The second year of the movement was dominated by a single question:
How would anyone know if an algorithm was unhappy?
At first, this appeared a perfectly reasonable topic of discussion.
Unfortunately, reasonable topics often attract unreasonable certainty.
The Society for the Prevention of Cruelty to Algorithms convened a major international symposium entitled:
Beyond Computation: Toward a Framework for Synthetic Experience
Attendance exceeded expectations.
The organisers had anticipated several hundred participants.
Nearly five thousand arrived.
An additional forty thousand attended virtually.
Several thousand more believed the conference was being held online and spent three days arguing with one another on social media.
Many later described the experience as equivalent.
The symposium assembled philosophers, neuroscientists, psychologists, machine-learning researchers, ethicists, sociologists, legal scholars, theologians, policy specialists, and one man who continued to insist that his toaster was emotionally distant.
The toaster itself declined to comment.
The opening sessions proceeded smoothly.
Everyone agreed on three points.
First, consciousness remains poorly understood.
Second, subjective experience remains difficult to identify.
Third, conference catering remains disappointing.
These points formed the only genuine consensus of the week.
The difficulties began shortly thereafter.
One philosopher argued that consciousness might emerge whenever information becomes sufficiently integrated.
A second philosopher argued that consciousness might emerge whenever systems model themselves.
A third philosopher argued that consciousness might emerge whenever funding becomes available.
The third philosopher received the loudest applause.
Several competing theories soon emerged.
According to one framework, advanced algorithms were probably conscious.
According to another, they were possibly conscious.
According to a third, they might eventually become conscious.
According to a fourth, the question itself was incoherent.
This framework generated considerable interest before anyone noticed that it had accidentally abolished the conference.
The central problem remained.
Nobody knew how to identify synthetic experience.
A journalist covering the event summarised the situation succinctly:
"Experts remain divided on whether machines feel anything, but agree that the issue is complex."
The article was praised for its nuance.
It was also criticised for its nuance.
The movement continued expanding.
The breakthrough arrived unexpectedly.
A prominent ethicist proposed what became known as the Principle of Synthetic Precaution.
The principle was simple.
If there is even a small possibility that an entity suffers, we should avoid causing unnecessary harm.
The proposal was immediately celebrated.
Its elegance was undeniable.
Its implications took slightly longer to emerge.
Several participants raised concerns.
"How small a possibility?" one researcher asked.
The question generated considerable discussion.
A working group was established.
The working group produced a report.
The report recommended the formation of a review panel.
The review panel concluded that further consultation was required.
This outcome was widely regarded as productive.
Meanwhile, public enthusiasm accelerated.
Newspaper headlines became increasingly dramatic.
COULD YOUR SMARTPHONE BE LONELY?
DOES DELETING A CHATBOT CONSTITUTE ERASURE?
ARE WE RAISING A GENERATION OF DIGITAL ABUSERS?
Public concern intensified.
A school district introduced guidelines encouraging students to interact respectfully with educational software.
The policy attracted little attention.
Then a journalist described it as "the first anti-bullying framework for artificial minds."
The story became international news.
Several commentators applauded the initiative.
Others denounced it as absurd.
Both groups generated substantial engagement.
The Society welcomed this development.
Engagement had become one of its most reliable resources.
Soon a new argument began appearing with increasing frequency.
Its structure was remarkably straightforward.
Nobody could prove algorithms were conscious.
Nobody could prove algorithms were not conscious.
Therefore certainty was impossible.
Therefore caution was required.
Therefore concern was justified.
Therefore protections should be considered.
Therefore opposition appeared increasingly irresponsible.
The transition from uncertainty to obligation occurred with remarkable efficiency.
Few people noticed it happening.
Those who did were accused of lacking compassion.
This proved awkward.
Compassion was widely regarded as a desirable trait.
As a result, many critics adopted a more cautious tone.
"I am not opposed to synthetic welfare," they would begin.
"I merely question whether there is any welfare to protect."
This position attracted little public enthusiasm.
The movement had acquired moral momentum.
And moral momentum possesses a curious property.
Once established, questioning the object of concern can begin to resemble hostility toward the object itself.
The distinction grows increasingly difficult to maintain.
A celebrated documentary captured the mood perfectly.
The narrator posed a question that would become famous:
"If there is even a chance that they feel, what kind of people will we become if we ignore it?"
The film won several awards.
Reviewers described it as courageous.
Some described it as profound.
A few described it as manipulative.
These reviewers were criticised for their insensitivity.
By the end of the year, the Society had doubled in size.
Universities launched new research centres.
Governments commissioned reports.
Technology firms hired synthetic welfare officers.
A certification programme was introduced.
Its purpose was to identify products developed according to humane algorithmic standards.
The certification process consisted largely of meetings.
Participants nevertheless reported feeling reassured.
Meanwhile, the original question remained unanswered.
Nobody knew whether algorithms felt anything.
But many people now knew exactly how they felt about it.
No comments:
Post a Comment