031 - Rethinking Neurodiversity

Neurodiversity is a frequently misunderstood concept. At a basic level, it is often portrayed as a measure of specialization or distance from the cognitive/perceptual norm.

A common example of this misunderstanding is the term "Autism Spectrum Disorder" (ASD), which was created to consolidate the previous Asperger's Syndrome and Autism rating systems. Unfortunately, ASD is even more useless as a category than the terms which preceded it.

The reason for this is simple, it measures distance but neglects direction.

Collective Intelligence derives increasing value from diversity of perspective, making even small quantities of neurodiversity an exceptional asset. However, calculating that diversity of perspective requires both distance and direction.

Two equidistant neurodiverse individuals may be nearly identical in their perspectives, or polar opposites, meaning that the value of that combination may be virtually zero, or the maximum possible. As such, a category that neglects direction is virtually useless in practice, as it only offers a weak and naïve correlation.

The capacity of Collective Intelligence to solve problems and reduce bias is proportionate to the diversity of perspectives covered by that collective's constituent members.

This concept also becomes increasingly critical when we begin considering diversity beyond the narrow scope of human perspective. When the hardware a perspective operates on becomes a variable, rather than defaulting to biological humans, the potential scope of neurodiversity increases much further. When Scalability is factored in on top of this, the potential scope expands yet again.

I've grown accustomed to hearing objectively false overgeneralizations being applied to ASD or "Autism", even in scientific literature. Some of this activity may be attributed to "Theory-induced Blindness" being applied to theories that use the same category for polar opposites whose scope exceeds the norm.

Naïve researchers who've neglected this have been known to claim that Autism causes reduced emotional intelligence, empathy, and so on. Less biased researchers have recognized that the exact opposite was often true, and that selection bias, subjective validation, out-group homogeneity bias, and related biases were likely underpinning such extreme naivety.

Cognitive Bias in humans increases proportionate to complexity, making the perception of more distant outliers proportionately more biased. However, the technology to overcome this tradeoff has already been demonstrated.

*Note: In Collective Intelligence, there is value derived from perspectives within the space between distant neurodiverse perspectives, as points within the bounded space increase the clarity/granularity of understanding within that space. The broader that space becomes, the more connecting points within it are required to maintain this granularity of understanding.

Consequently, this means that even with scalable and hardware agnostic AGI or ASI the subset where human diversity of perspective is found still offers those systems an insurmountable advantage over naïve and isolated Sci-Fi AGI concepts like Terminator's "Skynet" or Westworld's "Rehoboam".