it is focused then is up to what we collectively value and how much we want to invest in determining where it will find value.
Well, no. Banks provide liquidity, investment opportunity, and loans. Just like McDonald's, there is a cost to this service, and the negatives of this cost may or may not be worth the "positive constraint" but to say there is no positive constraint is utterly rediculous.
"Of all human institutions, these seem the most parasitical and dangerous."
What of those institutions that wage wars and actually people killed?
keep in mind that these are the same institutions that at some point saved the rest of the European jews, the rest of the European gays, the rest of the European gypsies, the rest of the European communists, mentally or physically disabled etc.
AIs have no such constraints. However alien a financial corporation may be, it might as well be a single human compared to what AIs could be. If the problem is real, it's a different problem, and the problem of large human corporations is only the smallest, smallest taste of the problem.
I first heard of this as the 'Santa Claus' argument on a forum thread. Short version: just because you can't point at a single person and say 'he is Santa', doesn't mean that Santa doesn't exist. Despite living solely as a meme in people's heads, he has desires (make kids happy on Christmas, make kids behave), and can act on them (put presents under the tree).
Those who supported this position used Microsoft as an example. 'Microsoft', like every company, is at its core an idea or a meme that everyone just accepts exists. Parts of it get written down, of course, and various people hold different parts of it in their head. Nevertheless, the company is its own entity, with its own goals and means of achieving them, yet exists only in people's heads.
Krugman's model of currency crises for instance ends up with "markets" as a kind of Cthulu-esque omnipotent, amoral, and inscrutable actor.
[1] http://en.wikipedia.org/wiki/NP-hard [2] http://en.wikipedia.org/wiki/Local_optimum [3] Making this point in the service of Internet Clarity. I don't think we disagree (unless I've misread).
https://en.wikipedia.org/wiki/Philosophy_of_Ghost_in_the_She...
Long story short: People hear about an event or idea, which did not happen or does not exist, and aspects of that thing become a part of reality, purely as an extension of the collective psychological reaction of those who become aware of the rumor or myth.
Similarly, corporations can't do anything without the people that own them and work for them doing something. So viewing corporations as quasi-autonomous agents is misleading; they can be viewed as collective agents in some respects, but in the end everything they do has to come down to some human or humans taking some actions.
Basically, the author of this article is punting: he's saying the problem is corporations being "effectively independent of human control", instead of actually looking at the actual humans whose decisions and actions constitute what the corporations do.
Which just punts on the next question: why? Who set up the corporate structure that way? Answer: other humans.
Corporations and othe large organizations don't come into existence by magic, and they don't develop structures that make people, even at the CEO level, into replaceable parts, by magic. Human beings have to make choices for those things to happen. Human beings can also make choices to change those structures, if that is what needs to be done. But to do that you have to stop ignoring the fact that it is human agency, not some amorphous "corporate" agency, that is causing the problems in the first place.
So corporations are definitely NOT run by humans behaving sensibly, but instead by a strange machinery specific to the corporate ecosystem.
Corporations don't "behave" at all. What you really mean is that human beings behave differently when they work for corporations:
You keep your job at a corporation by promoting the corporate interest.
That's true, but that by itself doesn't make the behavior immoral. If the corporate interest is best served by making products that people actually want or need, and selling them at a fair price, then keeping your job by promoting the corporate interest is perfectly moral.
If, on the other hand, the corporate interest is best served by something else, then we, as humans, need to revisit the whole issue of corporate governance: what is it that puts corporate interests out of alignment with our interests as human beings and members of society? And you can't just say "get rid of corporations" because corporations are necessary: without them we wouldn't have enough food, wouldn't have houses, wouldn't have cars, wouldn't have computers or the Internet, etc., etc.
None of us can survive purely by our own efforts; we need to be able to specialize and trade, and we need to engage in collective projects that required the coordinated efforts of many people. Corporations are a necessary part of doing that. The fact that the legal powers we give corporations have been abused does not mean all corporations abuse them. And the abuse doesn't happen because corporations magically do things without humans doing things; the abuse happens because some humans use corporations as tools to prey on other humans instead of providing value.
Now imagine some external driver of change appears on the horizon. How will the company respond? How will individuals within that company respond? I don't think it is so clear how or why people make the choices that they do.
In terms of right and wrong, the law identifies the act, but justice identifies with larger patterns; it is elusive yet tangible.
That's often true; but that doesn't mean people aren't making choices. It just means they're making uninformed, and often bad, choices. And to fix the problem, they need to make better choices. Pretending that it's not human choice at all, but some "corporate agency", that is causing the problems is not the way to fix them.
eg
http://www.huffingtonpost.com/2009/06/02/government-sachs-go...
The article states that the financial industry* is the "hostile AI" not all corporations (in fact, it clearly states that even the hated "energy" corporations actually provide a product - even if that ).
Having recently re-read Dan Simmons' Hyperion series (sci-fi humans vs. AI masterpiece), I could almost see the analogy perfectly - finance is key to almost every part of our individual life, business and even political machinations. And in each, the big finance outfits are impervious, indifferent and have their own goals outside of their stated purpose.
* I'd expand slightly to include all of FIRE - finance, real-estate and insurance
The corporation that feed you might do other things other places to other people.
As an example the De Beers in the diamond industry.
If crime is so evil why has its pay fed my family, paid for my house
"If farmers are so evil why have they fed my family, paid for my house and created so many amazing products we surround ourselves with" - said the cow.
Believe me, I don't disagree with that. Maybe I should emphasize that, even though I live in the Silicon Valley now, I grew up under communist dictatorship in Eastern EU. So I think I know a thing or two, from a practical standpoint, about Dystopia - the indoctrination, the forced labor, the rationing of electricity, fuel and food, the permanent fear of the secret police. That - I have not forgotten, for how could I.
All I'm saying is - this golden age could be even better, far better, if it weren't for the constant skimming of the cream that goes on all the time at the top.
If anything, it's the relative prosperity that makes it unlikely that the situation will ever be fixed. People don't rise as one, in anger, unless they are literally hungry and cold, in a physical sense. That, too, I know from first-hand experience. I've lived the incandescent emotions of the rioting mob while being one of them, out on the streets, 25 years ago, back home, in the dead of winter.
But when they are more or less sated, and provided with adequate (if not highbrow) entertainment, people will allow astonishing amounts of corruption, bribery, and downright theft to keep going at the top of the pyramid of money.
Is my belly full now? Yes. Am I more or less free to do most of the things I want? Yes. But do I still see corruption and injustice at the top, to the benefit of Big Money? Yes.
I guess I'm one of those people who are strongly motivated by principle. Injustice remains injustice no matter how pretty the makeup.
"If the rancher is so evil, why does he give me bales of hay to chew on and allowed me to become the fattest cow?"
"Evil" is (typically) not about actions or results, but about motivations and intent. Sure, you got your paycheck, but what's the intent?
Now, government bureaucracies may be another story, especially once they grow lobbying arms ;)
Good luck convincing them to prioritize anything other than finding more heroin.
Anybody have an alternate link, or more information about where it might be found?
perhaps it is the "What are you really waiting for?" argument (probably a favourable shift in power, to answer my own question)
This is one of my big problems with LessWrong (and I've been reading it four years): for all their claims of relevance, a startling unwillingness to question the social structure in a manner that would decrease the privileges of Bay Area techies like themselves. Politics is the mind-killer, don't you know.
The offshoot Effective Altruist movement has the same problem: throw money at the symptoms (and denigrate anyone who doesn't do the same) while noticeably never, ever questioning the system the problems are in the context of.
Money isn't everything.
McDonalds would do anything to squeeze more money out of its customers. Like most corporations, they profit from manipulating the human masses.
Sure, you can argue all day against the usefulness of speculative speed trading firms - and even then you'd still have some decent arguments against you - but to argue that financial institution in general offer nothing of use is ludicrous.
Some relevant quotes:
Corporations are at least somewhat constrained by the need to actually provide some service that is useful to people. Exxon provides energy, McDonald’s provides food, etc. The exception to this seems to be the financial industry. These institutions consume vast amounts of wealth and intelligence to essentially no human end. Of all human institutions, these seem the most parasitical and dangerous.
it [the financial system] has interests that are distant or hostile to human goals
"Utilitarianism is a theory in normative ethics holding that the proper course of action is the one that maximizes utility, usually defined as maximizing happiness and reducing suffering."
He sees financial institutions as a waste, and therefore consuming vast amounts of resources to no useful or positive effect.
Finance firms provide lots of very sophisticated ways for money to get from people who have lots of money that they don't currently have a use for, into the hands of people who have a need for money they don't currently have.
Note this is a weak claim. I'm not saying anything about the efficiency of this process or whether the financial industry is good or evil or anything. But you can't criticize the finance industry without at least understanding and engaging its outputs.
The argument I would like to make is not that finance is bad intrinsically - as you've said above. The real problem is that the current instantiation of this concept has grown a series of goals of its own, and to fulfill those goals it is doing things that use the common, shared, limited pool of resources of the global economy, siphoning off stuff that the rest of us could put to better use.
In other words, they could perform useful services for us, and they do. It's all the extra layer on top of that that's the problem.
In a strict, mathematical sense, the OP probably exaggerated when saying they are entirely parasitical.
The financial industry is as evil as any profit oriented industry can be and to single it out as something completely wasteful and parasitic is unreasonable.