My friend Paul was one of Dick's new research students. We had been undergraduates in the same intake year, although we knew each other only slightly then. He was a tall and exceptionally skinny young man, permanently stooped to avoid doorway lintels. He had a large bony and angular face mostly concealed by curly black hair and a full beard, giving him the appearance of the Tallest Dwarf in Show Business.

Paul, like me, had elected to stay on at the University to start on a PhD, although he had taken up one of the bursaries that Dick's newly-minted academic stature had created. After Finals, most of our peer group had left the University, and he and I had drifted closer. Fairly frequently, when the desire for a little company was upon me, I would call at Dick's laboratory to invite Paul for an after-work drink in one of the back-streets pubs nearby.

On these evenings, I would knock on the heavy steel firedoors, invariably firmly secured by that keypad combination lock whose code was rumoured to be changed daily. I would be required to wait a few moments, and then the door was opened by one or another of Dick's Research Associates. I was never allowed inside the suite of rooms, but I could glimpse diligent work being carried out within.

An array of workstations and graphics displays was laid out on desks in the outer office while the black cabinets of the supercomputer itself could just be glimpsed through the glass doors and partition beyond. As I watched, Paul logged himself off from one of the workstations and gathered his few belongings ready for the short walk to the public house.

It is not my field of speciality, but from my limited understanding - an understanding mostly gained over a couple of pints of beer, it has to be said - Dick's modelling process required three steps. First, a standard model for one particular intellectual area was loaded into the supercomputer. Typically, this would be a model developed by other researchers, and generally accepted as a reasonably accurate representation by others in that research domain - at least judged by the number and quality of papers published.

Academic endeavours around the world have produced a large number of computer models developed for specific fields. To get copies of the programs, a certain amount of negotiation must have been required. I assumed that Dick's charm must have come to the fore and he seemed to be well able to persuade groups in other Universities, especially those in other parts of Europe, to provide copies of their software.

Once the model files were acquired, there was an extensive and very manually-intensive process of matching of each of the variables in the model against the terms in a hierarchically-structured ontological database: that is, a standardised representation of concepts or "things". This kind of intellectual drudgery is of course exactly the reason why professional academic researchers engage so many research students.

The second step involved a heuristic process whose details I do not really comprehend, but seem to be some kind of genetic algorithm, to allow the machine to evolve a simplified model - in the sense of using far fewer machine resources - which nevertheless behaved in the same way as the original one. This compressed model would have identical inputs and outputs for all cases tested, and could therefore be said to represent the same system. Of course this optimisation took a vast amount of time on the computer, but after a few tens of thousands of iterations, a stable simplified model would usually have been created.

These first two steps were repeated for models for each of these different areas. From Paul's account, Dick had managed to acquire integrated models for global climate and weather, for the availability and consumption of natural and physical resources as diverse as oil and arable land, for international economic and monetary systems, and for the dynamics of human population growth. All in all, it was enough to represent nearly all of the myriad global interactions that human beings make on our planet, and vice versa.

The final step was to load all of the Heuristically Optimised models into the supercomputer simultaneously, and allow the Semantic Matcher to combine them in a single matrix, based on the mappings to the standard ontology from the database. This integrated matrix embodied a holistic representation of the combined space, a hugely complex model which might well have been impossible to assemble without the compression that Dick's approach employed, even on the world's largest supercomputers.

Now, the well-known computer dictum of "Garbage In, Garbage Out" may well be applied here. The approach required so many bizarre transformations far away from any kind of human scrutiny that it is compelling to conclude the outputs from such a composite model would be rubbish regardless of input. On the other hand, it was certainly the only practical way of getting such a complex composite to produce any kind of comprehensible output, at least with the kind of computational resources continually available to a minor British University research group.

Dick and his merry men - he did not seem to attract female research students - worked away for many months, largely separate from the rest of the School. This met with numerous snide and disparaging remarks, and a high level of general distrust. Even so, there were academic papers written and accepted for publication, which were grudgingly accepted as a useful contribution to the Research Score: the total weight of paper produced by the School and therefore the academic quality rating of the University as a whole.

The published papers following the usual sequence: first, a positioning paper, setting out the objectives and approach, and outlining the techniques in broad theoretical terms in order to establish the earliest possible claim of intellectual precedence and prior art. This seminal work was followed by an irregular sequence of research monographs, conference submissions and the occasional refereed journal paper, detailing specific parts of the whole process in mathematical and often excruciating detail.

Each paper had, as is typical in the outputs of research institutions, a long list of authors, always including Doctor Dick's name - near the head of the list, of course - together with a variable subset of his team. There was nothing truly outstanding in each publication, Paul acknowledged wryly, but each was enough to "advance the sum of human knowledge" as is required by the submission guidelines for academic publications.

Often, a grandiose academic undertaking like this either collapses, unable to deliver on the aspirational results, or else morphs into something almost unrecognisable as some newly-uncovered and exciting aspect expands to consume the entire budget. So it was unusual to hear from my friend Paul that the original objectives were indeed being pursued with very little deviation, and that they were on-track to deliver a successful combined model, although he confessed that one would probably not have realised this from the materials published so far.

I caught up with Paul again just a few days ago. He was looking worried and, unusually, it was he who suggested a quiet late-night pint. I had been working late, struggling with a research problem of my own, and rapidly agreed to meet him in the pub later on. I arrived first, bought the usual drinks for the both of us and I sat for a few moments sipping my pint. Paul arrived shortly afterwards, ducking under the door and looking around the bar in what struck me as a furtive fashion until he spotted me sitting in a quiet corner of the Snug.

He hurried over and slumped into the seat next to me, still looking around the otherwise unoccupied room as if expecting to see someone following him. He grasped his glass as if his life depended upon it, and drank of about half of his pint in one go - most unusual for a serious young man who rarely consumed more than two pints in the entire evening.

Speaking so quietly I found it hard to hear clearly, Paul told me about the state of development of the composite model - he referred to it as a Sustainability Matrix - and that Doctor Dick expected to publish a preliminary results paper very soon. He then leaned forward over his pint and lowered his voice even further.

All variants, Paul announced, of the matrix now executing on Dick's Tool demonstrated an inevitable decline in both global economy and human population - so rapid, he said, that the term "collapse of civilisation" was not too dramatic a phrase. I stared at him for a second, at first not sure whether I had heard him properly, and then for a longer moment while the import of his pronouncement sunk in.

"You mean it's inevitable?" I asked, aghast.

Paul nodded his head, then took another long pull from his drink.

I asked him whether the matrix, the simulation was telling the truth. He looked so serious for a moment.

"I think it is," he replied morosely, "I simply can't see any alternative"

He explained that Dick's researchers were trying more variations - the machine was running at full capacity all the time now - and there was a continued investigation to demonstrate coherence with the original models by substituting in turn each of the original full-scale models for the heuristically-refined version. Doctor Dick was desperately worried about finding a hitherto undiscovered flaw in the approach and that they expected academic challenges from many quarters.

"Look, I've got to go," he said, draining the last of his pint, "There's lots more work to do, and Dick wants to get a paper out tomorrow."

After my drink with Paul, I wandered back to the grotty terraced house I shared with several other postgrads, alternately musing on what I had heard in the pub and on my own technical problems. I slept badly, with obscure and formless dreams roiling through my head all night. Finally, I gave up and got out of bed, and was back in the School very early - at least for me.

The route to my own office took me past the double doors which opened into the machine room occupied by Dick's Tool and his team. Something made me look again, some subliminal change in the appearance of the doors made me turn back. The doors were not quite shut and the combination lock was not set.

Gingerly, I pushed open the door. The whole area was entirely empty, and the glass partitions stood open. There were just a few black scuff marks on the floor tiles to show where the supercomputer had stood. The desks were devoid of the workstations Paul and his colleagues had been using so busily only last night, and the filing cabinets stood open and empty. There was no-one around - the whole group must have left, completely, in the middle of the night.

The official announcement later that day, in the form of an "All Research Staff and Students" email from the Head of School, was short and singularly un-illuminating. "By mutual agreement," it read, "the Synthetic Computational Modelling group led by Dr. Dick" - his full name was included here - "has moved to another research institution with immediate effect."

There was no mention of which Institution - I noted that the word "University" had not been used - Dick and his team had moved to. Nor was there any reason given for the rapidity of the transfer or for the lack of any prior announcement.

I was curious to learn more. I tried to get in contact with Paul by email using the School's own mail address, assuming that the Computer Services Systems Management Office would have set up email forwarding as a matter of course. The note bounced back immediately, marked "undeliverable". I accessed the University's online staff and student directory. There was no entry for Doctor Dick, or Paul, or indeed for any of the other members of the group. All references had been completely expunged.

I asked around the other research students and some of the younger and more approachable member of the staff. No-one seemed to know, and indeed few seemed particularly concerned. Rather, they appeared more interested in squabbling over who was going to take over the prime office and machine room space previously occupied by Dick's group.

Finally, I tried the only other method readily available to me to track down Paul - the use of those Internet search engines which seem so adept at pulling up information on the most unlikely of subjects. I just found a few rather stale and generic references to the research work that, thanks to Paul, I was already aware of. In all honesty, I had expected to find an announcement that a "prestige research group" had joined some - presumably foreign - University. But there was nothing; it was exactly as if the entire research group had disappeared of the face of the Earth - or, at least, off the face of the Internet.

I have been forced to conclude that Dick and his Merry men had been co-opted into some secret - and probably Government - organisation, some top-secret Think Tank or advisory body, some establishment associated with climate change or some policy planning group, or perhaps the research arm of some multi-national company with deep pockets and vested interests.

I remained curious, browsing the Web frequently. I was expecting to see, perhaps, a policy change announcement from Brussels or the United Nations, or even Dick's long-awaited seminal paper on his results, vindicating his modelling approach after all the years of disbelief. Nothing at all, not a peep anywhere on the destination of Paul and Doctor Dick, or on the earth-shattering implications of the Sustainability Matrix.

Over the last few days, googling the same search terms, I have noticed the search produced progressively fewer results referencing Dick's research output. Right now, there are just a few vague and third-hand references to his modelling approach, more-or-less buried in the noise that keyword searches tend to elicit from the search engines. It is almost as if records were being carefully and progressively deleted.

The realisation has finally been forced upon me: what kind of secretive cabal or international company or government organisation has the kind of clout to extirpate the academic output from a university research group from the collective knowledge of the world academic community? More importantly, I have to ask the question: who or what is preventing us all from knowing the likely impact of the current human occupation of this planet? And, why on Earth are they doing it?

Part 1 Afterword