One kind of intelligence, our kind

For me, intelligence is something which is inherently human. This is it, this is how we think about things and that is the only way. Artificial intelligence is something that we build by organising those same thoughts
in a “mechanical” system. So any kind of automation that we(humans) do, is a leap in the same direction.

We are trying to make the elements around us to work and behave for us, and they represent our own thought process. So when we made the circular wheel, we manipulated the elements around us in a way which represented our abstract thoughts about what a circle would look like.

Programming and machine learning is a natural extension of that same process, with more efficient and faster way to manipulate and design things. Because the feedback loop is very tight here.

When I think of the future of AI, I see it more of an extension of ourselves, rather than it being an external self-conscious alien entity. I believe this is really good way to look at it, and a great way to build for it.

Ghost in the Shell

The current programs that we have in our machines are a form of intelligence, it is programmed intelligence and it is an extension of the intelligence of all those programmers who wrote the code for the application — SDKs, APIs, libraries, OS, compilers, ABIs, Assembler, Instruction sets, Digital Circuits and everything in between.

In that sense, this programmed intelligence is a form of extended intelligence, just that it has a big temporal component, and it is NOT controlled by all of those programmers in real-time unless some of them have put a backdoor somewhere.

Machine Learning and AI is more(mostly?) dependent on the environmental data than on the explicit programs written by someone. In that sense, this kind of intelligence is more of a piping directly from the environment to the system, instead of it going through our brain for explicit “vital force” processing to convert it into a program. A kind of automation of programming itself.

The Interfacing Problem

The problem with the current development of smart machines is that it is in the hands of very few people. The capability of manipulating computing machines using a sophisticated interface, like modern programming languages (java, C++, javascript etc.) is still very rare. Very few people are comfortable writing good software, Further that skill is being used under the corporate system to concentrate that power to even fewer number of people at the very top. Nothing new there.

The most important area where we can focus our learning and expertise about AI and ML is in making of a truly fluid interface between us (humans) and them(machines). To a point, where there is no difference. An interface, where you don’t have to spend 10 years to learn one programming language, to manipulate computing machines in a marginal way.

Neural Lace

I am talking about a truly immersive and fluid one-on-one interface between a person and all the digital resources of a computing system, over a highly sophisticated protocol. Maybe a protocol, which is already part of our brain.
A protocol, which is already inside us, something our brain already uses to communicate between its own parts.

Yes, I am talking about the PLUG or Neural Lace

The Matrix (1999), Neo being jacked in

Something which we can just connect to and extend our intelligence. But this time, it won’t be built by some “alien machines” to put us in a simulation, to get the chance of perpetually serving us (yeah, we are not good batteries) in order to fulfil their own purpose. This time we would make the interface ourselves, and give us access to all the RAM we need so that we don’t have to memorise the Lyrics of “All along the Watchtower”

As a side note, the actual interface won’t be this clunky as in the movie above, in fact it would be completely invisible. But that’s just a hardware detail.

Singularity Renaissance

The Animatrix (2003), Animatrix: Matriculated

Now, I want you to picture a situation where this is possible. Where you are connected to all (or a lot of) computing resources over such an interface. An interface which is as fluid, as you are connected to your fingers. The way you control you fingers, how you grab your mouse, how you can so diligently type on that keyboard, how you can take an eye lash from your eye, you have such fine control of this piece of hardware. And you don’t even realize its existence, it is so deeply part of your way of life. And it is invisible.

Think of it in the same way, what if you can get any piece of information, that human kind has ever known at the speed of thought. You don’t have to make a google search, pick a page to read, or find a book, and spend
the next 8 years to learn General Theory of Relativity. You can just come up with a thought, “hmm, what is General Relativity all about?”, and just like that, in a couple of seconds, you know everything about it, even more than what Einstein knew about it. You know about it, with as much familiarity as you know your name, your parents’ name, where did you do your schooling and what you had in breakfast.

Feel like going to a neighbouring galaxy, by yourself? Just think about how to pick the lock of a Trans-galactic Spaceship, and you can borrow your neighbour’s “Vogon 3000” for a joy ride which he built over the weekend.

You want to communicate your thoughts using music, you pick up a guitar for the first time in your life, and right there you start composing music with an instrument, which right now takes years to learn.

Think for a second. What human civilization will be capable of, if this was possible. The kind of creative leaps we will make in days, which took us centuries until now. But more important than that, think of the amount of personal joy it will bring. You can now learn anything you want, and then go ahead to work on your dream projects, faster than you can finish this sentence.