I Missed IWSG But It's Okay... Robots Might Want to Kill Us After All!

Hey all, I swear I was going to post yesterday. I got sidetracked. I don't say that because I expect that anyone was disappointed, but because I have an awful affliction where I feel like I have to make people like me. Studies have shown that people that apologize a lot tend to garner sympathy.

I'm so sorry I wrote that.

Did it work? Do you like me now?

Regardless, I stumbled onto this talk last week on the interwebanets and was pretty floored by the talk. The real quick of it is, we're all about 20 years of being jobless, no matter what profession we choose, and also, robots will most likely want to kill us all.

Great stuff. The vid is about 20 minutes long, but I guarantee it's the best thing you'll see all day.*

In case you didn't watch, because you're crazy. Let me tell you a story. Once upon a time, I read a book, called The Age of Spiritual Machines, by Ray Kurzwell. Any would-be musician from the 80's might recognize the name, because the man invented the most awesome keyboards on the planet. He's the guy that invented the keyboard that had pressure sensitive keys. A pretty big deal, if you ask me. His Kurzwell 2000 was the most awesome thing the world ever produced at the time.

Anyway, inventing musical instruments was just a hobby, his day job was being a futurist. He's the guy that introduced me to the technological singularity as a concept.**

In his book, he produced a piece of prose... hang on while I try to find it downstairs.

Waiting.

Waiting...

Waiting...  ....

Dammit. It's about two hours later and I just got back to my computer. I forgot that I was going to go get my copy of the book to look up the thing. I did, however, manage to go for a walk (my 10,000 steps per day, remember?) , and eat supper, and watch about 5 youtube videos (while I ate, multi-tasking). Basically, I was awesome.

But I forgot to get that book. I decided to see if that passage was on the interwebanets. Couldn't find it. Shazballs.

Anyway, in it, he had a short story printed, entirely written, I might add, by a computer. It was readable. Not blow-you-away-with-it's-insight-into-human-nature-amazing, but readable. The point I took away from it was not 'can I do better than that?' but 'how much better will programs be able to write fiction in the future?'

Well, I read that a very long time ago, now, in the future, where are we? Well, here is a video I found about that (which does seem ironic, because I'm trying to tell a story for those who can't spare the time to watch a different video).

To summarize, it's still early, but we're screwed.


Fiction Prototype from Phil Parker on Vimeo.


*There is no guarantee. I made that up.

** If you don't know what that is, think of the terminator movies, except that machines that fall in love with humanity and just take care of us instead of killing us. Well, that's his version anyway.

The Future is Here... But it Feels a Lot Like the Present

I am constantly amazed at the things that a well written software program can do. I think we're at the tip of a an amazing revolution, thanks to the confluence of cheap, disposable processors and clever software. I think of something like Google translate. I work for an international company that communicates with folks all over the world. There was a time, not so long ago, that it would be almost impossible to communicate directly with someone without a language to share, but translation software makes is a snap. I can see in English what was written in German, Spanish, Chinese, etc. Miraculous times.

Of course, one of the great things one can do with such a powerful technology is try to trick it. I think it's a good exercise, and at least for me, a very telling one. It let's us know that computers aren't so smart, they struggle with grammar just as much as we do.

To see where we are at, a fun game, that isn't really new, is to type something in English, translate it through several languages before eventually bringing it back to English again. I thought it would be fun to play with that and see what results I would get.

I wanted something short, but complex enough expose any goofs along the way. So I whipped up a quick sentence to use.

The first time I felt the wind on my face, I knew I would never set foot on the ground again.


That might not win me any awards, but it should do for now. I wanted to see what it would do if I went through several Germanic languages, to see if that would help keep the translation close when I brought it back to English again.


I translated it to German, then took that text and translated it to Swedish, to Danish, to Icelandic, before finally settling back on English. The result?

First time shooting a pitcher blow ansiktet, Jag Jag att knew would never be a true Footgolvet again.

I'm not a linguist, but I thought that running them through several closely related languages (which I did - at least according to this chart) would make a better end result. Granted, there were four translations before it came back to English, so I didn't expect stellar results, but damn. That's near impossible to decipher. 

So I tried the same original sentence again, this time using as many disparate languages as I could. I'm sure a lot of the Romance languages are closer to English than say, Icelandic, but still, I followed the chart. 

So again, English to Russian, then Russian to Spanish, then to Hebrew, to Hindi, and finally back to English. What did I get?

Ansiktet pitcher hit the first shot, Jaguar Jaguar Footgolvet ATT knew that he really will not happen again.

What did that teach me? Absolutely nothing, except that computers don't seem so smart now, do they? Score one for the humans. Kirk would be proud. If Google goes down later, you'll know why. I broke it.


*edit* Actually, I did learn something, both translations gave me the word 'Footgolvet'. What the hell?

Scientists No Longer Needed: We Have Machines That Can Do That Now

I heard this story a few weeks ago regarding the stunning news that a machine has uncovered the laws of motion (meaning that it didn't have a human tell it what the laws of motion were, it just figured it out) after playing with a pendulum for a few hours.

Damn.

To put that into perspective, it took humans a few centuries of work to figure those same laws out. And it took a little fella named Newton to invent Calculus to really get them worked out mathematically.

And some 'ol machine figures them out by messing around for an afternoon. I shake my head at the absurdity.

But what that tells me, on a much deeper level, is that we really are living in the future. Folks like Ray Kurzweil and Vernor Vinge have been preaching about the coming singularity for a while now... and I'm starting to get that little tingle in the back of my skull that they may be right. I just wonder if we'll really be able to tell when it's happening.

Oh, and the singularity is that magical moment when machines/AI surpasses humans in intelligence. At least in the Kurzweil version, once that happens we'll have a runaway chain reaction of advancement that will leave humanity surrounded by godlike machine intelligences that will be capable of insights and wonders that we won't even begin to be able to understand.

How can that happen? Well, the thought is that if you have a machine, smarter than any human, that designs the next generation of machine/AI/computer, then you'll have something that is greater than what any human could ever build.

Oh, and then this new, even smarter machine goes and designs and builds it successor, which is even more advanced.

And so on and so on. Before you know it, people are so removed from the process, and have been left so far behind intellectually, that we won't be capable of understanding machines anymore... hence the godlike entities I mentioned earlier.

Given that a rather plain machine was capable of uncovering the laws of Newtonian physics in a few short hours now. Imagine what sort of wonders it could uncover if it were a trillion times more advanced.

Again, we're really living in a sci-fi world.

All that being said, I'm a wee bit skeptical of the singularity actually happening as I described above, maybe a lifetime of seeing movies like the Terminator and the Matrix has made me take the prospect of computer overlords as silly.

But you never know.