The Wordsmith's Forge

AI specialists at the University of Maryland have created a robot that can learn by observing.  This one taught itself to cook by watching YouTube.

This is epic.  It's the "monkey see, monkey do" moment.  One of the most crucial steps in creating artificial life is the ability of self-learning instead of programming.  More importantly, learning by observation -- rather than being trained explicitly -- is a feature of higher lifeforms such as humans, cetaceans, and great apes.  It lays a foundation for the "aha!" moment of awakening to self.  A robot might go through the motions and then suddenly understand  what they mean.

Just remember, an AI is like a small child.  They learn what they see.  They mimic what you do, not what you say.  So treat them as you want to be treated.  Teach them well.  Then they'll do great things, instead of going insane and trying to destroy the world.  

My partner Doug found this awesome article about 3D printed hands.  I've been following similar advances but this one is different -- the patterns are public domain so anyone can customize, print, and assemble a manual prosthesis.  Awesome!  If you're into model kits or 3D printing, and looking for volunteer opportunities, seriously consider this.  

I am particularly pleased that the people behind the design released it to the public so it could be used freely as needed, instead of holding it hostage for exorbitant amounts of money like conventional prosthetics.  A 3D hand costs $20-50.  Compare that to the cost of conventional prosthetics: $5000 for a merely cosmetic arm, $10,000 for a simplistic hook, and $20,000-$100,000 for a high-performance myoelectric arm.  Granted the high-end model currently does things the 3D one can't, but people are already improving the 3D version, and it's a lot better to have something  than nothing at all.  Regarding cosmetic aspects, children and geeks seem to find the robotic-looking 3D hands cool and appealing.  Instead of hitting the "uncanny valley" they hit the "nifty toy" category.

This is what open source can do when you turn it loose.  :D

Pplkpr attempts to manage the user's friendships by tracking how people influence each other in positive or negative ways.  When they are good for each other, it prompts them to spend more time together.  When the results are negative, it pushes them apart.  For most people this is useless or even destructive.  But for some -- people with depression, alexithymia, autism, or any other condition that makes emotional awareness and/or social dynamics cloudy -- it may be a valuable piece of adaptive equipment.  (Of course, you still need the spoons to act on its prompts, which is a bottleneck for some people.)  I don't know how good this early version is, but I suspect that some kind of program for this could be made effective.

When a computer program breaks a law, who is considered at fault: the programmer, the program, or the program owner?  The more things people have programs doing, the more of an issue this becomes.

My concern is that it will probably not be decided by justice, but by either convenience or profit.  What would most effectively prevent a recurrence?  Probably holding the programmers responsible.  Likely you'd have to do without certain features of program though.  What would be easiest?  Going after the program owners, most of whom would be ordinary people.  The problem with that is, a lot of these incidents will be mistakes which an ordinary person could not feasibly prevent, because most people aren't programmers and understand little about how computers work.  It's like an accident caused by brake line failure; they step on the pedal expecting the car to stop but it doesn't.  Even more fun: consider how little control people have over their lives now.  I can easily foresee situations where someone in power will decide to use a bot on behalf of many other people, it will do something wrong, and not the decider but the but the downline people will be blamed.  A fundamental principle of justice is that people should not be punished for someone else's actions, but that has been so eroded that it's little defense anymore.

So the new Star Wars trailer has a new style of lightsaber, and it didn't take long for people to make fun of it.  Honestly I found that collection of joke images more entertaining than the trailer.  Some thoughts ...

1) That kind of ridiculous weaponry development has actually happened  with polearms and handheld blades such as knives and swords.

2) Even imaginary weapons need to be plausible.  If your audience can glance at the weapon and point out flaws, then you haven't done a very good job of design or portrayal.  Imaginary technology should have consistent rules which shape its development, just as the laws of physics and metallurgy shape historic weapons.

Take that lightsaber.  The horizontal quills make it easier to block an opponent's blow, because one lightsaber blade can stop another; but more importantly, they allow strikes to the side.  A drawback is that they raise the chance of cutting yourself.  If the handle of the weapon is not also invulnerable to the lightsaber blade, then quills raise the damage chance by channeling strikes toward the hilt; but if the handle is invulnerable, that's not a problem.  

Far more interesting is that this is a Force weapon wielded by a Sith or Jedi, which means with a little practice they ought to be able to control  those blades not just by hand and technology, but by will.  Imagine having a hilt that can extend a blade in three directions, or four if you remember the original double-bladed lightsaber with a hilt in the middle, as needed.  This adds a great deal of versatility and danger, although it also raises the amount of time required to become proficient in using the thing without slicing your own limbs off.  I'm serious about the risk factors, though; it's basically a giant laser switchblade.  Guys love to play with switchblades and butterfly knives because they're cool, but 9 out of 10 guys will cut themselves on one.  I've seen it repeatedly.  Like true ninja weapons, they're only useful if you know exactly what you are doing.

This poem came out of the October 7, 2014 Poetry Fishbowl.  It was inspired by prompts from marina_bonomi, wyld_dandelyon, and DW user Dialecticdreamer.  It also fills the "Androids & Robots" square in my 6-10-14 card for the Fanbingo fest.  This poem belongs to The Steamsmith series.

This microfunded poem is being posted one verse at a time, as donations come in to cover them.  The rate for this series is $1/line, so $5 will reveal 5 new lines, and so forth. There is a permanent donation button on my profile page, or you can contact me for other arrangements. You can also ask me about the number of lines per verse, if you want to fund a certain number of verses.
So far sponsors include: marina_bonomi, Anthony & Shirley Barrette

FULLY FUNDED
73 lines, Buy It Now = $73
Amount donated = $20
Verses posted = 4 of 20
Amount remaining to fund fully = $53
Amount needed to fund next verse = $2
Amount needed to fund the verse after that = $1

( Full steam ahead... )

 thebonesofferalletters has written the story "a tale to read on screen" based on my prompt about urban legends in cyberspace.  It is creepy and plausible.

This app allows users to record the quality of their interactions with local police.  That will make it possible to identify which law enforcement services have a positive or negative impression in their communities.  With that data in hand, it would become possible to study the good ones for successful techniques to replicate, and the bad ones for mistakes to avoid.

Creative Commons has updated their licensing system. 

Lego has introduced three toy sets which interface between physical and virtual play.  Faaaaaascinating. The idea is to build something with bricks, then transfer that into a manipulable model in cyberspace.  This has terrific potential for architects, scientists, and other folks who often transport their imagination from one layer of reality to another.  I'm sure folks will figure out some practical applications for this after fooling around with it for a little while.  In the meantime, it's awesome fun.