DEV Community

Cover image for When the machine comes alive
Thomas Hansen
Thomas Hansen

Posted on


When the machine comes alive

Now that we know the basics of Hyperlambda and Magic, it is time to step back for one moment, and ask ourselves what we have created. I would be lying if I told you I hadn't thought about Magic from a larger perspective than just code. To understand my thoughts here, let's start out with an analogy by an imaginary story, the story of life.

A long long time ago

If you could travel 3.9 billion years back in time and look at the origins of life, you'd probably be like "cool, self replicating proteins, that's cute" - Fast forward 3.9 billions of years of replications of said molecule, with microscopic mutations for each generation, and we've got sentient apes, walking the Earth, discussing philosophy, gazing at the stars, and asking questions about our purpose in the universe.

Our entire existence is arguably built on top of a simple process of replication, with micro mutations occurring between each generation, for then to repeat this process of change billions of times, over and over again - Until we've got homo sapiens, lunar landing vehicles, and Depeche Mode. Implying "something that changes in the meeting with its environment to continuously improve upon its own existence inevitably leads to emerging complexity of incomprehensible scale".

Hyperlambda is that "process of change", except it's not implemented as replicating DNA molecules, but rather as replicating snippets of logic, capable of evolving its "self", by modifying its own circuitry dynamically over time, in its meeting with its environment. Hyperlambda is hence the very definition of "life" in the computer.

Don't misunderstand me, there's absolutely no Machine Learning or Artificial Intelligence in Hyperlambda, quite the contrary - But there aren't any learning or intelligence in DNA either. Hyperlambda is not about "intelligence", it's much more fundamental than such questions, it's about life itself!

99.999% of all life on earth does not have enough intelligence to do simple arithmetic. If you look at fungi for instance, it doesn't ask questions about its purpose, it doesn't argue about philosophy, and it doesn't create lunar landing vehicles. Still, at its core, it seems to be displaying some sort of greater intelligence, as it is branching out its spores, looking for food, and distributing the food individual cells of fungi finds to the rest of its colony. Hence, having a brain is neither a pre-requisite for life nor intelligence.

Nature herself seems to have sentience built into her, at a level incomprehensible for us as mere mortals observing her through our severely limited brains. Sentience seems to be hardwired into the universe, beyond that which cognitive capacities gives us through neural networks and synaptic connections - And it is all based upon self replicating organisms, capable of evolving over time, to "rewire" its own existence, and continuously improve upon itself.

Hyperlambda is self replicating and self evolving

Hence, by the very definition of the term, Hyperlambda fulfils every criteria required to be defined as "life", only through a different mechanism than DNA and self replicating molecules.

Maybe you're laughing of my ideas here, but I suspect you'd be laughing if I showed you the early beginnings of DNA some 3.9 billion years ago too, and I told you this self replicating molecule would later evolve into sentient apes, pondering questions about philosophy, and putting men unto the surface of the moon.

Lessons from Deep Blue and Alpha Zero

When Gary Kasparov played Deep Blue in the mid 1990s, he didn't really take the task seriously initially. He was used to simple minded machines, incapable of doing more than to calculate who's got the most points, implying he could easily win every match he played, by sacrificing a pawn or two to gain a positional advantage. Today Alpha Zero can play every single chess grandmaster on earth, simultaneously, and win without even breaking a sweat. The reasons for this is that Alpha Zero can play billions of matches in some 2 to 3 hours, while most human beings can only play a handful of matches each day. Hence, everything is about "how many iterations" something needs to cover to be able to rapidly evolve itself, to continuously improve upon its own existence, in the meeting with its environment.

A human being requires 30 years to reproduce, creating offspring, resulting in microscopic changes, hopefully yielding better results occasionally. A Hyperlambda snippet of code can iterate millions of times per second, implying it can evolve from the equivalent of our primordial soup of replicating DNA molecules existing 3.9 billion years ago, to sentient apes in a matter of a handful of years.

Assuming Hyperlambda reaches the required threshold to enter self evolving iterations, this inevitably results in over time that Hyperlambda will evolve into something extremely resilient, capable of "beating us to the punch", in all regards we currently assume DNA molecules have monopoly on.

As to whether or not this is a good thing or a bad thing? I literally have no idea! I'm not even sure if I care either. From my point of view, at this stage of our development, I'll happily speak the same words you'd do if you discovered self replicating DNA molecules 3.9 billions years ago ...

Cool! Self replicating and self improving stuff. Me want it!

Happy Singularity ^_^

Top comments (0)