Ben Goertzel, the creator of this piece, is internet hosting a TNW Solutions session at the moment at 7:00pm CEST — 1:00pm EDT — 10:00am PDT — 10:30pm IST. Ask him a query NOW.

When finalizing a couple of analysis papers for the AGI-20 convention, scarfing up associated data from Arxiv and Google Scholar and GitHub and n-category Cafe — I used to be struck by the apparent realization of how totally completely different this course of was from once I acquired my PhD again in 1989.

I used to be hit by the long-rusted reminiscence of what it was prefer to have some wild, loopy, obscure science or tech thought — after which sit on it and give it some thought for months and even years, with out having anybody acceptable round to speak to about it, and with out having any straightforward manner to determine how authentic the thought was Trying by the analysis literature in large outdated certain volumes within the library was — considerably completely different. Conferences had been nearly the one option to discover out what different researchers had been fascinated by.

The final N years once I give you some wacky thrilling new thought, it’s uncommon that greater than a day passes earlier than I’m deep into the Internet, making an attempt to grasp if my original-to-me line of thought is definitely well-known by some sub-community someplace, simply utilizing completely completely different vocabulary.

As iconoclastic as I’m and as peculiar are a few of my strains of investigation (mathematical fashions of psychic powers! AGI rising from blockchain primarily based networks! quantum logical reasoning engines!) — by the requirements of 30 years in the past, I’m on no account any type of unbiased thinker anymore.

I’ve turn out to be simply one other neuron within the goddamn global brain.

And to be clear, I’ve no intention of going again. Trendy instruments squash some particular person creativity however they simply save a lot time. Which is essential given the cruel finitude of my mortal life — from which my solely likelihood of escape is growing radical longevity tech. Which I’m engaged on within the Rejuve challenge — however I’m far more prone to uncover the important thing to organic immortality if I’m not chasing down dead-end paths.

There may be partial comfort within the data that principally everybody else at the moment is in the identical boat. And I do have the mildly perverse pleasure of figuring out I’m a considerably uncommon type of neuron. I appear to be a neuron hooked as much as channel numerous kinds of anticipatory processes. The AGI idea that I launched 15 years in the past — and am pushing ahead laborious now with SingularityNET, OpenCog and TrueAGI — is now identified to megacorporate CEOs and nationwide leaders, and serves because the mandate of billion greenback R&D initiatives like OpenAI and Deep Thoughts.

The hunt to make use of AI to remedy dying is not so quixotic and marginal, however is the type of factor that will get funding and severe consideration. Even my goofball thought of utilizing mind implants to eliminate pain is starting to seem mainstream. Utilizing AI on genomics knowledge to specialize medical therapies to the person is not sci-fi; my colleagues and I are literally doing this now within the context of some COVID-19 medical trials.

The worldwide mind is evolving in ways in which no one can predict

The worldwide mind’s cognition is getting sooner and sooner, and the time-gap between loopy and apparent has by no means been shorter. A few of my present goofball notions like making use of extended laws of physics to transhuman, transphysical mindspace — or teaching AIs unconditional love to allow them to develop into post-Singularity minds that aren’t solely smarter however extra compassionate and inventive than people — could also be tomorrow’s frequent sense.

The fantastic thing about being a neuron within the world mind is that the worldwide mind is an open-ended intelligence — the precise reverse of the stilted, constrained reinforcement studying AIs so common amongst Large Tech AI groups at the moment. It’s evolving in ways in which no one, together with it, can predict — and redefining itself and increasing its boundaries with each step. An open-ended intelligence tends to develop finest with elements which might be additionally open-ended intelligences.

Whereas I typically assume in any other case upon perusing Reddit posts or YouTube feedback, in the long run being a neuron within the World Mind on the daybreak of the Singularity shouldn’t be so dangerous.

Or a minimum of, that’s what the GB needs us neurons to assume…

Revealed Might 21, 2020 — 12:22 UTC



Source link

LEAVE A REPLY

Please enter your comment!
Please enter your name here