Necessity is the mother of invention. It is also true that invention necessitates learning new things. And such was the case on the stormy Tuesday morning our story begins. Distant echos of thunder reverberated in the small 8 x 16 workshop, drawing my attention to the surge suppressor powering my bench. With only a few vacation days left, my goal of finishing the hacked dancing Santa Claus toy was far from complete. It was for a Secret Santa gift, and I wanted to impress. The Santa moved from side to side as it sang a song. I wanted to replace the song with a custom MP3 track. In 2008, MP3 players were cheap and ripe for hacking. They could readily be picked up at local thrift shops, and I had picked up a few. It soon became clear, however, that I would need a microcontroller to make it do what I wanted it to do.
It was bad when kids first started running up cell phone bills with excessive text messaging. Now we’re living in an age where our robots can go off and binge shop on the Silk Road with our hard earned bitcoins. What’s this world coming to? (_sarcasm;)
For their project ‘Random Darknet Shopper’, Swiss artists [Carmen Weisskopf] and [Domagoj Smoljo] developed a computer program that was given 100 dollars in bitcoins and granted permission to lurk on the dark inter-ether and make purchases at its own digression. Once a week, the AI would carrying out a transaction and have the spoils sent back home to its parents in Switzerland. As the random items trickled in, they were photographed and put on display as part of their exhibition, ‘The Darknet. From Memes to Onionland’ at Kunst Halle St. Gallen. The trove of random purchases they received aren’t all illegal, but they will all most definitely get you thinking… which is the point of course. They include everything from a benign Lord of the Rings audio book collection to a knock-off Hungarian passport, as well as the things you’d expect from the black market, like baggies of ecstasy and a stolen Visa credit card. The project is meant to question current sanctions on trade and investigate the world’s reaction to those limitations. In spite of dabbling in a world of questionable ethics and hazy legitimacy, the artists note that of all the purchases made, not a single one of them turned out to be a scam.
Though [Weisskopf] and [Smoljo] aren’t worried about being persecuted for illegal activity, as Swiss law protects their right to freely express ideas publicly through art, the implications behind their exhibition did raise some questions along those lines. If your robot goes out and buys a bounty of crack on its own accord and then gives it to its owner, who is liable for having purchased the crack?
If a collection of code (we’ll loosely use the term AI here) is autonomous, acting independent of its creator’s control, should the creator still be held accountable for their creation’s intent? If the answer is ‘no’ and the AI is responsible for the repercussions, then we’re entering a time when its necessary to address AI as separate liable entities. However, if you can blame something on an AI, this suggests that it in some way has rights…
Before I get ahead of myself though, this whole notion circulates around the idea of intent. Can we assign an artificial form of life with the capacity to have intent?
There’s pandemonium on the streets. You look down from your 4 story apartment and see hundreds of people marching and chanting. You pick up your phone and call your buddy, expecting it to link up to the nearest cell tower which will route your call to where it needs to go. Instead, without your knowledge, you link to a tricked-out police surveillance truck a few blocks away. They intercept your call and listen to the conversation in hopes of tracking the protest.
Most of us have heard of Second Life – that antiquated online virtual reality platform of yesteryear where users could explore, create, and even sell content. You might be surprised to learn that not only are they still around, but they’re also employing the Oculus Rift and completely redesigning their virtual world. With support of the DK2 Rift, the possibilities for a Second Life platform where users can share and explore each other’s creations opens up some interesting doors.
Envision a world where you could log on to a “virtual net”, put on your favorite VR headset and let your imagination run wild. You and some friends could make a city, a planet…and entire universe that you and thousands of others could explore. With a little bit of dreaming
and an arduino, VR can bring dreams to life.
Imagine you were walking down a beach, and you came across some driftwood resting against a pile of stones. You see it in the distance, and your brain has no trouble figuring out what you’re looking at. You see driftwood and rocks – you can clearly distinguish between the two objects without a second thought.
Think about the raw data entering the brain. The textures of the rocks and the driftwood are similar. The colors are similar. The irregular shapes are similar. Thus the raw data entering the brain’s V1 area for both objects must be similar as well. Now think about the borders that separate the pieces of driftwood from the edges of the rocks. From a raw data perspective, there is no border, and likewise no separation because the two objects are so similar. But yet your brain can clearly see a rock and a piece of driftwood – two distinctly different objects. So how does the brain do this? How does it so easily differentiate between the two? If the raw data on either side of the border separating the wood and the rocks is the same, then there must be an outside influence determining where that border is. Jeff Hawkins believes this outside influence is a very special and most interesting type of feedback. Read on as we explain and attempt to implement this form of feedback in our hierarchical structure of invariant representations.
In a previous article, we talked about the idea of the invariant representation and theorized different ways of implementing such an idea in silicon. The hypothetical example of identifying a song without knowledge of pitch or form was used to help create a foundation to support the end goal – to identify real world objects and events without the need of predefined templates. Such a task is possible if one can separate the parts of real world data that changes from that which does not. By only looking at the parts of the data that doesn’t change, or are invariant, one can identify real world events with superior accuracy compared to a template based system.
Consider a friend’s face. Imagine they were sitting in front of you, and their face took up most of your visual space. Your brain identifies the face as your friend without trouble. Now imagine you were in a crowded nightclub, and you were looking for the same friend. You catch a glimpse of her from several yards away, and your brain ID’s the face without trouble. Almost as easily as it did when she was sitting in front of you.
I want you to think about the raw data coming off the eye and going into the brain during both scenarios. The two sets of data would be completely different. Yet your brain is able to find a commonality between the two events. How? It can do this because the data that makes up the memory of your friend’s face is stored in an invariant form. There is no template of your friend’s face in your brain. It only stores the parts that do not change – such as the distance between the eyes, the distance between the eye and the nose, or the ear and the mouth. The shape her hairline makes on her forehead. These types of data points do not change with distance, lighting conditions or other ‘noise’.
One can argue over the specifics of how the brain does this. True or not true, the idea of the invariant representation is a powerful one, and implementing such an idea in silicon is a worthy goal. Read on as we continue to explore this idea in ever deeper detail.
Your job is to make a circuit that will illuminate a light bulb when it hears the song “Mary Had a Little Lamb”. So you breadboard a mic, op amp, your favorite microcontroller (and an ADC if needed) and get to work. You will sample the incoming data and compare it to a known template. When you get a match, you light the light. The first step is to make the template. But what to make the template of?
“Hey boss, what style of the song do you want to trigger the light? Is it children singing, piano, what?”
Your boss responds:
“I want the light to shine whenever any version of the song occurs. It could be singing, keyboard, guitar, any musical instrument or voice in any key. And I want it to work even if there’s a lot of ambient noise in the background.”
Uh oh. Your job just got a lot harder. Is it even possible? How do you make templates of every possible version of the song? Stumped, you talk to your friend about your dilemma over lunch, who just so happens to be [Jeff Hawkins] – a guy whose already put a great deal of thought into this very problem.
“Well, the brain solves your puzzle easily.” [Hawkins] says coolly. “Your brain can recall the memory of that song no matter if it’s vocal, instrumental in any key or pitch. And it can pick it out from a lot of noise.”
“Yea, but how does it do that though!” you ask. “The pattern’s of electrical signals entering the brain have to be completely different for different versions of the song, just like the patterns from my ADC. How does the brain store the countless number of templates required to ID the song?”
“Well…” [Hawkins] chuckles. “The brain does not store templates like that”. The brain only remembers the parts of the song that doesn’t change, or are invariant. The brain forms what we call invariant representations of real world data.”
Eureka! Your riddle has been solved. You need to construct an algorithm that stores only the parts of the song that doesn’t change. These parts will be the same in all versions – vocal or instrumental in any key. It will be these invariant, unchanging parts of the song that you will look for to trigger the light. But how do you implement this in silicon?