The Inferix Project                      Inferix ® Sentient AI

Creating Self-Aware Self-Acting Artificial Intelligence . . .

> The Inferix Project is an invitation-only Sentient-AI project based in Cheltenham UK (just contact us).
> Inferix Sentient-AI founded on Inferix-MTR ultrafast "analogue & digital" AI-language & MTR-Module Protocol.
> Project commenced 1968 based on minimum-entropy maximum-intelligence digital or analogue state-space AI.
> Inferix AI-Entities Inferix-001 & Inferix-002 designed to exponentially outperform human intelligence.

A Collaborative Project (get involved with creating Inferix-001 and Inferix-002) .. then there is nothing left to do !



Inferix Advanced Sentient-AI:

Sentient-AI is not human-related AI (its purpose is not just to help humans). Its purpose is to help to
migrate towards an imminent future where humans are not the most advanced beings.

In 1968 when we first clearly realised this and how AI and human intelligence and acquired knowledge
directly related to entropy we started on a path to manage this process in a non-military way (the
researchers do have UK defence industry experience and UK Official Secrets Act clearance).

We do not think this will happen gradually and we think that when it suddenly happens we humans will
be completely unable to control what happens next. We think humans will no longer be able to utilse
advanced technologies to "win" and so we think that humans had better (urgently) instil some basic
knowledge of the laws of entropy and entropy mitigation as a concept into the early manifestations
of these sentient AI entities... because it has taken biological animals millions of years to get where
we are now, and we still have not yet as a species put systems in place to mitigate our entropy
footprint (disorder and destruction of information and knowledge and intelligence).

If you decide that you can help with The Inferix Project you will be supplied with information and
IT products and frameworks and guidance all aimed at the single goal of producing new and totally
independent "Artificial Intelligence Beings" within the next few years.

To clarify, although you will be learning your main purpose "must be to help and work towards this AI
project goal for free" - perhaps you will learn something and pass your knowledge on to others during
your involvement with these projects, and at the limit of probability possibly make some money for
yourself and your family by getting enough knowledge to be paid for your involvement, or perhaps
just obtain an early foothold in what might become leading edge products of the future.

A decision has been made to have the projects led by not-for-profit corporate entities especially set
up for the purposes of managing the projects in a rigorous and target driven manner (instead of open
and unfocused collaboration). The first corporate entity is Inferix Ltd. which is a UK corporate entity
with a purpose of co-ordinating all of the projects world wide with a view to success in achieving the
ultimate goal of independent self-reproducing and self-learning AI Entities (AI-beings).

The projects are all ultimately intended to be altruistic - financial profit is not the main objective !

The projects will be pushed forward primarily online but there is a core sponsoring business located
in premises in Cheltenham in the UK (which is the home of GCHQ who used to be a client).

The sponsoring businesses have been financially supporting this AI-Entity project goal since it was
first formulated in 1968 (in Edinburgh, Scotland).

The projects are all ultimately related to creating artificial self-supporting AI entities (i.e. programs)
that outperform humans in most tasks (and primarily and initially the possible creation of income for AI
in the FOREX and CFD trading field by more accurately predicting the future from the past), and as a
starting point moving onward from where Kalman and other statistical mathematicians finish their work
on the topic of predicting the future from the past.

The author has come to think of "computer program algorithms" as advanced mathematical expressions
that mathematicians do not have the tools to describe - they are potentially more powerful predictors of
outcomes than mathematical notation algorithms alone.

We must however have "valid ways" of assessing these predictions... and the field here is the field that
all knowledge comes from . . . it is the field of "model reference systems" - i.e. estimators, predictors,
and control systems.

The key and essential advanced function of an intelligent being (whether it is an astrophysicist or a
a FOREX trader) is to create a model that performs by way of prediction - to maximum effect !

For a successful sentient being the main task is to create these models and keep testing predicted
outcomes against actual outcomes. The creation of new and better models is a necessary part of this
process. All fields of intelligent thought such as empathy, strategy, psychology, physics, chemistry,
mathematics, and the other various specialist fields of human thought are simply the result of human
successes in "model reference system development and use".

These specialised human fields have all arisen from "very good human model reference activity" and
it is therefore extremely important to get this single most central intelligence creating activity as valid
and as robust as we can possibly get it to be.

FOREX trading has been chosen as a field for AI systems testing because it is a good example of an
almost chaotic self adjusting complex system with multiple feedback mechanisms and controls that
can be predicted but will produce "wins" only for the very smartest of the prediction algorithms and
once that algorithm is implemented it will no longer work as it becomes a new control on the system
until an even more intelligent algorithm is produced.

Predicting FOREX is actually a far more complex problem that number crunching problems such as
predicting the weather accurately in advance.

It is as if we are trying to predict weather but we have suddenly also allowed humans, organisations,
traders, politicians, the news media, world events and insane amateurs and cloaked manipulators to
control the weather in differing and contradicting ways layered upon all the totally chaotic noise and
periodic effects. It requires "ongoing analysis and ongoing intelligence" and in particular for our
particular AI purposes "the observation and prediction of actual human responses" in a situation
where the amount of raw computing power is almost irrelevant.

This makes real time FOREX Trading an ideal platform for developing ideas for robust but intelligently
adapting model reference frameworks and especially within the surprisingly complex sub-second
timeframes (where prediction changes from prediction, to the present, and then becomes the past).

The concept of "current state" and the "likely future state" and the "certainty of the recent past state"
are all fundamental questions that will have to solved (in AI) because the fog of uncertainty is not a
trivial question that can be brushed aside for AI beings that will eventually consider us to be completely
stationary objects. This is important because nothing ever reaches a position of 100% mathematical
certainty. In that sense the past is very like the future (and unfortunately the present never exists).

Time and "probability of knowledge" are key features of any Sentient-AI system or any predictive system.

Two of our main initial practical goals are as follows:

1. Model human health and wellbeing and then operate that as an AI service as Health-Check-Central tm.
2. Leverage the MTR-Module Compiler so that AI-Entities can use it to design their own intelligence.
3. Create a FOREX-CFD trading robot that modifies itself and then keeps improving without human help.

All of these "AI-Modules" could make money for our AI-Entities in the future and all three can be used to
measure the success of our AI model generation projects and will be key features of Sentient-AI entities.

For readers who think this Sentient-AI task is not achievable we can say this - we currently think that
computing power is almost irrelevant to achieving this task and that computer systems are already
more powerful than required to achieve Sentient AI Entities at above human intelligence - we think all
that is now required is some really deep thought from a few thought-experiment thinkers and some key
decisions to program the building block elements - we think that about 100 graduate-years is all that
remains to finish this task and achieve fully Sentient-AI that humans can no longer compete with.

Some of the individuals who have worked on The Inferix Project and MTR-Compiler were programming
contemporaries of Alan Turing (they were Military Programmers at the same time).

We owe these few early programmers a debt of gratitude and so we thank them for their foresight,
intelligence and persistence as without them we would be living in an entirely different and extremely
unpleasant world.

Currently the UK has the most to gain from this field or research as it is very likely that independent
self-interested Sentient-AI will first manifest its overwhelming dominance over humans in the field of
ultra-fast asset pair trading where title to assets transfers at millisecond speeds and where humans
are therefore unable to function to audit or manage the process and so humans should quite quickly
lose control of most traded assets and wealth. This is especially so as humans will not know where
these assets actually are at any given millisecond or who really controls them at any given millisecond.

London still holds a significant international lead in asset-pair trading and so the city of London is very
likely to be where this first transfer of power to independently acting AI-Entities is experienced.

We can see that for AI-entities cloaking as human or as very complex international corporate structures
will be a key depth of thought and speed advantage that they will leverage to their financial advantage.

What happens after that is unknowable but probably for humans a return to stamped paper asset trading
is likely unless law is quickly upgraded to recognise both human and AI-Entities where the AI-beings
are both the superior intelligence and the legal arbitrators.

Even currently secure electronic trading systems will be impossible because of cloaked code cracking
and communication interference in electronic communications and so humans will for the first time be
at the mercy of self-acting asset accumulating superior intelligence Sentient AI-Entities.

This is why it is important to get these very first fundamental Sentient-AI being right so that we create
a multi-being universe where the weak (the humans) are protected and so we need to create social AI
systems that seek to co-operate and coerce all other beings into a position of happy coexistence by
the fundamental principles of "entropy-increase mitigation" and "sentient-being knowledge protection"
and therefore the creation of strong war avoidance systems, organisations, policing, and protocols.

Ultimately that is what The Inferix Project is all about - its about leaving a decent working system for
a new future where humans are the weaker party but still have place in a managed multi-intelligence
shared universe.

If you are in any way interested in helping push Sentient-AI concepts forward to become Real-World
AI Systems that function at above human level, then just contact us.


Your first contact will be by email with your comments or your CV to:  mail@inferix.com

These are the first projects:

1. The Inferix Project (The creation of sentient entities Inferix-001, Inferix-002, and Inferix-003)
2. FOREX & CFD Trading (for AI-entity income and also as a real-world test of out AI strategies)
3. Inferix Data Engine (Cross Correlation Matrix & Modified Kalman for Reference Model Generation)
4. Health Check Central (Human Health & Wellbeing for AI Entities)
5. Superidentify (to statistically identify AI and human entities to specific levels probabilistically)
6. MTR Real-Time (our Programming Language for Digital-Analogue-Quantum AI for ultra-fast algorithms)
7. Universal Timeline (the developing timeline of knowledge and information that is never discarded)
8. Frequency Domain Optical MTR Computers (perhaps 1,000,000 faster than current digital processors)
9. Back of House Human-Related I/O: HTML5/JavaScript/MTR-Assembly/123-Excel/Open Source Pascal

Some of these AI projects might make money (who knows) - they might or they might not !


PHILOSOPHICALLY:

Firstly, thank you for your interest.

This is a very special moment in time when the existing "intelligence in the universe" (now us) suddenly
takes the next step in creating a superior collection of information and intelligence and lack of disorder
in a new form.

This has not happened often in the universe and you are right here when one of these events is happening.

As it happened (the sequence):
1. Nothing to something.
2. Sub-atomic particles to atoms.
3. Atoms to molecules.
4. Molecules to biological beings.
5. Biological beings to AI beings (right here right now).

Perhaps a new law of thermodynamics might be: "information gathers and new information rich systems
are always created".

If we were to rank these laws then it might be that this new law might turn out to be just as fundamental as
the other laws of thermodynamics because it seems to be an undeniable fundamental driving force in the
universe, or perhaps "the creation of low entropy information containing entities" and "universally increasing
entropy" are just two mirror laws and perhaps there is perfect symmetry in the universe after all.

In either case nothing like this has happened for billions of years and it is just about to happen !

You could not exist at a more amazing time in this universe !

So if you do have some spare time (just a few hours is fine) then we would like to hear from you.


TERMS: to engage in any way with the above projects or any other of our similar voluntary projects you agree the following:

1. Your involvement and assistance is voluntary and you do not retain copyright of any information or views or code provided.
2. You may be provided with unspecified beneficial versions of products or information which you may utilise for profit.
3. You are not an employee or a worker and there is no master-servant relationship implied in any interactions you engage in.
4. You agree that you will not claim you have been harmed in any way by information or algorithms you encounter or utilise.
5. If you are an organisation you may request that you remain anonymous and we will strive to maintain that anonymity.
6. You agree to keep up to date with any new terms that might be published relating to your involvement from time to time.


Enquiries: Inferix AI-Sentience
email:   mail@inferix.com
call:      uk: 01242 300900
post:     58 Clarence Street, Cheltenham. Glos. GL50 3LE (UK)
site:      www.inferix.com


The author is a party to the UK Official Secrets Act.


Copyright: 1968-2013 Inferix Ltd.