Oh right, these. I used to do these, right? I can do another one. Sure.

I swear to God, the science news just got boring for a while. Honest!

First up on our news roundup : scientists are improving computers’ AI by teaching them regret.

Insert generic pop culture joke here along the lines of “what did they do, [undesirable thing involving something or someone currently unpopular]?”

What did they do, [take the computers to see Atlas Shrugged?]. Like that.

Anyhow, the idea is that if you can teach the program to recognize situations it wishes to avoid ahead of time and hence let it avoid them, you have effectively taught it an analogue to the human emotion of “regret” and this will lead to better, faster, and more accurate results.

It seems to make sense. Imagining how we would feel in potential future situations is the basic way the all important “future” aspect of sentience operates. When choosing a path, we can project our minds into the future and essentially imagine various things happening to us, and use that as a basis for our decisions.

So when they say they are teaching computers “regret”, what they are really doing is teaching them to imagine future regret. Which I suppose involves the capacity for real-time regret as well. Imagine that, a computer that can feel stupid for doing something.

I fail to see how that is a big improvement over just good old statistical outcome analysis, though. Other than sounding cool and attracting funding, of course.

Next up on the roundup turntable : from our For The Love Of God, Don’t Try This At Home department comes the story of how hooking a nine volt battery to your brain makes you better at video games.

Yes, they did a study in which a simple device powered by nothing more than nine volts of battery made the subjects twice as good at a video game designed to help soldiers train for ambushes. In point of fact, the device actually made the subjects twice as good at the game.

But before hordes of nerds rush off to get the batteries and baling wire and spit-stick electrodes to their noggins hoping to really improve those frag counts on Team Fortress 2, I feel compelled to warn you that this is just one study, and a small one at that point. Hardly enough of an experimental basis for alligator clamping the battery from your Geo to your earlobes.

And what is more, transcranial direct current stimulation is nothing to mess with on your own. This is not like bodging together your own ham radio receiver over a weekend. If you screw up this time, the breadboard you fry will be your brain stem, and brother, that’s no easy fix.

Plus, honestly, to me this sounds like the sort of thing that could easily be explained by the placebo effect. People were thinking “Wow, my brain is now SUPER CHARGED WITH POWER” and it’s that feeling, not any true enhancement, that drove the results.

Finally, we throw caution and quite possibly causality, not to mention sanity, by diving into that freaky jungle that is the world of modern quantum physics with a story about how they have succeeded in the quantum teleportation of light.

By “they”, I of course mean “scientists”, not the usual informal infinitive “they” we use in such statements as “they say the secret to a power breakfast is complex carbs and protein” or “I am looking forward to the new movie based on my favorite fantasy series, but they will probably screw it up. ”

Leaving the actual quantum mechanics questions delicately aside, the thing you need to know about this whole teleporting light thing is that they basically used quantum teleportation to transfer information (in the form of light) and that really opens up the possibility of a future of quantum computing, which besides also sounding cool and attracting funding means a future filled with computers so smoking fast that they make the fastest computer on Earth today seem like a one armed man with an abacus.

That, in turn, opens up a theoretical question I have been pondering for quite some time : at what point would we no longer have any use for faster computing (or networking)?

Certainly, we won’t hit that point any time real soon now, but it is an interesting thing to ponder. Certainly, there will be a limit to how much computing power you need to have in your pocket. Scientists might need something that can create realtime models of solar systems or something, but you just want to play video games and watch movies, and current computers already do that fairly well.

Plus, with current network speeds being perfectly capable of piping good quality video right to your viewing device of choice, one wonders how much more “pipe” we really can use.

I look forward to finding out!