I wrote this on impulse as a response to “A Reason Behind Bad Systems, and Moral Implications of Seeing This Reason”. It is essentially a first draft and is probably badly written. The plot also got away from me in the later sections; my mental processes surrounding a lot of these phenomena are too cross-linked. However, I sense it may be more useful for this to exist in the open than not, and I think if I decide to hide this away for rewriting, realistically I will not do so and it will languish forever… so here it is. Perhaps it can become something better later; perhaps not.
For copyright and moral rights purposes, as the sole author of everything in this post that is not a quotation, I place all such text in the public domain.
The programmer with the messy program did what any sensible software engineer would have done in her shoes: Set aside the old software and built something useful, that did what was needed, using the full power of the company’s modern computers.
A month later, Evelyn's manager at Yottascape, Henrietta, stormed into Evelyn's office with a notepad about the hundred different telephone calls the department had gotten complaining that the results from the new system were subtly inconsistent with those from the old. Other people who had little to no understanding of the code behind the figures they were seeing had built processes and expectations based on patterns they observed, patterns which had stayed stable for years.
Some of the new outputs had caused cascading instabilities. One manufacturing location had run out of slack on a key inflow of materials, temporarily halting outflow of parts to a second location, which revealed that workers at a second location had become lax in ensuring that the machine they were using—which needed to be kept at a certain temperature or suffer expensive damage that would put it out of commission for months—was actually kept at that temperature. The intake at the second location had been so constant before that it created a self-regulating effect, and they had stepped down control checks to one third the original frequency, a cost-saving measure which had been widely praised by management. Now the damage would lead to broken promises which would have to be documented in their next SEC filing. Even before then, it would involve a flurry of contract renegotiations with distributors, investors, and suppliers, some of whom would gossip among themselves, sensing that something was amiss and that their common counterparty had fallen off-balance, and push ruthlessly for favorable terms that would strip Yottascape to the bone.
Evelyn pointed out that her new outputs were, in fact, more correct than the old ones on several axes, and that she had documented these in the white paper she had sent out ahead of time. Hadn't they read her equations? The new code would also allow for much faster evolution, and it would be much easier to understand. She had projected that it would save a million dollars a year in maintenance costs, and potentially another million dollars a year in efficiency improvements. And her projections about software maintenance had always been correct before.
Henrietta replied that since Evelyn was a senior in the department who was given great operational leeway, it was part of her job description to ensure she got appropriate feedback and that the effects were properly understood before implementing such a drastic change unilaterally. Additionally, the finance department was projecting several million dollars a year of reduced revenue for several years yet as a result of the fault cascade. Further, less easily quantifiable losses might well accrue from morale loss leading to key innovative workers in the field leaving when their own efforts turned out to be wasted as a result of unpredictable events like this.
Evelyn was fired. Her code was rolled back, and the second most senior in the department, Saul, was promoted to lead the adjustment project in her stead. Saul gave an expansive presentation to Evelyn about how he would ensure his new code would not disrupt operations, and how there would be an extensive network of backup plans, rollback options, and safety checks, as well as additional advance training for anyone who might be affected by changes in this most critical program. He also noted that he would need a budget and access to managerial authority sufficient to take the time out of the workdays of other affected parties to ensure this would go smoothly. Henrietta worked with the executives above her and got all of this approved, despite wincing at the additional cost.
Six months after the original incident, gradual uptake was proceeding mostly apace—well, actually, it was notably slower than expected, as it seems Saul had been somewhat optimistic—but no further massive disruptions had accrued. Most of the acute damage had since been repaired, and while everyone still suffered from the effects of the more onerous contract terms and did not reasonably expect them to end soon, they had gotten used to it. Things had returned, mostly, to normal, with the prospect of new, safer improvements on the horizon.
II
Twelve months after the original incident, a hotshot new company, Zettaboom, announced product lines competing with three of Yottascape's main revenue sources. They were already delivering with more options and higher reliability at half the price, and the tide of consumer sentiment, including the lingering (if tarnished among anyone who read the business section of the newspaper) perception of Yottascape as the premier manufacturer in their space, was starting to turn away. Henrietta frantically checked to see whether Zettaboom had hired Evelyn—and they hadn't. But they had contracted with a consultancy that Evelyn had joined. The Zettaboom product managers were careful to keep the consultants' work at arm's length and retain complete control over acceptance criteria, but they also remained open-minded and lacked the burden of legacy procedure, and the code quality from Evelyn and her new coworkers was a key factor in their ability to pivot quickly and handle new market conditions while keeping costs and risks down. This appealed greatly to Zettaboom's investors, allowing them fast access to surprising amounts of capital to grow as a Yottascape competitor.
III
Five years after the original incident, unrest was rising among the public. Life had grown too hectic, somehow. Everything was shinier and glossier on the surface, but they all felt worse off. Those who could adapt to chaos well had gravitated to key positions in society, and then the chaos itself had become an evolutionary defensive measure, one not imposed by any dictat but only by the jitter of tense, barely-held-together, complex, implicit control systems. Old pillars had crumbled, somehow, from the weight of compounded individual choices. No one had consciously decided that Yottascape's old-fashioned community involvement should be diminished, or noticed the subtle differences in worker lifestyle, only made decisions using the information they had, and gotten a wave of recommendations from friends and acquaintances and ‘known’ that this was where society was going and they'd better get with the times.
Some people blamed everyone including themselves for getting in the habit of buying from Zettaboom and similar ‘soulless’ corporations, but very few had anything like a solution to propose, and no one could deny that the benefits had existed too—they could only lament that they hadn't coordinated on different priorities earlier. Others pointed out that executing on those priorities the way the first set of complainants wished had happened might not have worked out either. No one from either group could make any headway; the cultural and structural drift had made it almost impossible for someone of median intelligence or median social connections to make it very far in the world, creating a semi-permanent class rift.
Those of great mental capacity who did make it into the managerial class immediately entered a cycle of having to prove themselves by wielding all their intelligence to create more and more complex systems. Anyone who tried to simplify things would be seen as weak, and their market positions would erode, their influence would be treated with disgust, and they would quickly be expelled to become part of the underclass, below even the commoners. If they tried to rejoin the commoners, after all, they would be viewed with a different disgust as having contributed to The Problem, and they would have already rendered themselves unfit in habit to take part in any of it—so they were left with nowhere to go. Knowing that this could happen made those of the managerial class even more anxious.
I wrote this on impulse as a response to “A Reason Behind Bad Systems, and Moral Implications of Seeing This Reason”. It is essentially a first draft and is probably badly written. The plot also got away from me in the later sections; my mental processes surrounding a lot of these phenomena are too cross-linked. However, I sense it may be more useful for this to exist in the open than not, and I think if I decide to hide this away for rewriting, realistically I will not do so and it will languish forever… so here it is. Perhaps it can become something better later; perhaps not.
For copyright and moral rights purposes, as the sole author of everything in this post that is not a quotation, I place all such text in the public domain.
Epistemic status: fictional, quasi-allegorical, biased, exploratory.
I
A month later, Evelyn's manager at Yottascape, Henrietta, stormed into Evelyn's office with a notepad about the hundred different telephone calls the department had gotten complaining that the results from the new system were subtly inconsistent with those from the old. Other people who had little to no understanding of the code behind the figures they were seeing had built processes and expectations based on patterns they observed, patterns which had stayed stable for years.
Some of the new outputs had caused cascading instabilities. One manufacturing location had run out of slack on a key inflow of materials, temporarily halting outflow of parts to a second location, which revealed that workers at a second location had become lax in ensuring that the machine they were using—which needed to be kept at a certain temperature or suffer expensive damage that would put it out of commission for months—was actually kept at that temperature. The intake at the second location had been so constant before that it created a self-regulating effect, and they had stepped down control checks to one third the original frequency, a cost-saving measure which had been widely praised by management. Now the damage would lead to broken promises which would have to be documented in their next SEC filing. Even before then, it would involve a flurry of contract renegotiations with distributors, investors, and suppliers, some of whom would gossip among themselves, sensing that something was amiss and that their common counterparty had fallen off-balance, and push ruthlessly for favorable terms that would strip Yottascape to the bone.
Evelyn pointed out that her new outputs were, in fact, more correct than the old ones on several axes, and that she had documented these in the white paper she had sent out ahead of time. Hadn't they read her equations? The new code would also allow for much faster evolution, and it would be much easier to understand. She had projected that it would save a million dollars a year in maintenance costs, and potentially another million dollars a year in efficiency improvements. And her projections about software maintenance had always been correct before.
Henrietta replied that since Evelyn was a senior in the department who was given great operational leeway, it was part of her job description to ensure she got appropriate feedback and that the effects were properly understood before implementing such a drastic change unilaterally. Additionally, the finance department was projecting several million dollars a year of reduced revenue for several years yet as a result of the fault cascade. Further, less easily quantifiable losses might well accrue from morale loss leading to key innovative workers in the field leaving when their own efforts turned out to be wasted as a result of unpredictable events like this.
Evelyn was fired. Her code was rolled back, and the second most senior in the department, Saul, was promoted to lead the adjustment project in her stead. Saul gave an expansive presentation to Evelyn about how he would ensure his new code would not disrupt operations, and how there would be an extensive network of backup plans, rollback options, and safety checks, as well as additional advance training for anyone who might be affected by changes in this most critical program. He also noted that he would need a budget and access to managerial authority sufficient to take the time out of the workdays of other affected parties to ensure this would go smoothly. Henrietta worked with the executives above her and got all of this approved, despite wincing at the additional cost.
Six months after the original incident, gradual uptake was proceeding mostly apace—well, actually, it was notably slower than expected, as it seems Saul had been somewhat optimistic—but no further massive disruptions had accrued. Most of the acute damage had since been repaired, and while everyone still suffered from the effects of the more onerous contract terms and did not reasonably expect them to end soon, they had gotten used to it. Things had returned, mostly, to normal, with the prospect of new, safer improvements on the horizon.
II
Twelve months after the original incident, a hotshot new company, Zettaboom, announced product lines competing with three of Yottascape's main revenue sources. They were already delivering with more options and higher reliability at half the price, and the tide of consumer sentiment, including the lingering (if tarnished among anyone who read the business section of the newspaper) perception of Yottascape as the premier manufacturer in their space, was starting to turn away. Henrietta frantically checked to see whether Zettaboom had hired Evelyn—and they hadn't. But they had contracted with a consultancy that Evelyn had joined. The Zettaboom product managers were careful to keep the consultants' work at arm's length and retain complete control over acceptance criteria, but they also remained open-minded and lacked the burden of legacy procedure, and the code quality from Evelyn and her new coworkers was a key factor in their ability to pivot quickly and handle new market conditions while keeping costs and risks down. This appealed greatly to Zettaboom's investors, allowing them fast access to surprising amounts of capital to grow as a Yottascape competitor.
III
Five years after the original incident, unrest was rising among the public. Life had grown too hectic, somehow. Everything was shinier and glossier on the surface, but they all felt worse off. Those who could adapt to chaos well had gravitated to key positions in society, and then the chaos itself had become an evolutionary defensive measure, one not imposed by any dictat but only by the jitter of tense, barely-held-together, complex, implicit control systems. Old pillars had crumbled, somehow, from the weight of compounded individual choices. No one had consciously decided that Yottascape's old-fashioned community involvement should be diminished, or noticed the subtle differences in worker lifestyle, only made decisions using the information they had, and gotten a wave of recommendations from friends and acquaintances and ‘known’ that this was where society was going and they'd better get with the times.
Some people blamed everyone including themselves for getting in the habit of buying from Zettaboom and similar ‘soulless’ corporations, but very few had anything like a solution to propose, and no one could deny that the benefits had existed too—they could only lament that they hadn't coordinated on different priorities earlier. Others pointed out that executing on those priorities the way the first set of complainants wished had happened might not have worked out either. No one from either group could make any headway; the cultural and structural drift had made it almost impossible for someone of median intelligence or median social connections to make it very far in the world, creating a semi-permanent class rift.
Those of great mental capacity who did make it into the managerial class immediately entered a cycle of having to prove themselves by wielding all their intelligence to create more and more complex systems. Anyone who tried to simplify things would be seen as weak, and their market positions would erode, their influence would be treated with disgust, and they would quickly be expelled to become part of the underclass, below even the commoners. If they tried to rejoin the commoners, after all, they would be viewed with a different disgust as having contributed to The Problem, and they would have already rendered themselves unfit in habit to take part in any of it—so they were left with nowhere to go. Knowing that this could happen made those of the managerial class even more anxious.