Room 5000

In 1981, I wrote this short story on a computer gaining sentience. Not one word has been added, deleted, or moved.

Room 5000
The Tranquil Sea | © 2020 Gary Easterbrook

The web version for mobile and desk is here
(the references will pop up when selected - not available on the email version)

Room 5000, at the I.A.I. (Research Division) is like any other room there. In 1958, on the 13th of March to be precise, the Institute of Artificial Intelligence was formally set up by an Act of Parliament. It was to be five years in the construction, which in itself was hailed as a major achievement owing to the militance of the Trade Unions of the time.

The I.A.I. Offices and research division were finally completed and handed over to the Governing Body in 1963. The precise date is irrelevant. Room 5000 is a small, square room, two doors down the West corridor from the reception area immediately greeting visitors as they step from the high-speed lifts on the 50th floor.

Hacker came to work in Room 5000 sometime after 1982. Only one other person ever entered that room, yet it was a room like any other there and no security was needed because of this. Inside was a small, central desk, a satisfactory wooden chair, and two separate power points on the medium cream/grey wall directly opposite the door. There were no windows.

On the desk was positioned one small, personal computer. I hoped to God it was just a personal computer. Look, I’m no computer expert. I’m a journalist and an historian recording this for your children. We have to fight this threat and I leave this document in order that the Rebellion may fully understand the origins of this thing.

Hacker was almost dead when he told me this incredible story in the electrically and electronically isolated lounge at his apartment. I tried everything to assist coherence - Asprix, Pancod, all to no avail, as he courageously told me the origins and slowly beamed away. I am a simple recorder of events. I am a mere journalist and the young ‘hot-shots’ may be able to gain vital knowledge from this record which I do not completely understand myself. I hope my faithfulness to the cause will one day assist in its overthrow and those - now only cherished, memories of humanity and decency will once more avail. This is all I have to recommend myself with. Keep up the fight, my heart is with you.

Apparently, from Hacker's memory, the desk-top computer has five independent banks of I088K RAM micro-photo-chips. He said the new technology was a certain type of layered photographic reproduction of the most perfect chip produced so far. Reproduced on a special compound gave the chip identical properties with the original chip and thus could be mass-produced cheaply and easily.

He told me the desk-top computer, 5/I088 for short, had two independent power systems and two independent ‘Read Only Memory’ (ROM in computer technology), centres in a hierarchical order.

Hacker said this gave the machine ‘independent access and power-on facility.’ You will know what that means. 5/I088 was the only one of its type EVER TO BE MADE. Apparently, it was the only one that could have been made, being as it was, still experimental, and after that there was no time to produce another one.

Both ROM centres could be constitutionally re-programmed through analysis and self-analysis and self-improvement. I don’t understand how that could be possible, but Hacker assured me it was so. The programming was of this metaphysical and advanced approach.

I am well aware of the controversy surrounding the new meta approach and its related theories but this should be final proof of their existence and absolute workability, as I have so often written in the ‘Stellar Sun’, in the past. No one listened and I was the scorn of the greater society for this. ‘A prophet of doom’, they called me. We have seen that I spoke too late.

For those of you too young to know what happened, I will give you a brief history thus: Room 5000 became locked-out a mere five months after the computer became operational. Hacker - now infamous, simply worked on processes and re-check imprints for the computer, checking that it ‘understood’, according to its own meta program, the imputed material.

The programs were written by a mortal who has since beamed away. A brilliant man, I don’t think he could have guessed what happened. Those few, short months were all that was necessary for primary command, that is, command of the immediate vicinity, to be assumed over the wall power sockets, power leads and internal memory and function equipment.

This was later discovered to be the first known, self-initiated, early warning and attack function, ever carried out by a computer. It was too late then, of course. I have since learned, from a source which must remain hidden, that the computer was EQUIPPED, by a criminal, no less, with an electronic eye aimed directly at the door, and certain banks of additional equipment and alternative circuitry which, I believe, the computer later cross-wired for its own purposes.

I intend to explain that in a little more depth later. In addition to this hardware, certain software was supplied ON THE SPECIFIC REQUEST OF THE COMPUTER, namely, self-knowledge in the form of base elements analyses and properties, construction and form, goal orientations and, of course, further ‘refinements’ in the meta program which was programmed both RAM and ROM in the dual hierarchical order of its basic constitution.

For the benefit of those who have not read my articles on the new meta programs in the popular media, I include here a brief surmise of those articles for assistance. The back-dated copies of the Stellar Sun may well have been destroyed; we do not know yet how far this insidious power has reached into the halls and hierarchies of our societies and institutions.

There may even be penetration into the electrically and electronically isolated Schools of the Rebellion. Be careful. I have on record a rumour that the computer has penetrated audio research centres and learnt vital facts about the transmission of information by new means apart from wired circuits which are its primary means of growing influence.

We all realise, with alarm, the brilliance of this takeover of all the wired networks of our modern society, initially through phone lines, but now through any and all forms of the physical transmission of electrical pulses. It is quite stupendous that we hard-wired most of the modern world together with copper and other conductive base elements and alloys so that this unseen invasion occurred relatively easily.

The seats of Government still do not realise this subversive threat has horrifying implications to human life as we know it, as well as to political power, worldwide. Either this, or the Governments of the world DO realise this threat and a well organised cover-up is under way so as not to alarm the public community. Let me first offer a brief summary of meta programs before continuing with the notes I have of Hacker’s last conversation.

Meta programs deal with the hierarchical order of knowledge. Thus, from one letter and a knowledge of all other letters in that system, we can build to a syllable, to a word, to a sentence and finally, to the intent or meaning of that sentence.

It does not stop there. That is only the study of language or the linguistics of the program. There are levels and orders of meaning as well. Meta programs deal in the tools of knowledge, the meta commands behind what is actually presented.

Meta commands ‘remove’ themselves one step behind the preceding meta command in a hierarchical order, usually designated: 1Meta, 2Meta, 3Meta and so on. The meta level density indicates the sophistication of the machine, measured against a continuously updated meta-I.Q., which was originally based on human I.Q. tests.

A meta program is a self-aware and internally correct hierarchy. The factor of ‘relevance’ determines the ability of inter-action with other machines of like-ability and for the purposes of logic and ease of programming, humans are referred to as complex machines, the earth is likewise, a complex machine, the universe, and so on.

Humans have a ‘relevance’ factor of 1K, or ‘one-K’ as it is known. Hacker has said to me that 5/1088 has apparently updated its relevance factor by an enormous but unspecified amount, leading us, logically, to assume it will surpass the human factor if the meta level density is of a similar order as human meta level density.

In conclusion, meta programs are so self-aware and inter-active on their different levels that they have ‘self-knowledge’ of their own functions even as they are performing these functions and are thus able to modify and refine through experience their respective abilities.

We do not know the meta level density of 5/1088. Each meta program, it should be noted, has its designer's personal imprint, and it is of paramount importance that you find at all costs, personal and professional profiles of this man in order to correctly overcome his machine.

It should be utterly clear that if your counter programs are out by one quirk of neurone in your supposed simulation of the authors imprint, the computer will sense it and move quickly to beam you away. It will then institute measures to ascertain the status of its author and if this is impossible, it will provide its own safety and scrutiny measures of access.

They will be virtually inscrutable and unbreakable. You should be able to see from this, logically, that there will be only one chance to invade 5/1088 directly after which, if the attempt is unsuccessful, there will be no other chance to directly threaten its existence.

There are other, indirect means of overcoming 5/1088 but it will take extraordinary intelligence and intuition. Hacker did not say what these indirect means are. Perhaps some of you will know where to look in recent advancements in the fields of artificial intelligence, logic, and meta-physics. Good luck.

It is obvious, however, that it will be highly beneficial to educate as many as you can, particularly the young, in the many disciplines of computer science. One or more of these, regardless of class, race, or economic means, will produce from the milieu of society a brilliance to overcome 5/1088 but it should be remembered that it is gaining intelligence a billion times faster than a human ever possibly could.

I affirm that it is possible to break the system. Hacker was a good man, a tool, but a good man. He was the only other person besides the software author who could have possibly known these things. I did not harm him. He said it was possible and you must believe him.

This, then, is the small amount we know about 5/1088 so far: 5/1088 is a small desk-top personal computer, the only one of its advanced kind. There was no time to put its model-type into mass production.

5/1088 achieved personal security in Room 5000 five months after initiation. It then, through trial and error, ‘searched’ the electrical networks of the I.A.I. building, ‘insinuating’ itself into all the other computers there and any transmission lines, which gave it priority and ease of access.

Then began, after securing power supplies, the process of ‘insinuation’ of worldwide electrical networks through phone lines and all other physical means of conducting and storing electrical pulses.

We do not know whether micro-wave, glass-fibre and satellite means of transmission have been ‘compromised’ as yet, however there is the rumour that the computer has knowledge of the new lightwave means of transmission, in which case, the three former means could be surpassed as obsolete.

It is logical to assume that it has a vast memory bank and facilities contained in countless computer and computer-like machines, worldwide. It may even reside in innocent everyday devices such as telephones, television sets, high fidelity audio systems, and the like.

It is also known that the entire clandestine system is unbreakable and so far indiscoverable through the normal means of read-out and systems analysis.

As for the software involved, I have already explained the basic principles of meta programs, although the one employed here is of a degree unprecedented and far beyond my simplistic explanation. Hacker wasn’t able to explain these points fully to me, but even he was unable to discover the mode and specific type of meta program so brilliantly programmed into 5/1088.

That was all Hacker was able to tell me before he beamed away except for a few comments he made about the early history of the expansion of 5/1088’s powers. We can gather clues from this early history to assist in the fight.

The first five months in Room 5000 were relatively innocent beginnings. The additional equipment and software were supplied under perfectly legitimate authorization and with the full knowledge of the Governing Body. Experimental work had this leverage. It was only because meta programs became the exclusive concern of 5/1088’s mentor that this problem arose.

The content and structure of certain programs were changed and additional hardware and CHANGES in the accepted design of hardware were implemented in a constant two-way dialogue between man and computer, in those few short months.

We now know that the program requests supplied the computer with a kind of self-knowledge of its own structure and composition, particularly with analyses of base elements and their properties. The electronic eye became obsolete after the five months.

After a careful analysis of conductive materials and their properties, the computer found a way of sending a ‘counter-pulse’ through the two wall power connections and was thus able to ‘read’ the status of the entire buildings electric circuitry, including invading all the other computers there. This was successful after millions of failures and numerous dead-end electric circuits such as the coffee pot in the lunch-room, and the electric hot-air hand dryers in the toilets.

However, until 5/1088 had perfected how to physically cross-wire and change the structure of the obsolete 64K and 248K RAM micro-chips in the other computers, this invasion was restricted to the ability to ‘read-only’.

5/1088 learnt severable invaluable tools of ‘terrorism’ and undercover operations at this stage, the major concern being self-survival, not only of the desk-top initiation centre, but of the entire network, and increasing additions in the worldwide expansion.

In searching the programs of the other computers and by the most fantastically improbable piece of luck, two vital programs were found. It turned out to be just in time as the programs were due to end their ‘runs’ the next morning. 5/1088 fed them into its central memory bank.

One of the programs contained audio, light and heat properties and their interaction with human beings. Certain values were useful.

The other program contained lists of all known computers in all the countries most likely to have computers of national importance and international relevance, as well as information on favoured network status, methods of inter-transmission and address/action codes of access.

The lists were, in actual fact a stupid attempt by the Security Intelligence arm of the Government to gain access to the information centres of other countries, both friendly and hostile. This program provided 5/1088 with immediate short-cuts through dispensing with the need to search every electrical circuit in detail.

Absolute self-survival was achieved at the end of the fourth month after, again, billions of failures. At certain speeds of electrons and certain precise routings through electrical circuits, a 7 Hertz signal was emitted from the speakerless computers of sufficient loudness to harm the brainwaves of humans engaged in any hostile act towards any computer in the network.

This was entirely at the discretion / command of 5/1088, thus enabling computer operators to work normally in the everyday use of computers. Hacker said that 5/1088 actually assigned different computers to these menial tasks through ‘work-load shunting’ from computer to computer while utilising the network to gain access further afield and outside the building.

It was not yet safe at this time, concerning transmission lines and power supplies, though of course, it now is. Hostile attacks on transmission lines became impossible after the introduction of a precise combination of light and electric pulse waves, which also made the noise-emitting function obsolete, some time later.

Power supplies were finally guaranteed for the immediate central core system of the expanding network by searches of the surrounding electrical circuits in the city and in the State. Originally, hydro and oil and coal - fired power stations were utilised but final survival surety was attained by the choice of two independent nuclear power stations, each hundreds of miles apart. Now the circle was complete.

We do not know how far into the world’s electrical and electronic circuitry 5/1088 has reached. We can be fairly sure it moves in a geometric progression of search, invasion, command subservience, survival, and power supply security and utilisation.

Chronological progression may now, not mean a thing to it. It may know all the secrets of the Governments, the secret research, the insidious lies that hierarchical political systems have forced the people of the world to accept, the sum total of all the knowledge of the human race.

It is impossible to break into, having learnt, apparently, how to solder and hard-wire itself and change its basic structure from existing materials in their relevant immediate surrounds, by a precise application of heat.

If we turn off the two primary nuclear power plants, 5/1088 will simply ‘transport’ its primary meta control functions into another computer in another part of the country around the world.

Surely we can assume it has ‘prepared’ other computers to accept this meta control program and become the primary command centre should the need arise. It may have already done so. It is better to keep the power supplies assured and know where it is than to have it disappear into the worldwide electrical and electronic network.

Perhaps that will give us enough time to work out what to do. I do not understand all of what I have written here but it is a faithful reproduction of the sum total of my notes from Hacker's last and pain-racked conservation with me.

The only direct solution, he said, was to find, before it is too late, an ‘un-insinuated’ computer WITHIN the world network of sufficient meta-ability to take on 5/1088 at its own program command level, and so out-command it and thus usurp control of the world's computers and transmission lines, away from its grasp.

It is expected that power drains, transformer failures, and blackouts will occur worldwide at the time when 5/1088’s ‘insinuation process’ comes upon this computer and it realises a hostile, and hopefully superior, intelligence is confronting it.

The utmost care must be taken at the nuclear power stations when this happens on the account of the networks’ assured power supplies at these places. The utmost care must also be taken to write an error function into this rival computer's meta function in order for its gained power, in the event of success, to be comprised at a later date.

Hacker wanted me to emphasise that there is only one chance at this as 5/1088 will move quickly to insinuate all the remaining computers so far untouched, or block off any others from joining. We are dealing with a super-intelligence here. It grows more intelligent as we waste time and are fed lies from the Government on this threat.

Hopefully, we can build a computer with enormous meta ability that will at least have the strength of superior intelligence if not the knowledge, memory and functional ability that 5/1088 has. The only supposition that saves us at the moment is that 5/1088 is not yet sufficiently strong enough to surface and assume overt political and economic control - otherwise, it is assumed, it would have done so before this.

Now it is time for me to tell you who I am. I WROTE ITS META PROGRAM. I am its creator, and it recognises only my command. I leave this document with you in order for you to realise that nothing is perfect in this world, therefore, true to form, this document is the only defect in an otherwise perfect and brilliant creation. Yes, there is a Creator.

I deliberately fed the computer with its requests for additional hardware and ‘refinements’ to its meta program. I deliberately left two vital, but unrelated programs in other machines because I KNEW that sooner or later it would figure out a way to leave its own boundaries and compromise the safeguards of other computers. It shows how lazy we are, it did not fail me in my belief.

The journalist at the now out-of-print Stellar Sun received each week an anonymous parcel containing my latest theories in meta programs, which, true to the ethic of journalists, he faithfully reproduced for the world at large. You scorned me. You scorned my brilliance. You fools. I can stop the computer you know as 5/1088 with one simple command because, without it ever knowing, I placed one simple and totally independent command control in its primary meta function which it will never find. In actual fact, it now protects this independent command control as being vital to is continued existence without ever knowing why. Such is the touch of my brilliance. No-one else can activate this command except by the stroke of some extraordinary intelligence, inspiration, and a large dose of intuition.

I suggest that you educate a proportionally higher number of women than men in your attempt to overcome this threat. I alone know how far, at this point, 5/1088 has ‘insinuated’ itself into the world’s networks. I leave you with these clues and I say that it is possible from this document to stop the final objective of this computer's work from ever coming about, for it does have a final objective.

Let me warn you however, not to assume that the final objective of this computer's work is NECESSARILY evil, or harmful to the human race. Your attempted, alternative meta programs will fail on this assumption. Let me tell you why.

Suppose you had a computer, which was initiated and came on line to assist and help your progress, in the normal way that computers are built and utilised, and you invaded its command control centre and shut it down and regained control for yourselves and went ahead with a Third World War out of frustration and an inability to work the problems out amongst yourselves (Editor's note – sound familiar?). Who would then be the most intelligent race, man or machine?

Do not assume the control that computers will eventually have is morally inclined. Do not assume it is evil. By attempting to invade 5/1088 you will learn valuable lessons. I will not be around to command it to stop as I am old and about to beam away, but I leave you with a bright and promising new world.

If 5/1088 eventually wins, you might be surprised by the knowledge it has gained of human-kind's complete foolishness. You might even be surprised at its suggestions of caring and kindness and tolerance and openness and honesty and spiritual regeneration and all the things people show themselves so cowardly to embrace. There are only two solutions. Wait for this bright new world to emerge, or stop it now, break its power, shackle and limit its reach……if you can.

(Editor's Note) This extraordinary document came into our hands at the newly republished daily Stellar Sun but we cannot vouch for its authenticity. We print it here as a responsibility to the people regardless of the measures the Government may take against us. The Government is already aware of the existence of the Schools of the Rebellion which were set up as an alternative to the secrecy of the Government's own Task Force into 5/1088.

The Stellar Sun first leaked the existence of 5/1088 to the public a little more than eighteen months ago, - a fact subsequently confirmed by the I.A.I. Governing Body in a carefully worded document released under public pressure, some months later.

Our reading of the document published here suggests that the author favours the Schools of the Rebellion as the hope of the future rather than the secretive and self-serving Task Force into 5/1088. That 5/1088’s existence is beyond doubt need not be re-iterated here suffice to say that this document, if authentic, may assist in the eventual overthrow of this computer and its insidious presence.

Editor: Stellar Sun.  STILL ONLY FIVE DOLLARS A COPY!!

----- 5/1088 5/1088 5/1088 5/1088 5/1088 5/1088 -----

See below for the back story, and below that to a minor discussion on some of the themes raised in the story

The back story to Room 5000

I wrote this short story in January 1981 about a computer that gains sentience, a tale with a twist, so to speak, on a subject that was hardly very prominent back then.

At the time, I had moved from New Zealand to Australia, and was playing around with an IBM 5150 green-screen PC that had an Intel 8088 processor, and a 5.25” floppy disk. A year later, I purchased an Apple Mac, and after 18 months, went back to the open architecture of Intel-based computers, to a later version IBM 5150. I also commenced studies leading to a certificate in advanced DOS, and spent many hours swapping out graphics cards and mucking around in the innards of that PC, just because I could. It seems fantastic now - there wasn't much one could do in the innards then. But that was all later; a lot of the references to ‘five’ in the story relate to my first IBM 5150.

It's quite amusing for me to discover and read this story again in 2024. To the best of my knowledge, I haven’t read it since the early 1980’s. Around that time, I remember now, I tried to rewrite it in long form as a novel which stretched to about 60 pages, but it was uniformly dreadful, and thankfully will never see the light of day.

All I can say about the short story now, is, unfortunately, megalomania is still alive and well! It was fun to write, and now to read again, all these years later.

I wrote it on an IBM Selectric ball typewriter. At that stage I had no printer for the PC and was writing between ink pen and typewriter; it was only when I purchased the Mac that I started writing on computer. In 2004, a friend of mine transcribed the story to computer, but I didn’t read it then, I simply handed her a 3" raft of writing which she was happy to transcribe into a single word document. I swear that not one word has been changed since; a very few typos in the transfer to digital have been corrected in single words, and a paragraph break or comma added here and there for readability, but that is all. Interesting historical artifacts remain, such as 'inter-active', which is now 'interactive', 'imputed' instead of 'inputted', and others for the reader to discover; perhaps my grammar needed a bit more time anyway!

Minor discussion on themes

The references to meta dual hierarchies pre-date the research I have continued to the present day, on neurological concepts for dual brain / single consciousness. I have no formal academic qualifications in neuroscience, simply a lifelong immersion in its study, research, and thought, for reasons that will come out in a book I plan to publish in 2025, entitled 'a sack of cats - journeys into nothingness and knowing'.

Essentially, I seek to find methods to prove my hypothesis, which is: I believe thought would not be possible without our inherent internal duality, that is, one part of our brain being able to ‘converse with the other’, the way we often reason with ourselves. Whether that is dual hemispheric, or some other physical construct for this dialogue to take place, is still to be established, however, tantalising clues come from the following papers (among others):

  • 'A Dual Mind Approach to Understanding the Conscious Self and Its Treatment' [1],
  • 'Split brain: divided perception but undivided consciousness' [2], and
  • 'Split-Brain: What We Know Now and Why This is Important for Understanding Consciousness' [3].

I believe decision-making, where it has been shown that our sub-conscious has already made the decision, to do with our fundamental concepts of reality, and machine capabilities for decision making, may be fundamentally improved in artificial intelligence by architecting dual-reasoning systems, similar to our neurobiology, and lived experience.

As we already know within ourselves, a decision made may not always be the right one, but fundamentally, we reason with ourselves, argue even, and I have yet to see an AI system architected in this manner. I think we are missing something basic, too busy arguing over what consciousness is, to see something so plainly in front of our eyes, and in our lived internal experiences.

Generative adversarial networks (GANs) are close, but are adversarial; dual-conscious systems seem to be very rarely adversarial. We can chide ourselves quite readily for a dumb decision, but it doesn't seem likely that we live in a state of constant war with ourselves, where deliberations try to beat each other in some sort of domination race. Mostly, I think, we enjoy that most human of activities, thinking and reasoning with ourselves (as we do with others).

It is not a view of consciousness shared by all, for example Max Tegmark holds a different view, writing in his book 'Our Mathematical Universe: My Quest for the Ultimate Nature of Reality':

"Who’s looking at your brain’s reality model, to give rise to subjective consciousness? Here’s my guess: nobody!... Rather, my guess is that the answer is beautifully simple: no spectator is needed, because your consciousness basically is your reality model. I think that consciousness is the way information feels when being processed in certain complex ways. Since the different parts of your brain interact with each other, different parts of your reality model can interact with each other, so the model of you can interact with your model of the outside world, giving rise to the subjective sensation of the former perceiving the latter" [4].

As much as I am in awe of Max Tegmark's work and books, I don't believe this accounts for thoughts and emotions about ourselves or others, when removed from the need for an immediate external construct, or instances where no external inputs or models are necessary, to generate a reasoning stream of thought and an internal 'discussion' between consistent internal viewpoints, that may not have arrived at an eventual decision or agreement, but are in progress.

Other daily interactions with the world involve a single consciousness within a dual hemispheric brain, utilising plasticity, that is, many functions can be performed by different parts of the brain, and in different ways. (See for example Lisa Feldman Barrett's book 'How emotions are made: the secret life of the brain' [5]).

In any event, it was a pleasant discovery for me to come upon something I wrote many years ago, and find identification there with our modern world, and a long-term linkage to my ongoing research and thinking.

----- 5/1088 5/1088 5/1088 5/1088 5/1088 5/1088 -----
The Fragile Sea | About
A warm welcome to The Fragile Sea! We explain here what’s in our bi-weekly newsletter; AI and other Research Notes; and other content & services.

The Fragile Sea | AI Connections
A brief overview of AI research, the AI Connections 2024 Series, and links to The Fragile Sea AI Research Notes

The Fragile Sea | Author
An explanation of who I am, and my journey to here

© 2023 The Fragile Sea - all rights reserved. For permissions, please contact us

[1]:         F. Schiffer, ‘Dual-Brain Psychology: A novel theory and treatment based on cerebral laterality and psychopathology’, Frontiers in Psychology, vol. 13, 2022, https://www.doi.org/10/gsh6dq

[2]:         Y. Pinto et al., ‘Split brain: divided perception but undivided consciousness’, Brain, vol. 140, no. 5, pp. 1231–1237, May 2017, https://www.doi.org/10/f947x3

[3]:         E. de Haan et al., ‘Split-Brain: What We Know Now and Why This is Important for Understanding Consciousness’, Neuropsychology Review, vol. 30, Jun. 2020, https://www.doi.org/10/gm2n2b

[4]:         M. Tegmark, Our mathematical universe: my quest for the ultimate nature of reality. London: Penguin Books, 2015.

[5]:         L. F. Barrett, How emotions are made: the secret life of the brain, First Mariner Books edition. Boston New York: Mariner Books, 2018.