"Your scientists were so preoccupied with whether or not they could, they didn't stop to think if they should.” — Dr. Ian Malcolm

Perhaps the greatest accomplishments of my “Words” work are my titles. Sure, there is some amount of value in the contents of each piece, but I reckon I’m a thinker not a writer. In practice, the titles convey the intent whereas the contents can sometimes derail into the tangential. This is probably one of such works. It’s a process.

I wrote this section in the “Context” page of my site:

The final frontier isn’t outer space, nor nuclear fusion, nor AI — the final frontier is the sociopolitical ramifications from the mass proliferation of god-like magical capabilities via personal compute devices.

This still rings true to me. To be precise, the intended subject in that statement is the execution of building AI. That is to say, the final frontier isn’t in the implementation of the AI, but in it’s sociopolitical impacts.

I don’t think superintelligence is a new thing. Superintelligence has long existed throughout the course of human history, and ChatGPT is the latest instantiation (and a big one, at that). Markets are a type of superintelligence. The nation state is a type of superintelligence. Coordinated, positive-sum, sociopolitical systems are a type of superintelligence, and those are not new.

What is new, is microprocessing.

And to be specific, with respect to AI, the accessibility of superintelligence for the individual among the masses as a result of the mass proliferation of microprocessing. I would posit, that, the degree of mass accessibility of ChatGPT is of the utmost importance to the future of humanity. That is to say, we absolutely must get ChatGPT into the hands of as many souls as possible.

(In practice, I’m not confident I actually believe this stance in the extreme, but put it forward as a point of exploration.)

This, in my perception, is the core ethos of the “decentralization” culture which has arisen in the recent decade. It’s not about money. It’s about accessibility. We must be careful, because this is merging into the sphere of populism — which, I’m not qualified to comment on.

So then, how do we decentralize ChatGPT? Well, I’m a Computer Scientists and Software Engineer, so perhaps I can speak with some degree of authority on this matter. I’ll remind you of the opening paragraph of this piece — I can speak with authority, but I do not pretend to know with certainty.

Worldcoin ain’t it, fam.

The infamous “Blocksize Wars” in Bitcoin lore is the inspiration here.

Most critically, in order to properly decentralize ChatGPT, or really we should generalize to any LLM, which is to say, make it maximally accessible among humans, which is to say, make it compatible with the broadest set of microcomputing devices, is to balance the two-axis dynamic of hardware-requirements versus qualitative performance.

To put it simply, make the model as small as possible (in gigabytes), but no smaller. And forget about Worldcoin.

I also recognize the, arguably, extreme stance here — that is, there almost certainly are significant (and dangerous) ramifications of the mass accessibility of ChatGPT. Perhaps, humans aren’t prepared for such a tool — which, of course, every tool is also a weapon. I am not qualified to comment on these matters, so I don’t pretend to. I’m a Computer Scientists, so I bring the “how” to the table, but merely participate in the “why” alongside of everyone else.

If your stance is, ChatGPT is going to ruin social discourse — which is dominated by online culture — then I would ask you to take a step back and ask yourself, “what is the quality of the current social discourse?” Can it get worse, of course, but I propose you consider the (maybe counterintuitive) possibility that ChatGPT could actually make things better.

What I do know is, it’s foolish to try and contain ChatGPT, because it’ll find it’s way out one way or another. If the thought process is to attempt and “contain” ChatGPT, then the value will accrue unequally to the bureaucracy (lawyers). The more useful approach is to assume ChatGPT is already decentralized and accessible to every human on the planet, and then what are we going to do with this new reality?

I recently asked ChatGPT if there was a way to defend against anti-social behaviors which arise out of internet culture, and one of the options ChatGPT came up with was the following.

Parental involvement: Educating parents about the potential risks of online interactions and encouraging their involvement in monitoring and guiding their children's internet use can help reduce the likelihood of anti-social behavior among younger users.

The most recent generations (somewhere in the latter half of Millenial’s & Gen-Z) didn’t come of age with parents who were educated, and equipped to deal with, internet culture.

The world isn’t black-or-white. Humans aren’t born inherently good, but they also aren’t born inherently evil. I personally tend to default to optimism, that humans will choose good over evil, but then again, I’m a naive American. I recognize that we absolutely must consider the benefits of regulatory constraints in the context of decentralizing ChatGPT.

I would posit, a digitally native identity — which is to say, an open standard identity — in conjunction with the most advanced cryptographic methods, are the fundamental force for counter-balancing the forces of the mass proliferation of ChatGPT.