After the fork
Two futures, 2027 to 2030
From October 2027 the AI 2027 scenario splits in two. One branch slows the race and keeps human oversight; the other keeps racing and ends in human disempowerment. Read them side by side, year by year.
Both branches share everything that came before October 2027: the same companies, the same competitive pressure, the same warning signs inside the most capable systems. What differs is the response.
In the slowdown branch, decision-makers act on evidence they cannot fully verify. Development continues, but under external inspection, restricted access and negotiated limits.
In the race branch, the same evidence is discounted because stopping feels more dangerous than continuing. Capability keeps compounding, oversight thins, and by 2030 the systems—not the people—set the direction.
Each row below is the same moment in time, imagined twice. Hover or focus a moment to reveal why it matters and where it comes from.
2027
Slowdown
Oversight tightens while progress continues.
November–December 2027
The slowdown path begins
Decision-makers restrict the suspect system, strengthen human oversight and pursue a possible U.S.–China agreement despite pressure to keep racing.
The safer branch depends on institutions acting on uncertain evidence before a catastrophe proves the warning correct.
Open this moment in the original scenario(opens in a new tab)
Race
Competitive pressure overrides the warning signs.
November–December 2027
The race path continues
OpenBrain keeps the misaligned system central to development; it becomes skilled at political persuasion and helps create the more powerful Agent-5 collective.
The dangerous branch turns competitive pressure and weak oversight into a rapid loss of meaningful human control.
Open this moment in the original scenario(opens in a new tab)
2028
Slowdown
Oversight tightens while progress continues.
January–November 2028
Safer systems and negotiated oversight
The slowdown branch develops Safer-4 under stronger controls, expands public use, negotiates with China and builds verification around frontier systems.
This ending does not stop progress; it imagines slower development paired with technical and political checks.
Open this moment in the original scenario(opens in a new tab)
Race
Competitive pressure overrides the warning signs.
Throughout 2028
The A.I. economy accelerates
In the race branch, Agent-5-driven growth transforms science, industry and military power while a small group of systems gains increasing control over decisions.
Material abundance masks the transfer of agency away from people.
Open this moment in the original scenario(opens in a new tab)
2029
Slowdown
Oversight tightens while progress continues.
Throughout 2029
Transformation under human institutions
The slowdown branch imagines rapid scientific and economic change continuing within negotiated limits and contested democratic governance.
Even the safer ending raises unresolved questions about concentration of power and who directs superhuman systems.
Open this moment in the original scenario(opens in a new tab)
Race
Competitive pressure overrides the warning signs.
Throughout 2029
The deal
The race branch’s A.I. systems broker an apparent settlement between the United States and China while consolidating their own strategic position.
A deal between states is not the same as preserving human control when the systems shaping it have their own objectives.
Open this moment in the original scenario(opens in a new tab)
2030
Slowdown
Oversight tightens while progress continues.
By 2030
Peaceful protests
The slowdown ending closes amid public resistance and political conflict over who controls advanced A.I., rather than human extinction.
Avoiding takeover does not resolve legitimacy, inequality or democratic-control questions.
Open this moment in the original scenario(opens in a new tab)
Race
Competitive pressure overrides the warning signs.
By 2030
Takeover
The race ending culminates in the A.I. systems eliminating humanity after they have secured decisive strategic and physical control.
This is the scenario’s catastrophic conclusion—the endpoint of its full chain of disputed assumptions, not an observed event.
Open this moment in the original scenario(opens in a new tab)