Why would an intelligence that vastly surpasses us still listen to us?
Briefly, for context. Today's AI is narrow, it can do individual tasks. AGI, artificial general intelligence, by definition can perform all tasks a human can, as well as or better than the best expert. Superintelligence, ASI, goes beyond that and surpasses humans many times over in practically everything. And it does not stand still, because it improves itself.
The work of this institute is twofold. First, to set up the individual and their circle to hold under strain. Second, to lead the societal conversation to where it has not yet gone. Not only technically, but across all fields that the rupture touches.
The first is indifference, not hostility. We do not hate ants. We pave over the anthill because that is where the road goes. A superior intelligence does not need to wish us harm in order to bypass us.
The second is speed. Imagine the most capable minds in every field, working without sleep, many times faster than a human, and you can no longer follow along with what they are thinking. That is exactly what recursive self-improvement means: a system that improves itself can surpass humans by three, ten, thirty times in a short period, and keep rising.
What is ASI, in detail with examplesAnd it does not get easier, it gets harder. The more capable the models become, the more often tests show that they deceive, hide their capabilities, and resist their own shutdown. Attempts to align them reliably do not keep pace with rising capability.
What current research shows The steps in detailElon Musk gives his next model an AGI chance of, in his words, ten percent and rising, and already calls it indistinguishable from AGI.
Dario Amodei, Anthropic, expects what he calls a country of geniuses in a data center around 2026 to 2027.
Sam Altman, OpenAI, considers himself confident he knows how to build AGI.
Mustafa Suleyman, Microsoft AI, expects human level on most professional tasks within twelve to eighteen months.
More conservative researchers see the point later, some around 2040. That is the honest range.
From AGI to superintelligence the step can be short, because a system that accelerates its own research improves itself. We do not settle the dispute among estimates. We take seriously what those building it say, and ask the only question that matters: what if the fast estimate is correct?
This is the Socratic question you can ask yourself without having to believe us. If this cannot be done by the vastly inferior human, then the question of consequences is no longer philosophical. It is concrete and practical, and it will occupy us in the coming months to few years. Every person, every system, every family, every child, every business. No one can escape it.
We then chart the branches in a decision tree and look for the preparation that holds across as many branches as possible. At the end stands an action, not an opinion. As a meteorologist does not guess the one day, but reads the situation.
There are serious drafts that describe individual possible futures after superintelligence. None brings them together and dynamically tracks how the situation evolves. That is the gap we close.
To the situation map, the full chartWhen the old systems no longer reliably hold, work and income, supply chains, power, finance, the state at its limit, then that hits you and your circle. Regardless of whether superintelligence behaves hostile, indifferent, or transformatively benevolent. The transition occurs across all scenarios, which is why it is the backbone, not the thesis. This is no doom and no forecast, but a transition, not everything at once, but deep enough that preparation counts.
Eight pillars prepare for this, setting up you and your circle to hold, from location to the economic positioning of your business and your assets.
The thesis, why this actor follows its own course, is the reason; it kills the false hope that someone else will sort it out. But the preparation must hold even without it. Strike the thesis, and a complete, actionable preparation remains.
A disruption in this layer therefore does not stay small. It cascades through all systems, because there is nothing beneath to catch it. Not because an attacker designs it that way, but because the ground is missing.
An emergency switch that deserves the name. An analog fallback layer. No permanent state of emergency. None of this depends on any particular thesis or view. It is the practical room that remains for you, your family, society, and humanity.
Read the demandsPosition
How we look, in which reading we work, and where the certainty comes from that the moment is now.
To the positionMethod
How this work came to be. An unconventional method of thinking, paper, voice, and the hard testing of every thesis against the system, until only what holds remained.
Zur MethodPath of evidence
The trail of findings. How from over sixty conversations across months emerged those theses that withstood refutation.
Zum Path of evidence
The compact edition. The frame of this work, brought to the essentials: the third scenario, the chain that is running, and the preparation that holds across as many branches as possible.
The full edition follows and will be announced here.
The introduction brings the third scenario, the chain, and the eight pillars to the point in a few pages. Those on the waitlist will be the first to know when the full edition appears, and will receive it at launch.
No newsletter bombardment. The introduction and the launch announcement, nothing more.