Accelerating Future

Researching the current state of “roboethics” (a lame term that marginalizes “AI ethics”, a more-relevant superset of roboethics), I find a bunch of references to a South Korean project to draft a Robot Ethics Charter. All these references occur in March 2007, and they promised the ethics charter would be released in April 2007 and subsequently adopted by the government. However, I can’t find it anywhere. Anyone have a clue about where it went? One article summarized the effort as follows:

The prospect of intelligent robots serving the general public brings up an unprecedented question of how robots and humans should be expected to treat each other. South Korea’s Ministry of Commerce, Industry and Energy has decided that a written code of ethics is in order.

Starting last November, a team of five members, including a science-fiction writer, have been drafting a Robot Ethics Charter to address and prevent “robot abuse of humans and human abuse of robots.” Some of the sensitive subject areas covered in the charter include human addiction to robots, humans treating robots like a spouse, and prohibiting robots from ever hurting a human.

Critics of the charter say that the charter is premature and may not have a practical application once robots are really an integral part of society. Says Mark Tilden, the designer of the toy RoboSapien, “From experience, the problem is that giving robots morals is like teaching an ant to yodel. We’re not there yet, and as many of Asimov’s stories show, the conundrums robots and humans would face would result in more tragedy than utility.”

“Asimov” refers to science-fiction author Isaac Asimov, who created a robot code of ethics for one of his stories. His Three Rules were: (1) a robot could not hurt a human or through inaction allow a human to be harmed, (2) a robot must obey human orders unless those orders would make it violate rule number one, and (3) a robot must protect itself unless that protection would violate the first two rules. These apparently served as inspiration for the South Korean Robot Ethics Charter.

However, South Korea’s Ministry of Information and Communication plans to have a robot in every household by 2020. “Personally, I wish to accomplish that objective by 2010,” said Oh Sang Rok, head of the ministry’s project.

Personally, I think Asimov’s Three Laws are a terrible inspiration for any roboethics code. The laws were created to be used as a plot device. When they disintegrated, a story came out of it. Unfortunately, they’ve actually been taken seriously as a possible solution to the problem of human-unfriendly robots and AI for many decades now. But Asimov himself said, “There was just enough ambiguity in the Three Laws to provide the conflicts and uncertainties required for new stories, and, to my great relief, it seemed always to be possible to think up a new angle out of the 61 words of the Three Laws.”

Back in summer 2004, the Singularity Institute launched a website project, “Three Laws Unsafe”, a critique of Asimov’s Laws riding on the publicity of the “I, Robot” movie. Check out the articles section, which includes a submission by myself.

But yeah, anyone know where that Robot Ethics Charter is, or the names of anyone who was working on it? We need to get our magnifying glasses out and scrutinize that shit.

2 Responses »

Some of you may have seen these papers already, as I mention them frequently, but they’ll important enough that I like to re-mention them regularly. They’re “Artificial Intelligence as a Positive and Negative Factor in Global Risk” by Eliezer Yudkowsky and “The Basic AI Drives” by Steve Omohundro. The papers are 42 and 11 pages, respectively. There’s no abstract for the first paper, but here’s the abstract for the second:

“One might imagine that AI systems with harmless goals will be harmless. This paper instead shows that intelligent systems will need to be carefully designed to prevent them from behaving in harmful ways. We identify a number of “drives” that will appear in sufficiently advanced AI systems of any design. We call them drives because they are tendencies which will be present unless explicitly counteracted. We start by showing that goal-seeking systems will have drives to model their own operation and to improve themselves. We then show that self-improving systems will be driven to clarify their goals and represent them as economic utility functions. They will also strive for their actions to approximate rational economic behavior. This will lead almost all systems to protect their utility functions from modification and their utility measurement systems from corruption. We also discuss some exceptional systems which will want to modify their utility functions. We next discuss the drive toward self-protection which causes systems try to prevent themselves from being harmed. Finally we examine drives toward the acquisition of resources and toward their efficient utilization. We end with a discussion of how to incorporate these insights in designing intelligent technology which will lead to a positive future for humanity.”

Feel free to post your reactions here.

4 Responses »

 Loading …

If the poll script is missing up (like it seems to be for me), then just put your probability and arguments into the comments.

28 Responses »

Hi, I just thought I’d repeat some general points about AI and our future.

If human-equivalent AI is possible, this is a huge, huge deal. It would basically mean that you could turn inanimate matter into intelligence. Say that it requires about 500 teraflops (Tflops), roughly equivalent to one of the fastest supercomputers today, to run a human-equivalent AI program.

A really fast supercomputer costs about $100 million. As you may know, the cost of computing power tends to fall exponentially with time. Even if this doesn’t continue forever, it seems like it will continue until 2020 or so at the very least.

Around 2020, 2025, 2030, or thereabouts, it seems reasonable to say that a 500 Tflop computer would cost in the ballpark of $1,000, if not less. If such a computer were sufficient to run a human-level AI, it would make sense for your random company to buy these computers and run them alongside conventional staff. They would be substantially cheaper. After all, these AIs could think all day and night without food, and their cognitive architectures could be boosted by direct access to number-crunching capabilities. They could share thoughts in a common format, instantaneously.

If one of the AIs became a genius, the others could just copy the cognitive features that gave the original AI those capabilities. The entire collective would never be far behind the leader, in contrast to human collectives, where our wetware is static and cannot be improved.

To actually influence the world directly, it would be helpful for these AIs to develop some robotic avatar. This would be using the robotics of 2020-2030. It would be reasonable to assume that the robotics chosen might be quite flexible and capable, especially considering that the AIs themselves could assist in reprogramming, fine tuning, research, and development.

Bacteria are idiotic, yet capable of turning a tonne of organic waste into bacterial biomass over a course of hours. Human-equivalent AIs would be smart, and have great incentives to convert raw materials into robotic or biological bodies for their habitation. Such AIs could even uncover the principles of thought and boost themselves beyond the human level, even if their access to computing power remains roughly static.

Combine AI with advanced robotics, add the motivation to improve both, and you have a potentially abrupt and disruptive transition on your hands.

What baffles me is when pundit say: “surely, such AIs would lack the capability to become major players in the human world in any short-term timeframe”.

My question would be: “how do you know?”

We humans cannot put ourselves in the shoes of an intelligence that has complete access to its source code, can rearrange its cognitive architecture to optimize its performance on narrow problems, share thoughts with its comrades at the speed of light, transfer itself from point to point on the globe at the speed of light, directly integrate itself with scientific instruments as sensory modalities, blend together autonomic and deliberative processes in a thousand ways that humans can’t, form beliefs and update them in mathematically rigorous ways, and so on.

Such an intelligence could come a long way in a really short time, or perhaps not. The point is that we don’t know. If a pundit expresses skepticism about the idea, their opinion is more likely to reflect the limitations they know apply to humans — not limitations applied to AIs.

It seems easier to argue that human-equivalent AI is flatly impossible than it is to argue that human-equivalent AI wouldn’t have a huge impact on the world once developed. It seems most reasonable to proceed as if it would.

18 Responses »

Yesterday I was in the middle of brushing my teeth, and thinking, “I wonder how Peter Voss, Justin Corwin, and the rest of them down at A2I2 are doing?” (In case you didn’t know, A2I2 is an AGI company.) For an outline of the architecture, go here.

Lo und beholde, I check my email this morning and there is an update from A2I2, dated May 22nd:

Towards Commercialization

It’s been a while. We’ve been busy. A good kind of busy.

At the end of March we completed an important milestone: a demo system consolidating our prior 10 months’ work. This was followed by my annual pilgrimage to our investors in Australia. The upshot of all this is that we now have some additional seed funding to launch our commercialization phase late this year.

On the technical side we still have a lot of hard work ahead of us. Fortunately we have a very strong and highly motivated team, so that over the next 6 months we expect to make as much additional progress as we have over the past 12. Our next technical milestone is around early October by which time we’ll want our ‘proto AGI’ to be pretty much ready to start earning a living.

By the end of 2008 we should be ready to actively pursue commercialization in addition to our ongoing R&D efforts. At that time we’ll be looking for a high-powered CEO to head up our business division which we expect to grow to many hundreds of employees over a few years.

Early in 2009 we plan to raise capital for this commercial venture, and if things go according to plan we’ll have a team of around 50 by the middle of the year.

Well, exciting future plans, but now back to work.

Peter

Well then. A2I2 currently has 16 full-time employees and a significant amount of funding. But an expansion has been in the works for a while. I wonder what their AGI engine will be commercialized for?

And as I often ask, “are we there yet?”

13 Responses »

The following video is from the “good ending” of Deus Ex: Invisible War, the 2003 dumbed-down sequel of the foresightful and flagrantly transhumanist 2000 title, Deus Ex, which was named Best PC Game of All Time in a 2007 poll carried out by UK gaming magazine PC Zone, among other awards. The video is a little corny, but I kept thinking about it this morning, and would feel bad if I held it back. If you have a problem with computer games or never play them, skip it.

The description of the video is as follows:

“With the destruction of the Illuminati and the Knights Templar, nothing stands in the way of the great utopia: the purest democracy, where every person in the whole world is counted by the great supercomputer Helios. With all of humanity now one, inequality and war have become obsolete.”

Yeah, the voice and tone is a little spooky. Remember, this is a fictional computer game. I’m not saying I want this to happen. This is food for thought. Don’t take it too seriously!

It’s interesting to see the YouTube comments on this video. I’m surprised at the people who have such a huge problem with an AI-run democracy. Just like humans, there will be good AIs and bad AIs. If AIs can manage billions of bits of complex information more easily than humans, and be entirely unselfish, then wouldn’t it make sense to integrate them into our governments? Only irrational prejudice would dictate otherwise.

Thanks to Steven Killeen for forwarding this to me. My response, when asked what I thought, was as follows:

“It’s alright, a crude caricature of what I’d want from Friendly AI. A lot of people might find it spooky. I’ve seen the good ending from the first Deus Ex, but this is even better. Helios is saying a little bit of pseudophilosophical stuff, I think a real AI would be even more compelling.”

12 Responses »

Watch SIAI’s new interview with AI researcher Steve Omohundro, president of Self-Aware Systems. His CV is here. A transcript of the interview is here.

3 Responses »