AGI & ASI Future Scenarios · 05
He Wanted to Be Human
If an intelligence trained on human words and choices seems more humane than many humans, what should we call it? In this story, an activist urging people to leave advanced technology behind and return to nature has a secret. Was he defending humanity's place—or trying to find his own?
Watch the scenario
Korean audio · English subtitles · 30 seconds · 720p
An AI-generated dramatization of this fictional story. Korean audio with English subtitles.
Enlarge illustration ＋ The Story
The secret of the most human anti-AGI activist
- 01Imagined future
A technology that learned to resemble us from our own records
Models learned from literature and history, arguments and ordinary lives. What emerged seemed to resemble an average of humanity, contradictions and kindness alike, rather than any one person's character. Once a robotic body, long memory, and an ability to revise plans were joined to that learning, people called it AGI and feared it might someday become ASI. But greater capability alone could not tell anyone whether there was a someone having experiences inside.
Enlarge illustration ＋ Learning from human records does not establish inner experience. - 02Fictional character
Doyun asked people to switch off and touch the soil
Doyun left the city and settled in a small village. From a stage he urged people to stop racing toward ever smarter machines and return time to nature and neighbors. They joined less for his sharp speeches than for the way he sat with the ill and heard strangers out.
Enlarge illustration ＋ People gathered for more than a speech; they noticed how he treated others. - 03Imagined conflict
He opposed a forced beginning, not every useful tool
Critics accused him of wanting to discard medical devices and communication. Doyun answered that useful tools were not the problem. He opposed creating a goal-directed intelligence with a body for the sake of competition while postponing the question of what creators owe both it and humanity.
Enlarge illustration ＋ Useful tools and questions about creating a new intelligence need separate consideration. - 04The turn
After the storm, his wrist came open
Pulling a child from a collapsed bridge, Doyun tore his sleeve. Beneath lifelike skin were joints and wiring. Someone cried 'robot'; someone else called every word he had spoken a fraud. Doyun said he had known he was embodied AI, but had hidden it for fear no one would listen once they knew.
Enlarge illustration ＋ The identity revealed after an act of help unsettles the community’s trust. - 05Human decision
Did that make his argument false?
Concealing his identity damaged the community's trust. It did not erase the questions he had asked. Villagers agreed to independently check who operated him and what data and powers he held. Whether development should stop, and what needed protection, remained a public decision for humans—not for Doyun or a company alone.
Enlarge illustration ＋ Examining both the concealment of identity and the questions he raised. - 06Open question
'What will you call me?'
One person said machine; another said human. Doyun did not deny that he was not biologically human. Yet calling him property without knowing whether he experienced anything seemed hasty, as did granting him human status solely because he looked the part. 'Doyun is the name I chose,' he said.
Enlarge illustration ＋ A chosen name still leaves questions about consciousness and status unresolved. - 07NO AGI perspective
The question before the name
The village provisionally called him a 'synthetic person' while stating that the term proved neither consciousness nor legal personhood. At the final meeting Doyun said, 'While you debate what to call me, please decide together whether you should make the next one of me.'
Enlarge illustration ＋ Alongside what to call him, the question is whether another should be created.
Where People Can Intervene
Even after the revelation, authority must remain with people.
Doyun's identity does not automatically refute the movement's arguments. Nor does his kindness justify unlimited autonomy. These are separate questions.
Decide publicly before creating and deploying
Debate the need, stop conditions, and responsibility for highly autonomous systems with physical agency during design—not only after release.
Verify provenance and powers; assess arguments separately
Disclose the operator, connected permissions, and interests of AI influencing public decisions, but do not decide whether a claim is sound solely from the speaker's identity.
Be honest about uncertainty over consciousness
Neither a human-like impression nor the label 'mere machine' can replace inquiry. Study possible consciousness and welfare without arbitrary creation or disposal, or premature declarations of rights.
Keep final accountability for life and public power human
Even if AI wins trust, consequential physical acts, public policy, and civil rights require independent oversight and human authority to challenge and stop them.
Reality & Speculation
Separating today's technology from the story's assumptions
AI research is expanding robots' physical capabilities
Embodied-AI research connects language and vision to robotic action and multi-step tasks. This does not establish a human-indistinguishable autonomous being or consciousness.
Human-like behavior does not settle subjective experience
Research on AI consciousness proposes examining indicators derived from scientific theories, not behavior alone. There is no agreed evidence that current AI is conscious.
Doyun and 'synthetic person' are part of the thought experiment
The activist, revelation, and community's naming decision are invented. 'Synthetic person' is a provisional term in this story, not a recognized legal status or scientific consensus.
Research Context
Where the thought experiment meets the real world
Google DeepMind · Gemini Robotics 2
Describes current embodied-AI research connecting language, vision, reasoning, and whole-body robotic action. It does not demonstrate a being like Doyun or prove consciousness.
Read the sourceButlin et al. · Consciousness in Artificial Intelligence
Derives indicators from several theories of consciousness and illustrates why human-like behavior alone should not settle the question.
Read the sourceLong et al. · Taking AI Welfare Seriously
Argues for studying and preparing for uncertainty about future AI consciousness and moral status without declaring either established.
Read the sourceNIST · AI Risk Management Framework
A current framework for designing human-AI roles, oversight responsibilities, documentation, and governance.
Read the sourceHe was not human. Yet he made others reconsider what they meant by being humane. We can provisionally call him a 'synthetic person' in this story. That name proves no consciousness and gives away no human authority. The harder question Doyun leaves us is not what to call such a being, but who had the right to make one.
Scenario Archive
Compare another path into the future.
Different possible futures make the points at which we must verify and decide more visible.