Last week I asked a question, then I did one of the most infuriating (well, maybe) things I’ve seen done on a blog. I denied comments, offered no way to express your opinion, and left you to think about it. At this point I’ve no idea if anyone did.
Cryptic questions and lack of audience engagement aside, I thought it might be about time to introduce my project properly. I’ve given a general overview in the ‘Start Here’ section; enough to let you know that I’m making an imitation of an artificial intelligence that I’d love to see developed and documenting the process. Some words that tell you what it is, and a little bit of why it is, but not enough to deliver a clear picture.
For lack of a better way to describe it let’s break it down into the base components so we can all read from the same codex.
Artificial Intelligence has been around in popular culture for decades. The Terminator had it (though he also had this desire to cause harm), Hal from Space Odyssey 2001 had it (along with a couple of fatal flaws), even an episode of Scooby Doo or two have mentions of intelligence houses, robots, and other mechanical monstrosities. In most appearances something has gone wrong which threatens the safety of us humans (apologies to any non-humans reading of course). Opinions vary across stories as to whether the problem results from a malfunction, a logical error, or malicious intent developing from years of ‘slavery’.
The most popular side of A.I.’s story is what could go wrong and what will happen when it does.
But, as with most things, the popular side of the story isn’t the only side. Nor is it necessarily the side most indicative of how things actually are or could be.
Primitive forms of Artificial Intelligence have already been developed, and it’s definitely a growing science. Video games use a form of it to allow more realistic and engaging scenarios. They’re being tested out in bomb disposal, and medical operations. Prosthetic limbs are starting to incorporate part of that technology in order to operate in a more natural way. The search for ‘True’ AI continues, with the closest to date being A.L.I.C.E. – who at times you forget isn’t human.
Within its own sector there are debates raging on the ethics surrounding the creation of something truly autonomous. One group has even created a set of ‘best practice’ theories and taken great pains to point out that anything we make wouldn’t have the same values or emotions that we do.
All in all, it’s a fascinating field that isn’t at its peak yet. Technologically success hasn’t been achieved but it’s getting closer.
With this artificial entity I’d like to create the idea would (eventually, once the technology is in place) be to create a being that learned and grew based on what you told it. It should eventually be able to take on a life of its own and become a true companion, with the ability to disagree, call you on inconsistencies, and have genuine conversations. The prototype has a couple of conversation tracks that you can play through, scripted, and serves only to offer a taste of what might be eventually.
In my introduction post I touched briefly on imaginary friends, talking to yourself, and the subconscious. The theories surrounding these topics are what built this entire project, and sparked the question. Entire careers have been built on looking at what exactly these things are, or what they mean. As a result the best I can offer is an overview and a starting point, later you’ll be able to read what I’ve based my research on through the bibliography.
Do you remember what it was like, being young and brimming with imagination?
I’m starting the discussion on psychological theories with this topic because it’s received a lot of press in the last few years. The conventional view was that young, lonely children invented companions to play with and talk to when they couldn’t find the real thing. As a result children talking about them were viewed with a mixture of horror and concern, or it was tolerated as a phase – something they’d grow out of.
When I was a child I can remember having several, sometimes more than one at any given time. I had real friends too, I wasn’t exactly a lonely child in my early years, but it was so fun to make believe. Every cuddly toy I ever owned had a name and a story, along with my horde of barbies and toy horses (yep). If memory serves (and it may not, I’m notoriously bad at remembering childhood things) I stopped talking about it to the other children when I was around 7-8, and then to everyone by the time I was 10. I continued to have them until I was at least 14!
Following conventional wisdom I’d be considered abnormal and in need of help. Maybe a few of you are thinking it right now.
However, when it was researched recently it was found that not only do more children have imaginary friends than previously thought, but on the whole they were more confident, eloquent and creative. Further, children weren’t the only ones to have such ‘companions’; there were adults that did too. Normal, happy adults with day jobs.
Talking to Yourself
No longer the first sign of insanity, according to psychologists. Rather, it’s something we all do most of the time; usually internally (particularly when we think other people could hear us) but sometimes aloud. According to a study carried out within the last five years, when two groups of people are given a self control task and one group is prevented from giving themselves any kind of self-talk (by the use of distraction) the group that was left alone performed far better.
We use it to work through problems, remind ourselves of important things, keep to our convictions, any number of reasons. People humming under their breath on buses feel calmer while travelling.
It’s still considered taboo and odd. Despite the benefits that’s unlikely to change any time soon.
This is the most interesting topic, and the most difficult part to write.
With the conscious mind we’re helped to understand, because through asking some questions and observing reactions/answers we can work out what typical behaviours are for different personalities. We can work out those personality types, in fact. Up to a point it can be measured, categorised, and operates within a certain set of parameters for each person.
The subconscious remains a mystery. Stabs have been taken at its nature for years; it’s where dreams live, or intuition comes from there, or even that it’s where habits are formed and broken. Different levels of dominance have been assigned to it too, with no conclusive set of results.
Recently scientists have discovered that the subconscious is susceptible to subliminal messaging and can even learn that way, without the individual being consciously aware. This has been met with mixed emotions, many of the main concerns surrounding brainwashing and advertising.
In fact, no one knows what it’s capable of or what the limits might be.
Formation of an Idea
Fascination for the subconscious provided the initial spark. For years I’ve been devouring everything I could get my hands on relating to the subject, and there are still so many questions. I’d almost become resigned to not knowing within my lifetime.
Talking to yourself and imaginary companions are different flavours of the same thing in my view. Working through problems (usually social), psyching yourself up for things, relaxing and unwinding, getting to know yourself better. The causes and the benefits are similar.
The three subjects are unified, in a sense, by public approval of them. Fringe subjects, taboo topics, hidden behaviours that many of us exhibit in privacy; I started wondering if there were other ways to receive the benefits such actions bring without incurring the wrath of the masses. Artificial Intelligence opened a door, and out popped The Question.
What could we learn, what would we gain, by having something that would mirror us so well that we could (in effect) talk to our subconscious mind? Out loud. In a ‘safe’ environment, without fear of ridicule.
When you strip away all the created materials, the pictures, the programming, and get right to the heart of what I’m trying to achieve, that’s it in a nutshell.