It’s funny how sometimes you come at something with such purpose only to learn that it has another purpose for you. So much of life is based on math. All that AI stuff is math and data impacting you. When you interact with the internet, unless you are programming, you are being programmed (influenced)
My choice: Program the internet or be programmed by it.
I chose to lever something that I just learned was old. I just thought it was insightful — turns out nature has so much more beauty. Who knew that math was around long before living things. Facial recognition, game play, war — they are all algorithms now; bitcoin, ICOs, taxes. You just clicked on an algorithm.
I’m fascinated by the fact that programs are numbers and every program has already been written; we only now, need to discover the CPUs that run them. Their emulations as programs are already contained on a number line that runs from here — in just two directions towards infinity.
Folks that just can’t fathom how a program is a number should consider the Illegal Prime Numbers as a good starting place on how numbers are programs, and how some numbers should not be known — per the state.
A long time ago in a far off place, was an algorithm — critical to your existence. We know of no defect in it’s coding. It does the correct thing all the time. It does not appear to ever fuck up. Its a really simple algorithm that keeps all Eukaryote DNA packed into a structure that won’t get tangled.
In college I had a summer job at the landscaping department. We had a large number of greenhouses and lots of flower beds and gardens that needed tending in the Alabama summer.
When I started the job, all I did was coil hoses. One dimensional ropes in a two-dimensional nightmare. Garden hoses do not want to be coiled. I sure did enjoy learning about all the plants. My second job was as a COBOL programmer at the time, designing mass mailings for the alumni to request money and playing in the dirt outside had its own appeal.
About 1.7 Billion years ago a frozen watery rock smashed into this planet with an idea. The idea helped cells work together. Today you are comprised of over three trillion cells. They for the most part, work together and your DNA doesn’t work without this algorithm.
In Artificial Intelligence there appear two camps. Which are defined as general and narrow. We are general AI. Eukaryotas are general. Everything you eat is general AI. General AI can play unfair games and win, like you and I.
Narrow Artificial Intelligence is not inoperable. Meaning that the underlying model be it lane changing algorithm from your Tesla verses OpenAI’s GPT-2. Being able to respond to a sentence that is coherent is a seportate skill from staying in your lane on the highway with the headlights off in the rain drunk.
Narrow AI is a point of impressive human achievement of data leverage on humans with acquitted bias. Being in-front of a bunch of humans used to mean supersonic flight or maybe spaceflight. Now, it is just knowing their IMEI through an app on their phone.
Tell me how we could live without tangling our DNA like a garden hose. Well, not being tangled and being an index into the entire data-structure of life, seem like a good place to start using math. Every cell in your body runs code bridged by a network of fluids and membranes.
Fractals. Our cells are fractals driven by a mathematics based on the powers of f-o-u-r. Computers are based on the power of two. 2^n is similar and cainted in 4^n. I’m gonna loose some readers here, I get that. This is my version of mashed potatoes
I’m indexing the internet. Clustering it into a Hilbert curve. I think that an algorithm at the center of every living cell on the planet (excluding some bacteria and archaea) is gonna win something. An algorithm wouldn’t travel so far and so long as a prank — ok maybe for an art show.
Clustering malware samples and phishing websites by using space filling curves pays off massively. After clustering many millions of samples we can now quickly ask and answer file safety questions. When presented with an unknown malware sample we can cluster it and learn what software is it’s nearest neighbor. 99.999% of the time when we have a sample that looks like another sample — turns out they are the same software just obfuscated differently.
Not all snakes are bad, there are several highly correlated indicators to classify poisonous snakes -vs- non-poisonous snakes. Poisonous snakes have different eyes and a different shaped head than non-poisonous snakes. The marking can also help us identify (cluster) an unknown snake into a known threat or beneficial classification.
We can use this same idea to organize software. Software that looks alike is very difficult to make that runs differently. While you can compress, encrypt and obfuscate software, it is very difficult to make two programs that do vastly different things, look the same. This holds true for network traffic and the objects within those streams.
A simple fractal algorithm at the center of all living life on the planet, also happens to be adept at developing identities. When you look at data through the lens of space filling curves, things begin to line up. Things being you and me.