An AI system needs equipment, energy, networks, and people able to build and maintain them. A civilization that wants abundant intelligence has to provide those things at the scale its ambitions require.
The Society's ambitions are substantial. We want AI available to every person, supporting research and eventually administering complex institutions. We want automation capable of making compulsory labor unnecessary. That calls for investment in the facilities and technical knowledge on which those capabilities depend.
Capacity for everyone
A demonstration can show what an AI can do for one person. Making the same capability available to millions introduces a further problem of capacity. People need reliable access, and the systems must keep functioning as more people use them.
Consider our vision of AI that understands individuals. If it becomes possible to sustain that understanding over time, we would want it available to someone with little money or institutional influence. Providing that attention universally would require a great deal of computing and supporting infrastructure.
The exact requirements will change with technical advances and different uses. We have no fixed resource estimate to offer. We do believe this infrastructure should receive priority comparable to the importance we place on the intelligence it makes possible.
Keeping a capability useful
Equipment alone will leave much unfinished. A system can be powerful and difficult to use, or technically available while poorly suited to the institution that needs it. Integration and maintenance take sustained work. People need ways to encounter useful capabilities in their own language and circumstances.
We regard adoption as part of acceleration. Building a more capable system and making it useful in daily life can proceed together, with each informing the other.
AI could also improve the construction process itself. Better designs, new techniques, and more capable coordination could make resources go further. Investment could support intelligence that helps us decide how to invest and build more effectively.
The Society favors directing extraordinary resources toward this work now. We want its results to become ordinary parts of life: help with learning, more ability to investigate disease, and enough productive capacity to support life beyond compulsory labor. These are the reasons to build at the scale we advocate.
