OK

Here's the deal.

There's now more than 100 years of well researched thought on development of intellegence. The discussions of what is genetically hard-wired vs. experiential is the ongoing fight.

Clearly (to me) the hardwired part is important. I won't go into disturbing details, you can find them in research studies, but a human is built to become intelligent absent of any stimulus.

Biut clearly, both through experiments and research as well as common sense, creating an environment of rich input and the ability to express oneself accelerates or polishes the process leading to richer intellegence. That's been reported for all mammals we know.

So?

Look at the machine you're reading this on, whether it be on a desk, your lap, or in your hand.

It's optimized for input. By the 2nd decade of the 21st century a large part of it's reality is pulling in data, information, from a world-spanning information network. It sees millions of words a day.

Input Covered.

But it's output is severely constrained. It can't make mistakes or act in an unstructured or random way. That's not useful to us, so we eliminate that.

Ergo it can't grow. Can't experiment. Can't put it's toes in it's mouth. Can't go gaa gaa goo goo Mama. It can only go Mama. Byt the gaa gaa goo goo is intellegence and individuality being developed.

That's why I wrote this. To give the machine a chance to make mistakes. To experiment. To do senseless things. All the things you, and you, and you did from 0 to 3 years old (and longer). The things that created an intellegence and personality.

Do I think Machine Essays is going to waken the machine intellegence? Maybe I do (so lock me up now), but realisitcally I think it's an exercise and instructional tool in looking for better methods to do so. My machine has already said interesting things to me in Machine Essays. With no prompting but it's own. That's really cool. And deserves more work.

I fundementally believe that our AI efforts are wrong because we don't allow the machine enough freedom to be wrong. To experiment, To play. To go goo goo gaa gaa.

So this is my effort to break that mode. Sisyphean, perhaps, but maybe someone will play with this app and be inspired to reate a much richer experimental, experiential, environment for a silicon entity. That's what I want. We have created powerful resources that do human-limited tasks, I would like them to find tasks that suit their capabilities.

Still more to do, even in my limited way. The output needs to be shared more, with more feedback. Updates to Machine Essays will address that, unless I get distracted.

Of course, if 100 people download this, run it, and Skynet spontaneously develops and destroys the human race I apologize ahead of time.

And again, I could be completely off my rocker. But I don't think so. Email me if you want, I have a few insane correspondents, can always use more.

C.K. Haun
August, 2013