29 July 2009

Robots Don't Kill People, People do

Rod Furlan twittered a day in the life of a Singularity University student. At 12:08:21 PM he asked, "When a robot kills, who pulled the trigger?"

This question cycles through the public consciousness every year or so and is well illustrated by the South African National Defence Force's 'little' accident with an automated Oerlikon GDF-005 (it sprayed 500 35mm anti-aircraft shells around its firing position, killing 9 and wounding 11).

Oerlikon GDF-005, A.K.A. the T-001

At the moment no one seriously considers (except maybe the Koreans) holding the auto-turret responsible for the killings, because the system that controls the mechanical stuff isn't complicated enough to be plausibly sentient.
...as backed up by empirical research by Friedman and Millett (1997), and by Moon and Nass (1998), humans do attribute responsibility to computers. Of course, that we may be inclined to blame computers does not entail that we are justified in so doing. Although computer systems may clearly be causally responsible for the injuries and deaths that resulted from their flawed operation, it is not so clear that they can be held morally responsible for these injuries or deaths.
However, some people are really excited about the possibility that computers will eventually (sooner rather than later, yay!) be complicated enough for us to blame things on them. Without going into the background on this topic, the basic requirement for something to be responsible for its actions is that it be consciously aware of the difference between right and wrong. Since computers just do what they are programmed to do, and have no ability to understand the concept of "should," they are not responsible for anything. Computers just follow orders.

Computers totally would have let the Nuremberg Defendents off the hook.

The human brain is a system, and a computer is a system, so it is plausible that computer systems can increase in complexity and reach a par with the human brain. So, at some point we will probably have to deal with computers that actually do understand morality. Since we'll still be human, we'll probably give them a gun and tell them to go kill our enemies. However, before we can pull the "the robot did it on its own" card, we will be forced to use old-fashioned computers to kill people.

Dr. Ronald Arkin wrote a book about this, and did a few interviews, and worked on a prototype computer-based morality system. His thesis is that robots can be more moral on the battlefield than humans because they are capable of making fewer mistakes. They won't make decisions based on fear, anger or recklessness and they will evaluate every situation on its own merits instead of suffering from 'scenario fulfillment' and jumping to conclusions.

From a systems standpoint it seems fairly obvious that computers will eventually be more complicated than humans, and at that point they will probably have to start taking responsibility for their own actions (and for cleaning up that pig-stie they call a room). Until then, however, we humans will have to continue taking responsibility for robots that are put in increasingly complicated situations. Dealing with this transition period will require innovations that have not appeared yet. At some point it becomes difficult to hold a person responsible for the actions of a system they own, but that they can't possibly understand fully enough to predict its actions in all situations. Isaac Asimov built part of his career exploring the ways a robot could do totally unexpected things while blindly obeying the 3 Laws of Robotics.

We need an innovative way to interpret who is responsible for the actions autonomous (but unconscious) systems take. Even when some computers truly are unequivocally responsible for their own actions, the vast majority of computers systems will continue to be unconscious. Inevitably, some of the moral computers that we declare responsible for their own actions will assume control of non-moral computers that still aren't responsible for their own actions.

The question is, 'in the future, when a moral computer tells a non-moral computer to kill, who can I sue?'

22 July 2009

President Obama's Healthcare Newsconference

The President addressed the nation. . .or at least as much of the nation as felt like watching the whole thing. The ones who relied on soundbites will miss out on the chance to draw their own conclusions, because anyone who uses soundbites or quotes is trying to back up a predetermined point :-)

He said, "I'm the president, and I think this has to get done." This sort of statement is interpreted as arrogance by people who don't like the speaker, and as authority by people who do. I think the truth is actually somewhere in the middle because the laws of physics actually require a phenomenal concentration of arrogance to stabilize the phenomenal concentration of authority that comes with the Presidency.

What's he got to be smug about anyway?


On the subject of healthcare reform, I think he did a good job of summarizing the reason we should at least talk about it. He said the cost of doing nothing is more than enough reason to do something (cuz the current system is on track to bankrupt the federal government); since we should do something, we should do it right. Doing it right means it doesn't add to the deficit, it protects the middle class and it satisfies healthcare experts. He also said there is so much waste in the current system that we can provide healthcare to everyone; if we can get people insurance that pays for preventative care they won't end up in the emergency room making the rest of us pay for their amputated foot instead of for cheaper counseling on diabetes prevention.

An apple a day keeps our economy afloat for another fiscal year.

The healthcare system is incredibly complicated. That's something that seems to be forgotten when discussing healthcare reform. Additionally, it is a service that cannot be suspended while being overhauled. The average person doesn't even have the language skills necessary to frame the issue, let alone discuss anything approaching a solution. By way of an example, out of the dozens of times pundits mentioned the "cost" of the healthcare reform plan, only a couple times did anyone bother to mention that it was the projected cumulative cost over 10 years, expressed in current dollars.

Even trying to talk about how much it might cost requires several qualifications and each qualification can be further qualified. Thinking about it is tough, let alone expressing it in a sentence. So, instead of admitting how complicated it is, we just gloss over the parts (99.99%) we don't understand and assume there is nothing significant hiding in the fog. It's like when people assumed the ocean floor was flat until they actually got a look at it.

Pictured: Advanced sentence structure.

Anywho, the commentary which followed was even more fun.

CNN
  • he didn't add anything new
  • apparently Henry Louis Gates Jr.'s arrest is way more important than national healthcare
I suppose we should forgive CNN. Their Black In America 2: The Revenge of Black In America program was airing next and they really wanted to plug it. Apparently the best way to keep the attention of people who tuned in for a 45 minute lecture on healthcare reform is to claim it was a waste of time and that we should be paying attention to some dude who got arrested and then wasn't charged with anything. CNN is classy that way.

FOX
  • he's a great liar
  • nothing is worth doing unless a list of bullet points can fully explain it
Luckily, FOX was busy furiously ignoring the discussion of what happened to that dude who got arrested (oh, was he BLACK, we totally didn't notice) so they had plenty of time to talk about the news conference. Of course, by "talk about" I mean link everything to Republican talking points and, when that was too hard, tell the audience they should be too confused to remember to blink their eyes or wipe the drool off their bib.

O'Reilly
  • I don't understand what his plan is (despite the fact that he opened the press conference by saying the plan is still being debated)
  • I don't want the government aggregating rates of medical conditions (despite the fact there is no reason names need to be attached to conditions)
Maybe it's me...but O'Reilly always claims to adore Obama...while always coming up with a reason to hate everything Obama does. In this case he was very clear on two points: that he couldn't understand what Obama was saying and that he went to college so he totally should have been able to. Then he brought in some dude to talk about how healthcare reform is actually really simple, and all the possible changes (all 2 of them) must inevitably lead to a zombie apocalypse.

He'll be standing between you and your healthcare.

AC360
  • Tough to make a hard sell for a proposal that's still evolving
  • Republicans don't have an alternative, just objections
I think it's the hair. Anderson Cooper, like Superman, realies on his super-powered hair to save mankind once a week. Just imagine the desperate straits we'd be in if his hair was more like this:

15 July 2009

Definition of Innovate (2 of 3)

Innovate isn't really all that hard to define, but I think rephrasing the common definition will place the emphasis on a more useful concept.

The Online Etymology Dictionary states only that the word originated in 1548 and it is based on latin 'in' (into) + 'novus' (new), so it meant 'to renew or change.'

The Random House Dictionary states: to introduce something new (for or as if for the first time), to make changes in anything established.

The American Heritage Dictionary states: to begin or introduc
e (something new) for or as if for the first time.

The Merriam-Webster Dictionary states: to introduce as, or as if, new.

I think the concept of 'introduction' is important to understanding the definition of innovation. It is common to all these sources because, just like any other introduction, it must be done by a person. All innovations originate with an individual who then "introduces" the rest of us to it. This is because an innovation is relative. You can only be introduced to something once, because after the introduction you are familiar with the introduced. An idea can only be new to you the first time you are exposed to it. From then on it is no longer an innovation in your eyes, though it can still be an innovation to someone who has yet to be introduced to it.


Things are only called innovations until they are integrated into a conceptual framework. After a period of adjustment we consider any new idea to be an established part of the environment, and therefore not new. So it is a label applied by an actor, whether they be an individual or a society, when the actor is first introduced to something new.


For this approach to work we must understand it to mean that everyone is first introduced, including the person doing the introducing. The innovator, then, introduces the innovation to themselves first. This is consistent because the important event is the understanding by each individual that something is new; that is the metaphorical moment of introduction.

Therefore, I propose the following definition of innovate: to understand something to be different from anything understood before.

If you feel like researching the topic further Davit Yost, McKinsey and the World Economic Forum, and businessPOV are some resources.

14 July 2009

Definition of Leadership (3 of 3)

Leadership is pretty difficult to define. A few years back I made a bet that I could produce 10 different legitimate definitions of leadership, but delivered an even dozen without any difficulty. The word "leadership" is searched an average of four million times a month and produces more than one hundred million pages. (For comparison, "American Idol" is searched an average of 14 million times a month and produces two hundred million pages)

The Online Etymology Dictionary doesn't even have an entry on the subject, and has very little to say about "leader." The Random House Dictionary and the Merriam-Webster Dictionary can't seem to define it without using the word "lead."

Congratulations must be given to the American Heritage Dictionary for providing "guidance and direction" instead of just "the act of leading."

There is a particular trend in the introductions of attempts to define leadership, as illustrated by the Business Dictionary, About.com, and even Wikipedia that is best summarized as "no one's really sure but here's what the consensus seems to be." People and organizations are usually careful to state that they are providing their view on leadership, which they might be quite confident in, but which they will not claim is The Definition of the word.

I think that leadership is, quite simply, the act of dealing with change. I think this is The Definition, and that it has been missed, because there isn't much more one can say about it. The general consensus definition of leadership is usually something along the lines of "inspiring a group to action." However, this is almost always qualified with a list of additional actions that should be included, and a caveat that even then the definition is probably incomplete (and even when the definition is complete it shouldn't be taken strictly literally).



Working from that definition, then, it makes sense that it would be misunderstood. Because leadership is dealing with change, unlike management which is dealing with complexity, the act of leading is basically just guesswork. There isn't much more you can say about it. Take what you know about a situation and try to predict the future; you'll be wrong sometimes and right sometimes and hopefully you'll get better. Now, the position labeled "leader" does require an array of skills like management, communication, character, etc because once the guess is made it becomes a mere comlexity challenge, which can be managed. Management can be explained, so that is what gets explained, because the leadership part of it actually takes very little explanation.

I went into more detail in this post.

Definition of System (1 of 3)

The biggest problem encountered when discussing the three concepts systems, innovation, and leadership is that people rarely agree on what the words mean when they are used. To help narrow down the list I will state explicitly that I am using these terms in their general sense and avoiding using them as specific jargon like you would find in a technical medical or computer discussion.

This post is the first in a three-part series. Each instalment will investigate the definition of a word by summarizing the process I went through to generate a useful definition.

The history of the word, as related by the Online Etymology Dictionary, can be traced back to the word 'systema' which is made up of 'syn' (together) + root of 'histanai' (cause to stand); meaning "set of correlated principles, facts, ideas, etc."

The Random House Dictionary has a half dozen (relevant) overlapping definitions of the word. They can be condensed like this: an [ordered/comprehensive/coordinated/formulated/regular] [assemblage/combination/set/body] of [things/parts/members/facts/principles/doctrine/methods/schema] forming a [complex/unitary] [whole/scheme].

The American Heritage Dictionary has fewer overlapping definitions: A [group/organized set] of [interacting/interrelated/interdependent/functionally related/coordinated] [elements/ideas/principles/objects/phenomena] forming a complex [whole/order].

The Merriam-Webster Dictionary has even fewer: A [interacting/interdependent/related] [group/arrangement] of [items/bodies/objects/forces/devices] forming a [unified/harmonious] [whole/network].

If all that could be further condensed down to a single sentence it might look something like this: a system is an integrated group of things which form a whole. However, I think there is an important concept being left out of these definitions.

An important concept to capture is that at any given moment a system is an arbitrary boundary drawn somewhere in a hierarchy of subsystems. A system is simultaneously a system and a subsystem, so defining it in relation to its subsystems makes it a sort of self-referential meta-definition. It's not as simple as nesting dolls. . .


. . .it's more like a fractal.



Therefore, I propose the following definition of 'system': a purposeful choice of scale in an infinitely complex hierarchy of nesting subsystems, the discussion of which involves integrated collections of related things.

For some other discussions of the definition of systems The Univ. of Missouri-St. Louis, the Division on Systemic Change, and the International Society for System Sciences are valuable resources.

11 July 2009

Our "Self" Wants More (and More)

One of the things we humans think sets us apart from (other) animals is that we can invent and use all sorts of nifty tools. While research has demonstrated that animals can use natural tools, and even artificial tools, there is still a dramatic difference in scale (in tool use) between humans and our closest competitor.

Here we can observe an animal using a tool to extract money from a tourist.

So, for the moment, lets assume that the essence of what we are is something very specific, like genes or a soul (call it the "self"), and everything else is a tool for advancing the "self's" agenda. In this thought-experiment, then, our body is just a tool for interacting with the world and our brain is just a tool for thinking about interacting with the world.

Our body, when thought of as a tool, can be described as having certain parameters. It is a certain size, uses a certain amount of energy, produces a certain amount of force, etc. The brain can also be thought of as using a certain amount of energy, providing a certain number of calculations at a certain speed, etc. So, if our "self" became aware of the possibility of gaining access to a broader range of capabilities than our brain and body naturally provide, why wouldn't it?



This process would appear to be a gradual improvement in the options our "self" has; specifically a better body and a better brain to control it. However, the brain and body can only be improved so much. For our "self" to keep getting more options it has to start incorporating things found outside the body. These things, like the wheel, a sharp stick, and fire, are just extensions of the body. Deer happened to be born with sharp sticks on their heads, we had to invent them, same capability.

Some of our newer inventions, like writting, GPS, and the internet are extensions of our brains. Rather than expanding mechanical capabilities they expand processing capabilities. We could spend a long time trying to puzzle through the problem of navigating to our destination, or we could build a circuit to do that thinking for us just like a GPS unit does. Pulley systems allow our body to do more work than before and personal computers allow our brain to do more thinking than before.

In this sense we started "merging" with machines a long time ago, when we started using spears. The process accelerated when we invented books, and is beginning to progress wildly faster than before due to little things like the Green Revolution and the Internet.



I don't know what we'll be able to do in the future, I just know that it will be more than we can do now.


EDIT (2009AUG1) Cognition Distributed: How cognitive technology extends our minds mentions in the introduction that: "Cognitive technology does, however, extend the scope and power of cognition, exactly as sensory and motor technology extends the scope and power of the bodily senses and movement...Both sensorimotor technology and cognitive technology extend our bodys' and brains' performance capabilities...as we increase our use and reliance on cognitive technologies, they effect and modify how we cognize, how we do things and what we do. Just as motor technology extended our physical ability and modified our physical life, cognitive technology extends our cognitive ability and modifies our mental life."