Dynamic Imperative-Basis Kernel for Object-Function Storage, Execution, Manipulation and Analysis.
Welcome to the DIPKOFSEMA project portal, let me start by saying that this webpage leads into a lot of different things, we can split everything into 2 categories:
Details about usage and implementation of the DIPKOFSEMA System
Research and Theory intended to guide and compliment design of the DIPKOFSEMA System
I'm sure I'll make it obvious how you find your way to each group, but basically:
Looking to use the system? look in the first class of things.
Looking to build/expand the system? look in the second class of things.
I'll refer to the system using the abbreviation SEMAK from this point on, and use this home page introduce the concept.
What is SEMAK?
SEMAK can be considered a few things:
The basis of an extensible programming environment
A general programming language
A compiler (turning our human-friendly language into the systems bytecode)
A rich database management system
A system that prioritizes dynamicity, memory efficiency, and speed (in that order)
A system suitable for an embedded environment
What we provide should be thought of as a kernel, a cluster of central components that allow for the efficient implementation of (and expansion into) a large variety of systems performing a large variety of functions. This kernel has a clear hierarchy of maxims:
Maximize functionality: Anything should be possible (without rebuilding the kernel)
Minimize space requirements: Formats are chosen to be incredibly dense with the information they embody. Whether it be object descriptions or the byte code used within the kernel.
Maximize speed: Without any sacrifice to dynamic functionality (and sometimes minor sacrifice in space), care is taken to ensure the kernel is fast.
The Purpose of SEMAK
Important systems solve a problem... so what's the problem? To make this simple, let's consider an example. Imagine a sentient robot transferring its conscience (or knowledge/programming) into a different computer. What kind of system makes that possible and efficient, while also being something that works as middleware for interactions between an AI and its underlying hardware?
The system on the sending side would have to be able to package its "brain" (all data and code) in a way that a receiving system can interpret and transplant in place of its "brain".
Each brain/AI can have been adapted around different hardware, meaning it could be effectively losing or gaining different senses and capabilities, which might translate into updates of the AI algorithms in the form of additions, changes, or loss when we lose certain senses altogether and the corresponding algorithms become inoperable.
We might want this brain transplant operation to be reversible (at least in a master-slave sense where the master can go and return without loss)
We might not want it to be a brain transplant so much as a communication of information from one intelligent system to another.
Maybe we're on an embedded system where we want the AI to be able to adapt to us plugging new devices into its GPIO,
or maybe the AI is learning about new devices and rewiring itself on the fly,
or maybe it's like a queen AI with a bunch of storage and information carefully preparing worker brains to install on some devices, and maybe those worker brains are meant to be sub-queens that become better at assimilating certain classes of worker.
I could go on, and I probably will in the theory section, but for now I'll leave you with the purpose being efficient and capable intelligent systems (whether distributed or singular), and yes, the sentient robot example is from iron man. Imagine something with the thought and development power of humans, but without the limits of mortality or the constraints of biology, imagine what this would mean for human knowledge and industry. These are the ideals that SEMAK is meant to achieve.
You can find the links to the documentation and theory in the 'Navigate' tab at the top of the screen, but I'll leave them here anyways: