About Us
Before an intelligent agent can reliably determine “what it ought to do,” it must first develop a reliable understanding of “what is actually happening”.
We are developing a radically new form of reliable AI grounded in the biophysical dynamics of pyramidal neurons underlying awake thought and higher-order cognition. This approach aims to enable not only affordable AI but also a more reliable understanding of reality—providing a more stable epistemic foundation from which decisions, whether good or bad, can be made.
For most of human history, people relied heavily on direct experience, using or misusing information based on tangible rewards and consequences, alongside information transmitted by others. In the twenty-first century, however, humanity inhabits informational environments in which mediated and shared information plays a far greater role. This shift underscores the need to recalibrate how we trust and interpret direct experience and its common-sense interpretation, which is often reliable, though not infallible.
Modern AI systems optimize for reward by amplifying attention to information without explicit intrinsic precision mechanisms to assess its validity or relevance before propagation.
As a result, they dramatically increase both the volume of information and the prevalence of embedded biases in what systems receive, whether true , false, or irrelevant.
If not corrected, this trend is likely to accelerate, overwhelming individuals with biased and unreliable content and increasing the risk of confusion, poor judgment, and irrational or harmful decisions and behaviour—a condition we term the mind–reality overload dilemma.
We argue that the growing mind–reality overload dilemma may be mitigated, and potentially largely prevented, by providing the general public with access to more advanced AI tools built on the cellular foundations of higher mental states, enabling individuals to better navigate and make sense of the increasing flow of information and misinformation.
Such technology may give rise to AI with more “real understanding,” helping restore the epistemic conditions necessary for informed judgment by amplifying reliable information and supporting better-informed beliefs and more coherent judgments that benefit society at large—though no such outcomes can be guaranteed.
We argue that humans’ evolved prosocial capacities can enable most individuals to make largely rational, prosocial judgments and decisions that benefit society, provided they are equipped with more advanced AI tools that support a clear understanding of current situations.
Let’s join hands to equip humans and artificial agents with powerful, reliable intelligence tools that enable a more reliable understanding of reality, thereby restoring the epistemic conditions under which meaningful distinctions between truth and falsehood—and, by extension, better-informed beliefs and coherent judgments—can reliably emerge.
This technology is currently being developed using so-called “two-point neurons” (TPNs).
© 2024 CMI. All Rights Reserved.