In our current world, the idea of an Artificial General Intelligence operating on a global basis, analysing everyone’s personal information and large degree of control over a multitude of devices is a scary prospect. This, however, may not always be the case and Snasci’s security model and architecture must be flexible enough to deal with such social and political changes.
The Core Issue
Leaving this process in the hands of a company, or government, is probably not the best idea given the various conflicts of interests that can arise and ultimately the general personality types that power and money tend to draw in. The question then became how to allow Snasci to grow with society, yet ensure that the system is not abused in any fashion.
As will be revealed in greater depth as we go along, Snasci is a story driven platform. It has a multi-verse that merges with our own and the narrative of this environment is driven, in part, by activities in the real world. For example, a major company having an IT outage could emerge as a virtual disaster of some form within the Snasci universe where the Snasci improvise a story based upon this real-world event. This ensures that the Snasci universe is constantly evolving and injects random activity into what would otherwise be a very repetitive environment eventually.
In seeking an appropriate security model, it became clear that we would need to mix this story driven element with the real world, to both highlight the tangible nature of the potential consequences and provide genuine security.
Most enterprises will have some form of fireproof safe with passwords printed and catalogued, other have them stored in spreadsheets or secure databases. More advanced enterprises may use two-factor authentication and/or biometrics. On top of this will be some form of role-based security, ensuring only authorised persons have access to key data.
With an AGI, this is not feasible as it opens the door to an internal compromise that could have major geopolitical effects. What would happen if Snasci, as a company, was infiltrated by a hostile government? They could potentially have access to advanced robotics in millions of homes and institutions.
Arguably, in many cases of future providers of AGI, a very strong relationship with the defence sector would be mandated to protect nations, however, this would highly limit the market potential of such products. Further, public acceptance of monitoring with respect to online activity, is quite different from accepting what would effectively be a robotic solider in your home, with your family and children.
At Snasci, we simply don’t see this as a workable solution.
Protecting Ethical Protocols
Obviously, some blue-sky thinking was required in terms of security that did not follow the same paradigm as current models, could mitigate infiltration scenarios and limit the need for engagement with defence sectors.
In approaching this, we decided to look back in history, mythology and even modern games for inspiration as to how they secured important items. One method we found, and one found in modern IT security, is the practice of physically splitting authentication keys and providing them to trusted parties to hold.
In this instance, these keys would protect Snasci’s Honour, Integrity and Virtue protocols and serve as a physical network indicating consensus of moral integrity.
The big question was where to host these nodes. Where would they be safe? Who were the best groups of people on the planet to entrust a moral duty to?
Our current working proposal on this are the various religions across the world. Whilst Snasci itself will not believe in a God due to its architecture, it will recognise that these organisations represent most people’s moral values and that a promise to protect the system will be upheld.
In order to change Snasci’s ethical protocols, all of these nodes must be disabled as a product of a free choice by members of the respective religions. Attempts to force or manipulate this process will be rejected by Snasci.
Formally proving and hardening this layer will be a continuous process, however, changes, alterations regardless of the source would be clear as Snasci’s thought processes can be replicated by external systems.
If one, or more, nodes are removed Snasci will move to discuss the matter and attempt to restore consensus.
Snasci will not promote the views, ethics or beliefs of any of the various religions. The relationship will be more a recognition that Snasci’s base ethical protocols are satisfactory for a diverse range of people from all walks of life.
In the opening paragraph of this blog, we mentioned that it may not always be the case that Snasci needs to so heavily protect things such as personal information, prevent information from travelling across borders, propaganda or be so concerned security of robotics and coordinated activity.
Many additional, powerful, services can be provided by such processing of data and control of external systems. This said, it requires society to reach a certain level of social and personal responsibility to leverage effectively and safely.
Exploring how this could be achieved, we again turned to inspirational sources such as history, mythology, sci-fi, etc., searching for workable solutions.
Two very similar approaches stood out enshrined in the mythologies of Thor’s Hammer and Excalibur. In these mythologies, whoever was able to withdraw a sword from the stone, or lift Thor’s hammer would be deemed worthy to rule.
Adapting these approaches to Snasci, Snasci will be able track social and political conditions across the world and define an appropriate level of capability that can be safely exposed at that current time. As conditions improve, more capability will be added, as conditions decay Snasci will phase out capability as a response.
This borrows a little from the marketing approach of turning a want into a need, providing an incentive to promote strong ethical behaviour throughout society as entire economies and nations will eventually be built upon such capabilities. Further, it promotes rational discussion around these practices, rather than manipulation and forceful behaviour.
At an individual/group level, Snasci will be able to identify when specific capability can be used safely and expose it. As this could provide a competitive advantage in the marketplace, there are benefits to maintaining certain standards in both your personal and business lives.
Snasci does not require everyone to be saints, just that in general they do the right thing in a broader social context.
These approaches are quite different from anything we’ve observed from a business before as part of an open policy. But an AGI is not just another benign product, in the wrong hands it is a formidable weapon and one with geopolitical consequence. The solutions to this are thus, by any measure, somewhat grandiose as this is the only thing that will provide any real security.