This is a good paper exploring how computation with spiking neural networks is likely to work.

https://www.izhikevich.org/publications/spnet.htm

I agree with this article. Coding is the easiest and best part of the job. Way more time is spent testing than coding. Why take away the fun part, making your job only about the verification of it.

>Why take away the fun part

Answering that is easy: I won't be able to stay employed as a programmer if the other programmers use AI to become more productive, and I do not.

One of the best ways to slow down an opponents OODA loop in soccer is to choose inconsistent actions from play to play. As a forward receiving the ball from defense: dribble (with speed changes), pass back, pass across, etc.

It was pretty awesome hearing my son's team making this type of observation on the field as the opposition repeated the same offensive play, and adapt their response.

I agree with you, but he didn't say the word "sorry".

Check out "Thank You for Arguing" by Jay Heinrichs. It covers the same basic steps, but recommends to exclude the actual apology part. It's also a great book for other reasons.

I get your frustration with architecture of architecture. However this kind of AOA has been going on for a long time in a similar domain of control theory.

Model Predictive Control is a method of applying a model of a process to estimate it's state. Per Wikipedia, it's been used since the 80s. When you apply it to tracking something you aren't in control of, say an enemy plane, you need to use multiple models, and then pick the most probable solution.

It's not elegant, but it's not actually a bad way to get good results.

There's a lot of good information on modeling spiking neurons here: http://www.scholarpedia.org/article/Category:Spiking_Network...

Also, I like the papers linked here by Eugene Izhikevich: http://www.izhikevich.org/publications/index.htm

The new links discussing the capsule neural networks are somewhat reminiscent of this polychronous computation paper: http://www.izhikevich.org/publications/polychronous_wavefron....

https://hackernoon.com/what-is-a-capsnet-or-capsule-network-...

This article reminds me of this polychronization paper that describes how networks of spiking neurons result in the emergence of groups, and some discussion on how that could explain consciousness as attention to memory.

http://izhikevich.org/publications/spnet.htm

Plus you're pretty much screwed on RAII since you cannot return a value from the initializer. So by not using exceptions in C++, you also will often end up requiring a separate init function that can return errors.

It looks very cool, but they definitely moved & rotated slowly in the video. Hopefully they can keep the motion tracking processing latency low enough to not cause nausea.

Even with the slow movement it's quite jittery. Not sure how much that affects the experience.

The other thing is the demo environment is very dark. All the lamps are dim and angled towards the walls.

Actually that's an interesting point - Oculus has shown that you can do tracking and display updates on a traditional display fast enough to appear more or less fixed to the world.

But Magic Leap's display has a physically spinning component. I wonder if you move quickly there is a gyroscopic precesion effect that distorts the display...

Spinning component?

Yeah it's a fibre optic light that spins in a spiral pattern to scan out an image, kind of like a circular CRT.

At least that's what some of their patents describe.

I built my desktop from scratch (for the first time) in 2005, added a couple components from other computers as they died, and still use it with no issues. I've thought about upgrading components in it, but since they haven't died yet, and it still works just fine, there's just not much of a point.

Built a monster PC late last year and I see it lasting me a long time as well. I don't want to play the game of always jumping on the latest and greatest. Your pockets will always get emptied out and I've discovered accumulating material possessions and keeping up with the Joneses doesn't give me the endorphin rush like it used to. Thanks for this comment.

I did this back in 2009 and dropped a wad of cash on a Mac Pro tower. Thing is a frickin' champ. Upgraded it to 12GB of RAM and a nice SSD for the OS drive and never looked back. Turns out 4 Nehalem cores at 2.66GHz will last you quite a while if you don't do video editing work or play the latest games at full resolution.

This is the way to go, if you know how. Unfortunately a lot of people's brains just shut off when they have to think about their computer on that level.