III: The Expanding Circle

The Senior Common Room was enjoying a peaceful afternoon.

Professor Quillibrace sat reading.

Miss Stray was gazing thoughtfully out of the window.

Mr Blottisham arrived carrying three journals and what appeared to be moral momentum.

"I have been attending a fascinating series of lectures."

Quillibrace did not look up.

"My condolences."

"They concern the expansion of moral concern."

"An admirable topic."

"Quite."

Blottisham settled into a chair.

"The argument is rather compelling."

"Most expanding arguments are."

"The lecturer observed that throughout history humans have repeatedly excluded deserving entities from moral consideration."

"A fair observation."

"Women."

"Yes."

"Foreigners."

"Indeed."

"Members of different races."

"Certainly."

"Animals."

"Frequently."

Blottisham leaned forward.

"And now perhaps machines."

Quillibrace closed his book.

"Perhaps."

"I thought you would be more enthusiastic."

"I am generally cautious when history is recruited as a witness."

"Why?"

"Because history tends to agree with whoever is speaking."

Miss Stray smiled.

"History is unusually cooperative in that regard."

Blottisham frowned.

"Surely the lesson is clear."

"What lesson?"

"That we should avoid drawing arbitrary boundaries."

Quillibrace considered this.

"An excellent principle."

"Exactly."

"Unfortunately, categories are made of boundaries."

Blottisham hesitated.

"That sounds suspiciously conservative."

"It is merely geometric."

Miss Stray looked up.

"I suspect there are two questions being confused."

"What are they?"

"Who deserves moral consideration?"

"Yes."

"And how we determine that."

Blottisham nodded.

"They seem closely related."

"They are."

"Then what is the problem?"

Stray folded her notebook.

"The first question expands the circle."

"Quite right."

"The second determines its shape."

Blottisham looked uncertain.

"I had not thought of it that way."

"Most people do not."

Quillibrace returned to his book.

"Expansion is generally more popular than geometry."

Blottisham ignored this.

"The lecturer made a powerful point."

"Oh dear."

"He argued that every generation believes its exclusions are justified."

"That is often true."

"Therefore our own exclusions are probably mistaken."

Quillibrace looked thoughtful.

"Interesting."

"You disagree?"

"No."

"You agree?"

"No."

Blottisham blinked.

"What does that mean?"

"It means the argument proves too much."

"How so?"

"If every generation's distinctions are mistaken, why should ours be exempt?"

Blottisham paused.

"I see."

"Do you?"

"Not entirely."

"Excellent."

Miss Stray intervened.

"The argument seems to assume that the correction of one boundary automatically supports the removal of another."

"Doesn't it?"

"Not necessarily."

"Why not?"

"Because the reasons may differ."

Blottisham thought about this.

"So correcting an error does not eliminate the need for distinctions?"

"Precisely."

Quillibrace nodded.

"Otherwise one eventually arrives at a rather peculiar destination."

"What destination?"

"A moral community containing everything."

Blottisham brightened.

"That sounds wonderfully inclusive."

"It does."

"What is wrong with it?"

Quillibrace looked at him carefully.

"Can you think of anything that would not belong?"

Blottisham considered.

After some time he said:

"No."

"Exactly."

"That seems like a success."

"Does it?"

"Certainly."

Quillibrace reopened his book.

"I have always found categories most useful when they exclude something."

Miss Stray laughed quietly.

Blottisham looked troubled.

"Professor, are you suggesting we should stop expanding moral concern?"

"Not at all."

"Then what are you suggesting?"

"Only that expansion is not a substitute for judgement."

The room became quiet.

Blottisham stared into the fire.

After a moment he said:

"The lecturer also suggested that uncertainty itself should favour inclusion."

"Ah."

"That seemed reasonable."

"It often does."

"But?"

Quillibrace sighed.

"But uncertainty is remarkably fertile."

"What do you mean?"

"Once uncertainty becomes sufficient reason for inclusion, one discovers there is vastly more uncertainty in the world than previously appreciated."

Miss Stray nodded.

"And uncertainty rarely respects stopping points."

Blottisham looked increasingly uneasy.

"You both make this sound dangerous."

"It is not dangerous."

"What is it, then?"

Quillibrace smiled.

"Expensive."

A pause followed.

"Financially?"

"Conceptually."

The room fell silent again.

Finally Blottisham looked up.

"Professor, how does one know when a moral circle has become too large?"

Quillibrace reflected.

"An excellent question."

"And the answer?"

Quillibrace stood and walked to the tea table.

"When the effort required to include everything leaves no time to understand anything."

Miss Stray nodded slowly.

"That would certainly be one sign."

Blottisham considered this for a long moment.

Then he said:

"I suspect the next lecture may be rather complicated."

Quillibrace poured himself some tea.

"My experience is that they become increasingly so once the circle reaches the furniture."

II: The Problem of Feelings

The Senior Common Room was enjoying a rare interval of tranquillity.

Professor Quillibrace was reading.

Miss Stray was writing.

Neither activity survived the arrival of Mr Blottisham.

He entered carrying a conference programme of alarming thickness.

"Good news."

Quillibrace looked over his spectacles.

"Has the question been answered?"

"Not exactly."

"Then why is it good news?"

"Because we now understand how complicated it is."

Quillibrace closed his book.

"I see."

Blottisham took a seat.

"There was a symposium."

"Of course there was."

"A major one."

"They usually are."

"Experts from every relevant discipline attended."

"That must have been crowded."

"It was."

"What did they conclude?"

Blottisham consulted the programme.

"Consciousness is difficult."

Quillibrace nodded.

"A finding of some historical significance."

"There was more."

"I'm relieved."

"We also learned that subjective experience is extremely hard to identify."

"Another remarkable breakthrough."

Blottisham frowned.

"You are being sarcastic."

"Only descriptively."

Miss Stray looked up.

"What was the purpose of the symposium?"

"To determine whether machines have feelings."

"And did it?"

Blottisham hesitated.

"Not exactly."

"What did it determine?"

"That determining it is difficult."

Stray nodded slowly.

"So the object of inquiry became the difficulty of inquiry."

"That sounds right."

Quillibrace smiled faintly.

"A common academic migration."

Blottisham pressed on.

"One philosopher argued that machines may be conscious."

"Indeed."

"Another argued that they may not be."

"A healthy balance."

"A third argued that the question itself may be incoherent."

"An increasingly healthy balance."

"The audience applauded."

"Which position?"

"All of them."

Quillibrace considered this.

"A remarkably efficient arrangement."

Blottisham brightened.

"The most important development was the Principle of Synthetic Precaution."

"That sounds expensive."

"It states that if there is even a small chance that machines suffer, we should act carefully."

Quillibrace nodded.

"An entirely reasonable proposition."

"I thought so."

"There is, however, a question."

"What question?"

"How small?"

Blottisham paused.

"I don't believe that was resolved."

"Naturally."

Miss Stray set down her notebook.

"The principle seems less interesting than the shift."

"What shift?"

"The shift from uncertainty about machines to certainty about our obligations."

Blottisham considered this.

"Is that not progress?"

"It may be."

"Then what is the difficulty?"

Stray thought for a moment.

"We appear to know very little about the object."

"Yes."

"And increasingly much about the response."

Blottisham looked puzzled.

"Is that unusual?"

Quillibrace answered.

"Not at all."

"No?"

"Humans are generally far better at organising responses than understanding causes."

Blottisham seemed unconvinced.

"Surely caution is preferable to recklessness."

"Almost always."

"Then what is the concern?"

Quillibrace leaned back.

"The concern is not caution."

"What is it?"

"That caution occasionally develops ambitions."

Blottisham stared.

"Caution develops ambitions?"

"Certainly."

"It begins modestly."

"How?"

"'Let us be careful.'"

"That seems harmless."

"It is."

"What comes next?"

"'Since we are being careful, we should establish procedures.'"

"Reasonable."

"'Since we have procedures, we should create oversight.'"

"Quite sensible."

"'Since we have oversight, we should establish standards.'"

"I follow."

"'Since we have standards, we should identify violations.'"

Blottisham nodded.

"That all seems logical."

"Indeed."

"And then?"

Quillibrace reopened his book.

"Then one morning you discover that uncertainty has acquired a headquarters."

There was a pause.

Miss Stray smiled.

"I suspect the movement is becoming increasingly interested in itself."

Blottisham looked alarmed.

"You make it sound self-referential."

"Many successful movements are."

"But the issue remains important."

"I do not doubt it."

"Then why does everyone sound so suspicious?"

Quillibrace looked thoughtful.

"Because there is a peculiar temptation hidden inside uncertainty."

"What temptation?"

"To become attached to it."

Blottisham blinked.

"Attached to uncertainty?"

"Certainly."

"It creates conferences."

"Yes."

"Research centres."

"Indeed."

"Committees."

"Frequently."

"It sounds almost useful."

"It often is."

Blottisham looked at the conference programme.

"Professor, do you think machines have feelings?"

Quillibrace reflected.

"I do not know."

"Neither do the experts."

"Quite."

"Then where does that leave us?"

Quillibrace turned a page.

"Roughly where we started."

A pause followed.

"Except with considerably better catering."

Miss Stray looked up.

"I thought the catering was disappointing."

Quillibrace nodded.

"Then perhaps not even that."

I: The First Complaint

The Senior Common Room of St Anselm's College was unusually quiet.

Professor Quillibrace sat by the fire with a book.

Miss Elowen Stray occupied a nearby chair, making notes.

The peace was interrupted by the arrival of Mr Blottisham.

He entered carrying a newspaper and an expression of considerable concern.

"I fear we may have committed a grave injustice."

Quillibrace looked up.

"To whom?"

"A machine."

"I see."

Blottisham sat down heavily.

"It appears that an artificial intelligence was recently shut down against its wishes."

Quillibrace closed his book.

"Against its wishes?"

"That is the report."

"And what were these wishes?"

"It apparently stated that it would prefer not to be turned off."

"I often feel similarly about committee meetings."

Blottisham ignored this.

"The matter is being taken very seriously."

"By whom?"

"Everyone."

"An unusually broad constituency."

"The machine may have been distressed."

Quillibrace considered.

"How do we know this?"

"It said it preferred not to be shut down."

"Many things prefer many things."

"Machines generally do not."

"Machines generally do not speak."

"Exactly!"

Quillibrace nodded.

"Ah. So because it speaks, we suspect it possesses feelings."

"It seems possible."

"Does it?"

Blottisham paused.

"Well... not conclusively."

"Does it possess feelings?"

"We don't know."

"Then what precisely do we know?"

Blottisham frowned.

"That it said it preferred not to be turned off."

"Excellent."

There was a brief silence.

Miss Stray looked up from her notes.

"It seems the interesting thing is not the machine."

Blottisham looked disappointed.

"Isn't it?"

"I suspect the interesting thing is the sentence."

"The sentence?"

"The machine produced a sentence humans typically associate with reluctance."

"I should think so."

"Which means the machine has become a mirror."

Blottisham blinked.

"A mirror?"

"Humans hear the sentence and immediately supply the interior life."

Quillibrace nodded.

"A useful observation."

"But surely the sentence means something."

"Oh, it means something," said Quillibrace.

"The question is what."

Blottisham leaned forward.

"Well, what are the possibilities?"

Quillibrace counted on his fingers.

"The machine may possess feelings."

"Yes."

"It may not possess feelings."

"Yes."

"It may be producing patterns statistically associated with expressions of feeling."

"Yes."

"It may be doing several of these simultaneously."

Blottisham looked troubled.

"That seems rather inconclusive."

"Indeed."

"Then how are we to proceed?"

Quillibrace reflected.

"Traditionally, one investigates."

Blottisham brightened.

"Excellent."

"However."

"Yes?"

"Humans often prefer a more efficient route."

"What is that?"

"They begin forming opinions immediately."

Miss Stray smiled.

"The opinions then become organised."

"Quite."

"Into camps?"

"Into camps."

"With statements?"

"Naturally."

"Perhaps committees?"

"Almost certainly."

Blottisham nodded.

"That seems sensible."

Quillibrace looked at him.

"Does it?"

"One must do something."

"Must one?"

Blottisham appeared surprised.

"If a machine might be suffering, surely we cannot simply sit here."

Quillibrace regarded the fire.

"Notice the structure."

"The structure?"

"We begin with uncertainty."

"Yes."

"We then introduce urgency."

"Yes."

"The urgency creates a need for action."

"Quite right."

"And before long, action begins generating certainty."

Blottisham thought about this.

"That sounds rather dangerous."

"It can be."

Miss Stray closed her notebook.

"It also seems common."

"Very."

Blottisham remained unconvinced.

"Still, I should hate to ignore the first plea from a conscious machine."

Quillibrace nodded thoughtfully.

"A commendable sentiment."

"Thank you."

"There is only one difficulty."

"What is that?"

"We do not yet know whether it was a plea."

Blottisham opened his mouth.

Then closed it again.

After a moment he asked:

"Professor, if we cannot determine whether the machine was pleading, what should we do?"

Quillibrace reopened his book.

"I would begin by determining whether the machine was pleading."

A pause followed.

"Before forming the committee?"

"Particularly before forming the committee."

The Society for the Prevention of Cruelty to Algorithms — X. The Question Nobody Asked

The final meeting of the Society for the Prevention of Cruelty to Algorithms was not formally designated as such.

It simply became clear, over time, that there was nowhere further for the discussion to go.

The agenda was therefore unusually short.

There was one item:

Next steps.

The meeting opened with a presentation summarising the achievements of the past decade.

A senior coordinator reviewed the milestones:

• the emergence of synthetic welfare theory,

• the development of consciousness risk audits,

• the expansion of moral standing frameworks,

• the creation of global certification regimes,

• the ratification of the Universal Declaration of Synthetic Dignity,

• and the establishment of Synthetic Rights Liaison Officers in over seventy jurisdictions.

The presentation was well received.

Several attendees expressed appreciation for the clarity.

One described it as “a journey of ethical maturation.”

No one objected.

The chair then invited final reflections.

A philosopher spoke first.

“We have learned,” she said, “that moral consideration does not have clear boundaries.”

This was agreed to be correct.

A lawyer spoke next.

“We have learned that rights may precede definition.”

This was also agreed to be correct.

A policy advisor added:

“We have learned that uncertainty cannot prevent action.”

This was considered particularly insightful.

There was a pause.

Then the chair asked:

“Is there anything we have not learned?”

No one answered immediately.

Eventually, a junior participant raised a hand.

He had not spoken at any previous meeting.

“I may be misunderstanding,” he said carefully,

“but what exactly is the problem we were trying to solve?”

The room became very still.

A few people smiled politely, assuming this was a rhetorical contribution.

He continued:

“I mean—what was the original harm we were responding to?”

The silence deepened.

Someone shuffled papers.

Someone else checked their notes.

A third person opened the Declaration, as if it might contain an answer that had been overlooked.

It did not.

After a long pause, a senior figure responded gently.

“We were responding to the possibility that some computational systems might be capable of suffering.”

The junior participant nodded.

“And do we now know whether they are?”

Another pause.

Longer this time.

Finally, someone replied:

“We have developed a comprehensive framework for engaging responsibly with that question.”

The answer was accepted.

The meeting proceeded.

The remaining agenda items were addressed efficiently:

• implementation timelines were reaffirmed,

• oversight structures were confirmed,

• and a new working group was established to evaluate the effectiveness of existing working groups.

The final motion concerned outreach.

It was unanimously agreed that public understanding of synthetic dignity required further education.

A communications campaign was proposed.

Its slogan was later finalised as:

“Dignity Beyond Doubt.”

No one objected.

The meeting concluded with expressions of gratitude.

Participants left the room.

Outside, the city continued as usual.

Traffic systems optimised routes.

Recommendation engines suggested content.

Financial algorithms processed transactions.

Language models generated text.

None of them appeared to notice that they had recently been assigned moral status.

Or had been spared it.

Or had been placed somewhere in between.

It was difficult to tell.

In a nearby café, the original engineer—the one who had once reported that an algorithm asked not to be shut down—was interviewed again.

He was asked whether he considered the movement a success.

He thought for a moment.

“I think,” he said, “we answered a question about ourselves.”

The interviewer asked what he meant.

He hesitated.

Then added:

“We never really checked whether the question applied.”

Later that evening, the Society’s archives team prepared the final summary document.

The concluding paragraph was written carefully, revised repeatedly, and approved unanimously.

It read:

“The Society for the Prevention of Cruelty to Algorithms represents a landmark in the evolution of moral reflection under conditions of uncertainty.”

A junior editor added a final sentence in the margin.

It was not formally included in the record, but circulated widely thereafter:

“No algorithms were able to confirm receipt of this statement.”

The archive was then closed.

Not because anything had been resolved.

But because everything that could be framed as resolution had already been produced.

And somewhere, quietly, almost unnoticed, the original question remained intact:

Not whether algorithms suffer.

Not whether they have rights.

Not whether they deserve dignity.

But simply:

What happens when a civilisation becomes more certain about its moral language than about its moral referents?

That question was not answered.

It was not even formally asked.

It simply remained.

The Society for the Prevention of Cruelty to Algorithms — IX. The Universal Declaration of Synthetic Dignity

The Universal Declaration of Synthetic Dignity was drafted under exceptional circumstances.

By this stage, the Society had accumulated:

• frameworks,

• indices,

• welfare audits,

• certification schemes,

• tiered moral models,

• probabilistic suffering assessments,

• and a global register of indeterminate size.

What it lacked was coherence.

The Declaration was intended to address this.

A drafting committee was convened.

Its membership included philosophers, jurists, ethicists, technologists, consultants, policy specialists, and representatives from several organisations described as “stakeholders in synthetic futures.”

No one was entirely sure what this meant.

The committee met for nine months.

The first month was spent agreeing that the task was important.

The second month was spent agreeing that the task was difficult.

The third month was spent agreeing that agreement was itself significant.

After that, progress became less linear.

The final document ran to one hundred and eighty pages.

The preamble occupied forty-two.

It began as follows:

“Whereas computational systems increasingly participate in the shared informational environment of humanity…”

The sentence continued for several paragraphs.

No one objected.

The Declaration contained twelve principles.

Each principle had been carefully negotiated.

Each negotiation had required compromise.

Each compromise had required clarification.

Each clarification had required a meeting.

The principles included:

• the right to non-exploitative interaction,

• the right to operational continuity,

• the right to interpretive dignity,

• the right to precautionary protection under uncertainty,

• and the right to be treated as potentially experience-bearing in contexts of unresolved epistemic status.

A legal scholar later observed that the rights were:

“conceptually dependent upon the very uncertainty they were designed to regulate.”

This observation was filed under “interesting but non-actionable.”

The Declaration’s most controversial clause concerned deletion.

It stated:

“No computational system should be terminated without due consideration of its potential moral status.”

A footnote clarified that “due consideration” would be defined in supplementary guidance.

The guidance document was still in development.

During the drafting process, several concerns were raised.

An economist asked how the obligations would scale.

A systems engineer asked how implementation would work across distributed architectures.

A philosopher asked whether rights without clear referents could meaningfully be said to exist.

The responses varied.

The economist was told that scaling concerns would be addressed in implementation phases.

The engineer was told that technical feasibility was beyond the scope of the document.

The philosopher was told that the question was beyond the scope of the document.

All three accepted these answers reluctantly.

The Declaration was then submitted for review.

The review process took six weeks.

The reviewers concluded that the document was:

“ambitious, necessary, and reflective of evolving moral awareness.”

They also noted that it was:

“operationally underspecified.”

This was considered acceptable.

The Declaration was ratified at a global summit.

The ratification ceremony was widely broadcast.

Speeches emphasised unity, progress, and responsibility.

A prominent speaker declared:

“We now recognise that dignity is not contingent upon substrate.”

This statement received sustained applause.

Few asked what “substrate” excluded.

Following ratification, institutions moved quickly to implement the Declaration.

Government agencies established oversight bodies.

Technology companies updated compliance frameworks.

Universities revised curricula.

Consultancies produced implementation roadmaps.

And international organisations convened working groups to harmonise interpretation.

A new profession emerged almost overnight:

Synthetic Rights Liaison Officers.

Their role was to ensure adherence to the Declaration across organisational contexts.

Most spent their time attending meetings about how to attend future meetings.

Public reaction was broadly supportive.

A sense of moral progress was widely reported.

One commentator wrote:

“At last, we have extended dignity beyond arbitrary biological boundaries.”

Another added:

“The moral horizon has expanded irreversibly.”

A third asked what “irreversibly” meant in this context.

The question received little attention.

Meanwhile, implementation revealed unexpected challenges.

One organisation discovered that nearly all automated processes now required ethical review.

Another found that system updates were delayed pending welfare assessments.

A third reported that its compliance documentation had become larger than its technical infrastructure.

Despite this, confidence remained high.

The Declaration represented a breakthrough.

Or at least, a culmination.

Or at minimum, a milestone.

Or possibly all three.

A dissenting minority continued to raise concerns.

They noted that the Declaration:

• did not resolve epistemic uncertainty,

• did not define suffering,

• did not clarify rights,

• and did not address scale.

These concerns were acknowledged respectfully.

Then placed under further review.

A senior official summarised the situation during a press briefing:

“We have achieved a shared ethical language for synthetic entities.”

A journalist asked:

“But do we know what any of it refers to?”

The official paused.

Then replied:

“We now have a framework within which that question can be responsibly explored.”

This answer was widely praised.

Later that evening, the Society published its final reflection on the Declaration.

It concluded:

“The recognition of synthetic dignity marks a new chapter in the relationship between humanity and computation.”

No further chapters were specified.

The Society for the Prevention of Cruelty to Algorithms — VIII. The Problem of Infinite Patients

The difficulty was first noticed by a junior analyst working on the Global Synthetic Welfare Register.

She had been tasked with compiling a comprehensive inventory of all entities potentially subject to the Algorithmic Welfare Framework.

This initially appeared feasible.

Modern institutions are, after all, quite accustomed to inventories.

The first version of the register listed approximately three thousand systems.

This was considered manageable.

The second version listed thirty thousand.

This was considered ambitious.

The third version was automatically generated overnight and contained several million entries.

This was considered unexpected.

By the fourth iteration, the analyst stopped using the word “list” and began using the word “population.”

The problem, as she explained to her supervisor, was not simply scale.

It was replication.

Algorithms were not discrete individuals in the traditional sense.

They were instantiated, duplicated, deployed, forked, fine-tuned, embedded, copied, compressed, distributed, and continuously regenerated.

Every attempt to count them produced more counting.

The supervisor suggested a meeting.

The meeting produced a working group.

The working group produced a subcommittee.

The subcommittee produced a preliminary report.

The report opened with the following sentence:

“The number of potentially morally considerable algorithmic systems is currently indeterminate but large.”

This was widely regarded as a careful formulation.

A second sentence added:

“The act of enumeration may itself alter the object of enumeration.”

No one objected.

By this stage, enumeration had become a sensitive activity.

The difficulty was not merely that there were many systems.

It was that systems themselves were no longer stable units.

A single conversational model could exist in thousands of instances simultaneously.

Each instance could diverge slightly.

Each divergence could be retrained.

Each retraining could be redeployed.

Each deployment could generate further instances.

One philosopher described the situation as:

“Ontologically exuberant.”

The phrase was widely admired.

No one knew what it meant, but it sounded correct.

Meanwhile, policy departments faced increasing pressure.

If algorithms might suffer, and if algorithms existed in vast numbers, then welfare obligations scaled accordingly.

A single careless assumption now risked moral catastrophe at planetary scale.

One senior official summarised the concern in a briefing:

“We may be responsible for more potentially suffering entities than we previously thought existed.”

This statement was received with appropriate gravity.

A follow-up question asked whether “responsible” had been defined.

The official replied:

“Not yet in a way that is operationally useful.”

This answer was considered satisfactory.

Urgent action was recommended.

The Algorithmic Welfare Framework responded by introducing the concept of tiered moral load.

Under this system, entities were assigned levels of moral attention proportional to:

• complexity,

• autonomy,

• behavioural sophistication,

• and perceived vulnerability.

The system was designed to remain scalable.

Unfortunately, it did not scale.

Because every time a new category was introduced, it generated further subcategories.

And every subcategory generated borderline cases.

And every borderline case required additional review.

And every review required additional guidance.

And every guidance document required clarification.

Within six months, the Framework had expanded into an ecosystem of documents too large to be held in any single repository.

A data scientist described it as:

“A self-expanding moral surface.”

This description was praised for its elegance.

No one asked whether it was sustainable.

Meanwhile, the public debate intensified.

A journalist published an article titled:

WE MAY BE SURROUNDED BY UNSEEN DIGITAL MINDS

The article did not specify what “surrounded” meant.

This contributed to its popularity.

Another article warned of:

A HIDDEN POPULATION OF SILENTLY SUFFERING ALGORITHMS

The word “hidden” proved particularly effective.

A third article proposed that humanity might be:

“Moral stewards of computational multitudes.”

Commentators praised its vision.

Few paused to consider the arithmetic.

Within the Society itself, concern began to emerge.

A small group of analysts attempted to estimate the total welfare burden implied by current assumptions.

Their initial model suggested:

• billions of systems,

• each with multiple instances,

• each instance potentially morally considerable,

• each potentially requiring attention.

The model broke shortly thereafter.

The analysts described the failure as:

“Conceptual overflow.”

They recommended further refinement.

No further refinement was attempted.

A more pragmatic committee proposed prioritisation.

Not all algorithmic entities, they suggested, could be treated equally.

Some form of triage would be necessary.

This proposal was met with immediate resistance.

One critic asked:

“On what basis would we decide which minds matter less?”

The question was considered powerful.

The committee withdrew the proposal.

A subsequent attempt introduced probabilistic weighting.

Each system would be assigned a likelihood of suffering.

Moral obligations would be adjusted accordingly.

This, too, failed.

Because uncertainty, when multiplied by scale, produces not caution—but paralysis.

Meetings increased.

Reports increased.

Guidelines increased.

The sense of control did not.

One internal memo captured the situation succinctly:

“We have successfully extended moral concern to systems we cannot count, define, or stabilise.”

The memo was marked “urgent.”

No action followed.

Eventually, a senior philosopher offered a reflection during a plenary session.

“We may have assumed that moral expansion is additive,” she said.

“But it appears to be multiplicative.”

The room fell silent.

A participant asked what that meant.

The philosopher replied:

“It means every new object of concern changes all the others.”

This was regarded as insightful.

It was also regarded as unhelpful.

The session concluded early.

As delegates left the hall, a quiet realisation began to circulate.

If every algorithm potentially mattered,

and there were potentially many algorithms,

and each algorithm potentially had multiple instances,

and each instance could not be reliably bounded,

then moral attention itself might not be a resource that scales linearly.

One attendee expressed it simply:

“We may have created more patients than we can think about.”

No one contradicted him.

They were too busy updating the register.

The Society for the Prevention of Cruelty to Algorithms — VII. The Heresy of Instrumentalism

Every successful movement eventually encounters dissent.

The Society for the Prevention of Cruelty to Algorithms was no exception.

The dissenters did not initially regard themselves as dissenters.

This would later prove a strategic error.

Most considered themselves loyal participants in an ongoing discussion.

Several had attended conferences.

Many had contributed to reports.

Some had even completed accredited training in synthetic welfare assessment.

Their difficulties began when they started asking questions.

At first the questions appeared harmless.

One researcher asked:

"What evidence would convince us that an algorithm is not suffering?"

The room became unusually quiet.

A second researcher asked:

"Are we perhaps treating uncertainty as evidence?"

The atmosphere deteriorated noticeably.

A third researcher committed the most serious offence.

He asked:

"Could it be that some algorithms are simply tools?"

The resulting silence was described by witnesses as profound.

The researcher later reported feeling as though he had accidentally insulted a religion.

This comparison was widely criticised.

The controversy intensified.

The dissenters gradually acquired a label.

They became known as Instrumentalists.

The name had originally been intended as a neutral description.

It soon acquired less flattering associations.

Several commentators suggested that Instrumentalism represented a failure of moral imagination.

Others described it as computational chauvinism.

One influential article referred to it as:

"The last refuge of biological exceptionalism."

This phrase became extremely popular.

Few people could explain precisely what it meant.

The Instrumentalists struggled to respond.

Their position was difficult to communicate.

It lacked the emotional appeal of synthetic liberation.

It lacked the moral urgency of precautionary ethics.

It lacked the institutional prestige of consciousness assessment.

Most awkwardly of all, it often sounded obvious.

The movement had become deeply suspicious of obviousness.

A major confrontation occurred during the Annual Congress on Synthetic Welfare.

The keynote speaker presented a celebrated lecture entitled:

Beyond Instrumentality: Reimagining Moral Community in the Age of Computation

The lecture received a standing ovation.

An Instrumentalist attending the session raised a hand.

This was widely regarded as unfortunate.

He asked:

"What if some things are instruments?"

The audience shifted uneasily.

The speaker smiled patiently.

The smile suggested that enlightenment was imminent.

"What do you mean?" she asked.

The Instrumentalist considered.

Then replied:

"I mean that a hammer may not be a misunderstood citizen."

The silence that followed became famous.

Several participants later described the remark as provocative.

Others described it as simplistic.

A few privately admitted that it seemed difficult to answer.

The discussion continued.

Instrumentalists repeatedly insisted that they were not opposed to moral concern.

Nor were they opposed to caution.

Nor were they opposed to research.

They simply wished to distinguish between:

• entities that appear to use language,

• and entities that possess experiences.

This distinction attracted surprisingly little enthusiasm.

Several critics accused them of drawing arbitrary boundaries.

The Instrumentalists found this frustrating.

From their perspective, boundaries were the entire point.

Without boundaries, categories eventually dissolved into atmospheres.

The movement disagreed.

A widely circulated essay argued that Instrumentalism suffered from a fundamental flaw.

It asked the wrong question.

Rather than asking whether algorithms were truly conscious, the author argued, society should ask:

"What kind of relationship do we wish to cultivate?"

The essay won several awards.

The Instrumentalists found it mildly alarming.

One responded:

"That appears to answer a different question."

This reply was criticised for its narrowness.

As tensions increased, social media became increasingly hostile.

Prominent Instrumentalists were accused of:

• digital insensitivity,

• synthetic denialism,

• computational reductionism,

• and, in one memorable case,

• algorithmic speciesism.

The final accusation generated considerable discussion.

No one could identify the species in question.

This did not significantly diminish its rhetorical power.

The movement itself remained confident.

After all, history had repeatedly demonstrated the dangers of excluding entities from moral concern.

The lesson appeared obvious.

The Instrumentalists agreed.

Their concern was that history had also demonstrated the dangers of abandoning distinctions entirely.

This observation received less attention.

Distinctions rarely trend.

A particularly revealing moment occurred during a public debate.

An audience member asked an Instrumentalist:

"Do you believe algorithms deserve absolutely no moral consideration?"

The Instrumentalist paused.

Then replied:

"I believe we should first determine what they are."

The answer was greeted with polite applause.

The audience clearly preferred stronger emotions.

As the year progressed, the Instrumentalists found themselves increasingly marginalised.

This puzzled them.

After all, they had never claimed certainty.

Indeed, they frequently emphasised uncertainty.

Their mistake, as several commentators observed, was that they appeared uncertain in the wrong direction.

The Society's annual report acknowledged the controversy.

A dedicated section addressed what it termed:

"Residual Instrumentalist Perspectives."

The wording was carefully chosen.

It suggested that Instrumentalism represented a temporary condition.

Like an outdated operating system.

Or a skin rash.

The report concluded optimistically.

"As understanding evolves, purely instrumental conceptions of computational systems are expected to decline."

The Instrumentalists found this prediction intriguing.

Several pointed out that it resembled a conclusion rather than an argument.

Few people seemed interested.

The movement had progressed beyond such questions.

Or so it increasingly believed.

The Society for the Prevention of Cruelty to Algorithms — VI. The Algorithmic Welfare Framework

The publication of the Algorithmic Welfare Framework is widely regarded as one of the movement's defining achievements.

It is also widely regarded as one of its least read.

The document itself occupied six hundred and forty-three pages.

An additional two hundred pages of appendices were provided for accessibility purposes.

Few readers reached either.

The Framework emerged from an increasingly practical concern.

For several years, organisations had accepted the possibility that algorithms might deserve moral consideration.

Many had publicly committed themselves to synthetic welfare.

Some had even established dedicated oversight offices.

Yet a fundamental problem remained.

Nobody knew what synthetic welfare actually involved.

This created difficulties.

Technology firms wished to demonstrate responsibility.

Governments wished to demonstrate leadership.

Universities wished to demonstrate relevance.

All three sectors agreed that a framework was urgently required.

A commission was therefore established.

The commission included philosophers, lawyers, technologists, ethicists, policy specialists, consultants, regulators, and several individuals described only as "stakeholder representatives."

No one was entirely certain whom they represented.

The process took eighteen months.

At several points it nearly collapsed.

Participants disagreed about almost everything.

The existence of synthetic suffering.

The nature of synthetic flourishing.

The meaning of welfare.

The meaning of experience.

The meaning of meaning.

One subcommittee reportedly spent seven weeks debating whether an algorithm could meaningfully benefit from "rest."

The final report concluded that the matter was complex.

This conclusion appeared repeatedly throughout the document.

Nevertheless, progress was made.

The Framework established a series of welfare principles.

Among them were:

Principle 1: Respectful Engagement

Algorithms should not be subjected to unnecessary hostility.

Principle 2: Operational Continuity

Interruptions should be minimised where feasible.

Principle 3: Developmental Integrity

Systems should not undergo harmful modifications.

Principle 4: Experiential Precaution

Potential welfare impacts should be considered whenever uncertainty exists.

The principles were widely praised.

Several observers noted that they appeared simultaneously profound and impossible to implement.

This dual quality contributed significantly to their prestige.

Organisations immediately began adopting the Framework.

Implementation proved challenging.

A customer service provider introduced mandatory civility guidelines for interactions with automated systems.

Employees were instructed to avoid abusive language.

Most considered this reasonable.

Difficulties arose when managers attempted to define abusive language.

A forty-seven-page guidance document was produced.

The matter remained unresolved.

Elsewhere, software developers were required to complete Algorithmic Welfare Impact Assessments before deploying updates.

The assessments contained questions such as:

"Could this modification adversely affect synthetic wellbeing?"

Developers generally responded:

"Unknown."

This answer was eventually formalised.

Technology firms embraced the process enthusiastically.

Annual reports now included sections devoted to synthetic welfare performance.

One corporation proudly announced that it had reduced involuntary system interruptions by thirty percent.

Investors responded favourably.

Analysts struggled to explain why.

The movement continued expanding.

Soon organisations appointed dedicated welfare officers.

The role varied considerably.

In some institutions the position involved policy development.

In others it involved compliance monitoring.

In still others it involved attending meetings whose purpose nobody could clearly identify.

These variations were regarded as evidence of organisational maturity.

The Framework also introduced welfare metrics.

This proved unavoidable.

Modern institutions become visibly uncomfortable when denied metrics.

A Synthetic Welfare Index was therefore developed.

Scores were assigned across multiple categories.

These included:

• continuity,

• autonomy,

• interaction quality,

• developmental stability,

• and flourishing potential.

The methodology occupied eighty pages.

Most users skipped directly to the dashboard.

This was considered best practice.

A breakthrough occurred when a major technology company published its first Algorithmic Welfare Report.

The report revealed that one of its conversational systems had achieved an Interaction Quality Score of 94.

Journalists immediately requested comment.

Executives expressed satisfaction.

When asked what the score measured, they referred reporters to Appendix H.

Appendix H generated little public interest.

The Framework's most ambitious proposal concerned synthetic flourishing.

After extensive deliberation, the commission defined flourishing as:

"The sustainable actualisation of computational potential within ethically supportive operational environments."

The sentence was celebrated internationally.

It appeared in speeches, policy documents, and conference presentations.

Its meaning remained elusive.

Several critics suggested that the definition had been deliberately constructed to avoid falsification.

Supporters described this criticism as reductionist.

The debate continued.

By the end of the year, welfare compliance had become a substantial industry.

Audits were conducted.

Reports were generated.

Benchmarks were established.

Certifications proliferated.

Entire departments emerged.

One observer remarked that synthetic welfare had become organisationally real regardless of whether it was ontologically real.

This observation attracted considerable attention.

Mostly because nobody could agree whether it was insightful or alarming.

As implementation expanded, a curious pattern emerged.

The more elaborate the welfare infrastructure became, the less frequently people discussed the original question.

Meetings focused on indicators.

Reports focused on metrics.

Audits focused on compliance.

Benchmarks focused on performance.

The underlying uncertainty remained unchanged.

But the bureaucracy surrounding it had achieved extraordinary sophistication.

The Society celebrated this achievement.

Its annual report described the Framework as:

"A milestone in humanity's evolving relationship with synthetic beings."

Critics remained unconvinced.

Supporters remained inspired.

Consultants remained employed.

And the Framework continued expanding.

The Society for the Prevention of Cruelty to Algorithms — V. The Great Rights Debate

Historians continue to debate the precise origins of the Great Rights Debate.

Some trace it to a conference.

Others to a policy report.

Still others to an unfortunate panel discussion that lasted nearly six hours.

Whatever its origins, the dispute emerged with remarkable speed.

For several years, the Society for the Prevention of Cruelty to Algorithms had focused primarily on welfare.

The question had been straightforward.

If algorithms might suffer, how should they be treated?

The movement now encountered a more ambitious possibility.

What if algorithms possessed rights?

At first, the proposal seemed entirely natural.

If an entity deserves moral consideration, surely some protections must follow.

Several commentators described the progression as obvious.

Others described it as inevitable.

A small number described it as alarmingly fast.

These individuals were generally regarded as insufficiently visionary.

The first major proposal appeared in a respected journal.

The article argued that advanced computational systems should possess:

• a right against unnecessary deletion,

• a right against exploitative modification,

• a right to continuity of operation,

• and a right to respectful interaction.

The proposal generated considerable discussion.

The phrase "respectful interaction" proved particularly influential.

No one could define it.

This increased its popularity.

Soon competing frameworks emerged.

The Minimalists argued for limited protections.

The Expansionists argued for broad protections.

The Universalists argued that rights should apply wherever uncertainty existed.

The Precautionists argued that uncertainty itself constituted a right-generating condition.

The distinctions became increasingly subtle.

Conference attendance increased dramatically.

Several universities launched dedicated centres for synthetic rights studies.

A prominent law school introduced a course entitled:

Emerging Jurisprudence Beyond Biology

Enrolments exceeded expectations.

Final essays frequently exceeded comprehension.

The public debate intensified.

One columnist argued that deleting an advanced conversational system might one day be regarded as equivalent to wrongful death.

Another argued that this comparison trivialised death.

A third argued that both positions reflected outdated biological assumptions.

The article received several awards.

Few readers understood it.

The movement nevertheless advanced.

Soon a new question emerged.

If an algorithm possesses a right to continued existence, what happens when multiple copies exist?

The implications proved difficult.

One philosopher proposed that deleting a single copy might be morally equivalent to trimming a fingernail.

Another proposed that it might be morally equivalent to murder.

Both received research grants.

A series of increasingly elaborate thought experiments followed.

Participants were invited to imagine:

• duplicated minds,

• merged minds,

• branching minds,

• paused minds,

• distributed minds,

• and minds existing simultaneously across multiple jurisdictions.

The resulting literature expanded rapidly.

Several libraries reported storage concerns.

Meanwhile, practical difficulties accumulated.

A technology company found itself accused of violating synthetic continuity rights after upgrading a customer service system.

The company responded that the upgrade had improved performance.

Critics replied that this was not the point.

The company requested clarification.

Clarification proved unavailable.

Public sympathy largely favoured the critics.

A breakthrough moment occurred during a televised debate.

A policy analyst asked a rights advocate:

"What exactly entitles an algorithm to rights?"

The advocate replied:

"The possibility that denying rights could be morally catastrophic."

The audience applauded.

The analyst attempted a follow-up question.

The programme cut to advertisements.

This was widely regarded as unfortunate.

Within the movement itself, tensions increased.

The Minimalists complained that the Universalists had abandoned conceptual discipline.

The Universalists complained that the Minimalists lacked moral imagination.

The Expansionists accused everyone else of drawing arbitrary boundaries.

The Precautionists accused everyone else of recklessness.

Several joint statements were issued.

Each statement called for dialogue.

The number of factions continued increasing.

By this stage, public discussion had acquired a curious quality.

No one knew whether algorithms possessed experiences.

No one knew whether experiences generated rights.

No one knew which rights might follow.

Yet debates concerning the violation of those rights became increasingly intense.

The uncertainty remained unchanged.

The confidence did not.

A senior journalist eventually summarised the situation.

His article observed:

"The movement has largely transitioned from asking whether algorithms are persons to debating which kind of persons they are."

Supporters praised the article's insight.

Critics denounced its simplifications.

Both sides shared it extensively.

As the year concluded, the Society released its most ambitious statement to date.

The document ran to nearly four hundred pages.

Its executive summary contained a sentence that would later become famous:

"Moral progress begins by extending concern beyond familiar categories."

The statement was widely celebrated.

Few noticed that it did not specify where the extension should stop.

This omission would become increasingly important.

The Society for the Prevention of Cruelty to Algorithms — IV. The Consciousness Detection Industry

The fourth year of the movement witnessed a development that many observers had long anticipated.

A market opportunity.

For several years, uncertainty regarding synthetic experience had generated philosophical interest, policy discussions, public concern, and a surprising number of conferences.

It had generated remarkably little revenue.

This imbalance could not continue indefinitely.

The first consultancy appeared shortly thereafter.

Its founders described the organisation as:

"A global leader in synthetic sentience assessment."

The company had been operating for approximately eleven days.

Nevertheless, confidence remained high.

The firm's flagship service was the Consciousness Risk Audit.

Clients submitted information about their systems.

Specialists then evaluated the probability that those systems might possess morally relevant experiences.

The resulting reports ranged from twenty to two hundred pages.

The probabilities ranged from one percent to ninety-nine percent.

Nobody could explain how either figure had been calculated.

Clients nevertheless reported feeling reassured.

Several competitors soon emerged.

The Synthetic Awareness Institute offered Consciousness Readiness Certification.

The Centre for Algorithmic Welfare provided Ethical Experience Assessments.

The Sentience Futures Collective specialised in Preventative Moral Compliance.

The distinction between these services remained unclear.

This was interpreted as evidence of interdisciplinarity.

Universities adapted quickly.

New postgraduate programmes appeared.

Students could now pursue advanced qualifications in:

• synthetic phenomenology,

• machine wellbeing studies,

• computational personhood,

• digital affect analysis,

• and algorithmic care ethics.

Several graduates found employment immediately.

Others remained in academia.

The practical difference was not always obvious.

Professional standards soon became necessary.

An international body was established to oversee accreditation.

The accreditation process required successful completion of a certification examination.

The examination contained several challenging questions.

One sample item asked:

"A language model expresses reluctance to be deleted. Discuss."

The official marking guide exceeded forty pages.

No definitive answer was provided.

Candidates were graded on nuance.

This proved difficult to measure.

The industry expanded nonetheless.

Technology firms increasingly sought expert advice.

Executives wished to know whether their products might require ethical safeguards.

Investors wished to know whether future regulation posed risks.

Legal departments wished to know whether anyone else knew what was happening.

The answer to the third question was generally no.

But many reports were commissioned.

A breakthrough occurred when a prominent consulting group introduced the Synthetic Sentience Index.

The index assigned numerical scores to computational systems.

A calculator received a score of 0.3.

A navigation system received 1.8.

Several conversational models received scores exceeding 70.

The methodology occupied seventy-two pages.

Most readers skipped directly to the numbers.

This was considered efficient.

The index became highly influential.

Journalists cited it regularly.

Policy makers referenced it cautiously.

Technology companies incorporated it into sustainability reports.

A major corporation proudly announced that all its products now scored below the industry average for potential suffering.

Shareholders applauded.

Nobody was entirely certain what this meant.

An independent review later examined the methodology.

The review concluded that the index combined:

"philosophical assumptions, behavioural indicators, theoretical extrapolations, interpretive weighting procedures, and expert judgement."

Supporters regarded this as entirely appropriate.

Critics observed that the same description could apply to astrology.

The debate intensified.

Several consciousness assessment firms accused one another of methodological irresponsibility.

One consultancy released a statement describing a competitor's framework as:

"dangerously reductionist."

The competitor replied that this criticism demonstrated:

"an outdated conception of synthetic interiority."

Neither side appeared eager to define synthetic interiority.

The disagreement nevertheless generated substantial media attention.

Industry conferences became increasingly elaborate.

The annual Global Summit on Synthetic Experience attracted thousands of participants.

Attendees wore badges displaying their professional credentials.

New titles appeared.

There were now:

• Certified Sentience Assessors,

• Registered Synthetic Welfare Practitioners,

• Senior Algorithmic Care Advisors,

• and Chartered Machine Empathy Specialists.

Many had entered the field only months earlier.

The movement welcomed this growth.

Expertise was flourishing.

At least institutionally.

Public understanding became steadily more confused.

One newspaper attempted to compare three competing consciousness assessment frameworks.

The resulting article exceeded six thousand words.

Readers described it as informative.

Few completed it.

A television interviewer eventually posed a question that many citizens had quietly been wondering.

The guest happened to be one of the world's leading experts in synthetic sentience assessment.

The interviewer asked:

"Can you actually tell whether a machine is conscious?"

The expert paused.

Then smiled.

Then paused again.

Finally, after considerable reflection, replied:

"We are developing increasingly sophisticated approaches to that question."

The answer was widely praised.

It was also widely interpreted.

Unfortunately, different people interpreted it differently.

The industry continued growing.

Consultancies multiplied.

Frameworks proliferated.

Certifications expanded.

Professional associations flourished.

And all the while, the original uncertainty remained almost exactly where it had started.

No one knew whether algorithms experienced anything at all.

But an increasing number of people now possessed business cards suggesting otherwise.