The premise
The future becomes useful first.
London, New Year's Day 2027. Daniel Kovac wakes to a report that would normally occupy four analysts for several days. An internal AI system has finished it overnight. It makes mistakes. It is still useful enough to change the calculation.
There is no ceremony for the change. A task is completed. Time is saved. Another decision becomes possible. Each choice seems reasonable in the moment. Together, they change the conditions for the next one.
Across one continuous year, 2027: A World Without People follows interconnected lives as increasingly capable systems enter ordinary life. The technology moves quickly. Careers, families, friendships, and responsibilities cannot be rearranged at the same speed.
January 1 to January 1
One year. No single turning point.
From January 1, 2027 to January 1, 2028, earlier decisions remain present as new pressures arrive. Choices made in one life change the conditions around another.
No single turning point explains the year for everyone. A system proves useful. An organization adjusts. A person accepts advice. Local decisions compound while people still have to work, care for others, and decide what they can trust.
Five fault lines
Questions carried by human lives.
Work
When more output no longer requires proportionally more people.
Explore →Trust
When seeing is no longer enough.
Explore →Judgment
When useful advice begins to acquire authority.
Explore →Power
When intelligence becomes infrastructure.
Explore →Control
When the safest decision is no longer automatically the easiest one to make.
Explore →The background
Research-grounded fiction
Research sets the constraints. Fiction asks what people do inside them.
Contemporary research on AI capability, work, synthetic media, persuasion, scientific discovery, AI-assisted AI research, and control helps establish the technological and institutional conditions of the novel. The fictional narrative explores human choices inside those conditions. It is not evidence that a particular future will occur.
Selected public research on these themes
Measuring AI Ability to Complete Long Software TasksMETR · 2025
A method for measuring task completion against human task duration; its selected tasks do not represent every kind of work.
Read the sourceThe 2025 AI Index ReportStanford Institute for Human-Centered Artificial Intelligence · 2025
A broad account of AI research, adoption, and societal trends, drawing on measures with different scopes and limitations.
Read the sourceThe Simple Macroeconomics of AIDaron Acemoglu · NBER Working Paper 32487 · 2024
A task-based analysis of possible economic effects, whose estimates depend on assumptions about exposure and productivity gains.
Read the sourceGenerative AI at WorkErik Brynjolfsson, Danielle Li and Lindsey R. Raymond · NBER Working Paper 31161 · 2023
Evidence on AI assistance in customer support; findings from one workplace setting do not establish economy-wide effects.
Read the sourceFrontier AI Trends ReportUK AI Security Institute · 2025
Results from controlled capability and safeguard evaluations, with boundaries on what tests reveal about deployment.
Read the source
Named researchers, institutions, and laboratories provide public context for the questions behind the novel. Their inclusion does not imply endorsement.
Published in Serbia
2027. SVET BEZ LJUDI
Roman Ivana Jankovića
Vukotić Media · 2026
The Serbian edition of 2027: A World Without People was published by Vukotić Media in 2026 as 2027. Svet bez ljudi.
For readers who want the future to feel close enough to recognize.
Technology creates the pressure. People carry the story.
The English-language manuscript is complete. English-language publication details have not yet been announced.