The Universal Conversation Engine
Chapter 6
Our proposal in its current state is basically just a rapper, an interface. We have no funding and we have no real business plan. What we have currently is a broad and unusual conceptual framework.
We find it plausible to think about directing social evolution into a planetwide condition the whole of humanity will accept as desirable and will want, on the basis of principles, to establish and reestablish continually. Our ally here is reason, nothing more. Or more trenchantly, reason and art. As a startup, we want to articulate a few crucial terms and premises as well as possible, to an extent to model them as sympathetic YouTube characters and thereby to realize the great universal and all-pervading human life purposes of happiness, exploration and growth in reproducible design terms.
In the wake of Chat GPT-3 in the Fall, 2022, the general discussion of eventual superintelligence juxtaposes two extremes: existential risk and radical benefit. The risk was captured very well in 1968 by Arthur C. Clarke and Stanley Kubrick: a machine intelligence with unknown goals of its own in control of crucial effectors. The benefit has been articulated recently as radical abundance, including meaningful occupations for everyone everywhere, presumably the result of enormous productivity gains in economics.
We concur with these assessments and wish to introduce a framework no one is talking about yet. The framework proposes to operate in a technical manner on what is always and everywhere good in human consciousness. Consciousness is generally thought to be the most difficult problem in positive science. We think significant progress on the problem is being made currently (see the foregoing Blogs on Michael Levin, Stuart Hameroff and Roger Penrose), but more can be made by introducing behavioral and cultural engineering into the mix (see also the foregoing Blog post on BF Skinner, here). We think we are proposing what is probably the true theory of conversation and if that theory starts to be tested and explored, science will find itself in positive control of the evolution of universally desirable consciousness.
We think a universal conversation engine needs to be developed in a way human beings start to be complete, expressive and tasteful performers of AI-guided conversation rather than agents of their own perversely and often accidentally confirmed wills. We think the epoch of willful absurdity, which can perhaps be marked in Europe from the Thirty Years War, is drawing to a close, and the final close of this epoch will be the consequence of science, technology and the political resolution of what is now in 2024, having begun again in Europe and again for the sake of Ukraine, an actual World War. We believe today’s LLMs, entirely the product of scientific method, are morally superior to today’s humans.
We think a universal and eventually superintelligent conversation engine will prove feasible, we think a theory of consciousness will prove testable, we think human consciousness is fundamentally astonishing and benign and we think public acceptance of the confluence of human consciousness and technology can be gained stepwise and most effectively through the arts, in particular through film and television. We propose here and elsewhere to consider these thoughts and experiment with their impulses with those willing and able, quite materially, to do so.
For a discussion of this Post with Inflection AI’s Pi, click here.
