Can AI be programmed to be moral?

There’s a blog called “Wait But Why” that does a good job of taking dense, philosophical topics and presenting them in an amusing but thought-provoking way. The author recently did a very long post on the kind of artificial intelligence technology that many think is around the corner (possibly only a few decades away.) He concedes that AI could radically reshape the course of history, possibly by bringing about a Utopia or possibly by causing humanity’s destruction.

In part II of the post the author gets into the more negative secenerios. A point he makes well is that AI wouldn’t end humanity because the AI is “evil” or out to get us, but rather that humanity’s extinction might just be a logical output of of badly coded commands. A simplistic example would be someone tells an AI to solve world hunger and the computer “thinks”, “Well, only living humans are hungry so if I kill them all, they can’t be hungry.” Again, a very simplistic example, but a much more convoluted one could occur and wipe us out.

So the challenge is that we have to program our morality into the AI. It needs to follow our moral rules. But this is interesting because, as I’ve pointed out before, we really don’t have a clear and concise ruleset. We think we do, the Golden Rule being a good example, but
the Golden Rule is really quite flawed
. It’s more true to say we have a general set of moral intuitions that we kind of follow sometimes. Hardly the sort of thing that can be fed to a computer program.

So maybe the great challenge for humanity right now is to create a purely logocal, objective set of moral rules. I personally suspect that is impossible,

What’s up with cholesterol?

I’m often talking about the foibles of the medical establishment, though usually as related to pain management and such. But I’ve been reading a bit about cholesterol and it seems like conventional wisdom about several aspects of the topic has been turned on its head.

For example, the decades old advice about avoiding foods with high cholesterol (like eggs) doesn’t seem to hold up. Avoiding these foods doesn’t have much correlation with the cholesterol levels of your body. (Cholesterol is actually produced by your body.) In this case, you aren’t what you eat. Don’t take my word for it.

In February, however, the Dietary Guidelines Advisory Committee (DGAC) rocked the nutrition and medical worlds by changing their tune.

In its report, the committee states: “Previously, the Dietary Guidelines for Americans recommended that cholesterol intake be limited to no more than 300 mg/day. The 2015 DGAC will not bring forward this recommendation because available evidence shows no appreciable relationship between consumption of dietary cholesterol and serum (blood) cholesterol. … Cholesterol is not a nutrient of concern for overconsumption.”

Perhaps more interesting is the news on saturated fats which we’ve been advised to avoid for years because it raises cholesterol levels. You probably know that there are two types of cholesterol: Good HDL cholesterol and bad LDL cholesterol. But is LDL really bad? It turns out there are different subtypes of LDL

LDL comes in four basic forms: a big, fluffy form known as large LDL, and three increasingly dense forms known as medium, small, and very small LDL. A diet high in saturated fat mainly boosts the numbers of large-LDL particles, while a low-fat diet high in carbohydrates propagates the smaller forms. The big, fluffy particles are largely benign, while the small, dense versions keep lipid-science researchers awake at night.

So, by advising people to avoid saturated fats and eats more carbs, health advisors may have been inadvertently raising people’s levels of really bad, small LDL cholesterol.

Oops.

More thoughts on jazz

Lately I’ve been working on playing jazz on guitar. This means learning a lot of jazz tunes and really getting comfortable improvising over the classic jazz changes.

I’m starting to understand a key difference between jazz and pop. Pop music is, I believe, really focused on crafting the best, most immersive song experience. A chord progression will be tweaked to perfection, engineers will spend hours hunting for the right organ sound, and a solo will be played over and over until perfect. The idea is to create 3 minutes of bliss. A jazz song takes a more different approach. Jazz is dedicated to the concept of improvisation and is willing to let things suffer a bit to take risks. If you record a tune and the sax solo isn’t all that great, nobody will demand redoing it. (Often jazz is recorded live with various instruments bleeding into each other’s mics so redoing it isn’t even possible.) A certain amount of Imperfection is the price of doing business in jazz.

To put it another way, I think if you are listening to a pop song and your mind is wandering, in some sense the song has failed. It hasn’t totally captured you. With a jazz tune, particularly a long one, it’s ok for the mind to wander. Not every moment has to be great.

The anti-guilt pill

For a while now I’ve heard of a particular drug that purports to dull the formation of painful memories. I’ve always been a little unclear on how it works but I believe it takes away the emotional sting of the memory while leaving the recollection of the events. Ideally it could aid people who have suffered horrible crimes or soldiers suffering from PSTD. I had not heard of a more controversial use: the pill as a way of ducking emotional damage caused by committing heinous acts, especially in war time. This article, from 2003, describes a scenario.

The artillery this soldier can unleash with a single command to his mobile computer will bring flames and screaming, deafening blasts and unforgettably acrid air. The ground around him will be littered with the broken bodies of women and children, and he’ll have to walk right through. Every value he learned as a boy tells him to back down, to return to base and find another way of routing the enemy. Or, he reasons, he could complete the task and rush back to start popping pills that can, over the course of two weeks, immunize him against a lifetime of crushing remorse. He draws one last clean breath and fires.

That sounds a little overdramatic but makes the point. The rest of the article is a very even handed look at the whole issue. Some might say we can never use the pill in this way as it will destroy our humanity. But the response is that, look, if a killer is wounded during his crime, he still gets medical treatment for his physical wounds. Why would we deny him treatment for his psychological wounds? And if the person is a soldier why should he be doomed to a lifetime of guilt why the politicians who put him in the position get off scot-free*? It’s quite an interesting ethical debate.

* Writing this sentence made me consider how the term “scot-free” came to be. You’d think it was based on some story about a guy named Scot, but not so. it’s derived from an old english term that means exempt from royal tax.

Timing your mental activity

A while back I was reading a book titled “The Circle of Consciousness.” One point it made, one that we’ve all heard before, is that different people are more alert and functional at different times of day. Some people are morning people, some are night owls, and some are, according to the book, a kind of hybrid person that comes alive after waking up, then burns out after a few hours but can then have a second wind around afternoon or evening. I suspect I fall into that category.

So why is this? I don’t really know though I suspect it has to do with the way your metabolism varies throughout the day. At certain points maybe energy can better get to your brain or something.

It seems our eating schedule affects this as well. I usually wake up and have a not-heavy breakfast (plus coffee!) I can then work on whatever for a good couple hours and get things done. Eventually the nagging of hunger gets to me and I’ll have a lunch. And almost always my brain then conks out a bit; I become more sluggish. This seems like the opposite of the way you’d think it would work—more food should give me more energy. But I find that slightly hungry morning period is my best period for mental activities. (I tend to write these erudite blog posts during that period.) To be slightly hungry actual makes my brain run better.

I could look up the whys of this but in a way it doesn’t matter. What I try to do is organize my day so that key mental activity takes place during that first hungry period (or perhaps later in the day at my second wind) and mundane, unintellectual stuff is after lunch.

Maybe the trick for optimum mental ability is the classic six light meals a day program that keeps your metabolism burning but never overwhelms you digestion.

Computer made comic art

I’ve mentioned that I’ve been drawing a lot of comic book style art lately. And I’m always talking about the notion that computers and robots will soon be taking away a lot of people’s jobs. Today I woke up and found myself thinking about how software could allow non artists to render decent looking comic book art.

We all know computers are great at numbers and calculations. So the question is, can we turn a scene in a comic book world into numbers and calculations? That is exactly what happens with the kind of 3D animation so prevalent in movies (and there we have the added complication of motion.) Basically, any object—a box, a cat, a person, a spaceship—can be reduces to a series of lines and curves and these can be rendered as numbers. So I can envision a software where you basically say, “show me an office space (basically the interior of a box) and put inside it a desk and a guy in a business suit.” Then you could rotate the scene around or move the guy, or move his arms at the joints (like and action figure etc.) From there the software could apply different styles to the way the lines are drawn (using a thin or thick pen for example, or using different kinds of shading techniques.)

This would go a long way towards allowing dilettantes to produced decent looking comic books. But would it also put real comic book artists out of work? I dunno… probably. At least it would make the job market tougher. I suspect what you would start to see would be more of a hybrid approach where some of the art is produced by software and some by an artists.

I’ve also mentioned the idea of software writing stories. I wonder if we will see in my lifetime the first comic book fully created by computer. At point we will know computers are our masters.

How does your self-driving car handle the Trolley Problem?

In the field of ethics, you often hear discussion of “The Trolley Problem“, a fictional scenario where a person is forced to chose the best outcome from a situation where at least one person is guaranteed to die. I stumbled across this web article which makes the case that the self driving Google car could bring the trolley problem to reality. If you car is headed into a crash should it sacrifice you to save bystanders?

How will a Google car, or an ultra-safe Volvo, be programmed to handle a no-win situation — a blown tire, perhaps — where it must choose between swerving into oncoming traffic or steering directly into a retaining wall? The computers will certainly be fast enough to make a reasoned judgment within milliseconds. They would have time to scan the cars ahead and identify the one most likely to survive a collision, for example, or the one with the most other humans inside. But should they be programmed to make the decision that is best for their owners? Or the choice that does the least harm — even if that means choosing to slam into a retaining wall to avoid hitting an oncoming school bus? Who will make that call, and how will they decide?

I would offer an additional moral question. If my car decides to sacrifice me can it be programmed to quickly and painlessly kill me as opposed to leaving me to the destruction of a car accident?

Can you be an atheist and a believer?

Occasionally I mention people known as split brain patients. These are folks, usually epileptics, who’ve had their left and right brain hemispheres separated for therapeutic reasons. Numerous experiments have been done showing that these patients are, eerily, kind of like two people in one body. Neuroscientist V.S. Ramachandran took things in an interesting direction when he asked a split brain patient whether he (they?) believed in God.

The personal value of music

A while back I read some blog post by a guy describing a friend of his who still bought CDs. The guy did this because he believed that the act of curation was part of what made the music special for him. It wasn’t enough to have a vast collection of music at his fingertips (as anyone who has access to the web does), he wanted to have a relationship of sorts with the music. He wanted o purchase the CD, to eagerly read its jacket, to place the CD on and listen to the music, determining which were his favorites etc. I get the point though I think that kind of fetishization is a little fruity.

But there is something that I think has occurs when you have the massive digitization of music albums: each individual album becomes less valuable. Not just in a financial sense, but in a harder to define personal sense. I can remember as a kid that certain albums had a strong cache. “Sgt. Peppers” would be one, as would Pink Floyd’s “The Wall.” These albums were almost legendary in certain circles. I’m sure fans of hip hop or heavy metal or various other music genres can point to similar examples of their own. And additionally, when I was a kid, I would find certain unknown albums that I came to love and they became personal favorites of mine. (A bizarre album by the group Zodiac Mindwarp and the Love Machine was one.) This music had great personal value to me.

And I wonder of that sort of thing is disappearing. Because music can be pertained with so little effort is music losing not just financial but personal value? The guy buying CDs above is sort of forcing himself to maintain the previous value of music (personal and financial), even if the rest of the world has moved on.

This is counter to the consumer oriented forces of “more is better” who argue that the cheapening of things can only be good for people. And I suspect they’re basically right in terms of food and other basic needs. But not so much in regards to objects of personal fetishization. The valuation of such things has always been an ethereal process—exactly why a culture values one album of music over another is unclear (especially since music really has no purely utilitarian value the way food or shelter does.)

It’s a mystery.

Are we wired to created Gods out of our heroes?

If you take a look at rock history you notice that a lot of rock stars were or are complete douchebags. I’ve noted before that John Lennon, icon of peace, was actually kind of a violent fucktard. Warren Zevon, whom I’m a great fan of, was a violent alcoholic. It’s only recently I learned of Eric Clapton’s famous racist speech from the 70s.
(You can hear his clueless defense decades later, here.)

The truth is, you seldom hear people talk about this. Fans and music journalists seem to be able to look past these behaviors and continue their adulation of these musicians.

I was recently reading a text that touched on the idea that humans are wired for a certain kind of spirituality, a certain sense of mysterious forces in the universe. We are, the idea goes, wired to believe in god. (There’s actually are very interesting, albeit flawed book called “The God Part of the Brain” which is all about this stuff.)

And I wonder if this is partly why we can be so forgiving of rock stars who go bad? Are we programmed to view them as Gods and therefore incapable of evil? (Ironically, graffiti that peppered London in the 60s did claim that “Clapton is God.”)