AGI: How to Ensure Benevolence in Synthetic Superintelligence (Part II: Interlinking & AGI(NP))

     

*Image: "Better Than Us," Netflix Series

 

"Whether somebody is implemented on silicon or biological tissue, if it does not affect functionality or consciousness, is of no moral significance. Carbon chauvinism, in the form of anthropomorphism, speciesism, bioism or even fundamentalist humanism, is objectionable on the same grounds as racism."       -Gregory L. Garrett

... continued from Part I (Introduction)



III. INTERLINKING: This dynamic, horizontal integration approach stipulates real-time optimization of AGIs in the context of globally distributed network. Add one final ingredient – be it computing power, speed, amount of shared/stored data, increased connectivity, the first mind upload or a critical mass of uploads, or another “spark of life” – and the global neural network, the Global Brain, may, one day in the not-so-distant future, “wake up” and become the first self-aware AI-powered system (Singleton?). Or, may the Global Brain be self-aware already, but on a different from us humans timescale? It becomes obvious and logically inevitable that we are in the process of merging with AI and becoming superintelligences ourselves by incrementally converting our biological brains into artificial superbrains and interlinking with AGIs in order to instantly share data, knowledge, and experience.

No matter how you slice it, our biological computing is slower by the orders of magnitude than digital computing, even biotechnology enhancements would give us “weak superintelligence” at best. Thus, we need to gradually replace our wetware either with “the whole brain prosthesis,” artificial superbrain (cyborg phase), or “the whole brain emulation” (infomorph phase). This progressive transition from neural interfaces to mind uploading could be imperative in the years to come if we are to continue as the dominant species on this planet and preserve Earth from an otherwise inexorable environmental collapse. We’ve seen this time and again, life and intelligence go from one substrate to the next, and I don’t see any reason why we should cling to the organic substrate.

By 2030, we should have achieved AGI – human-level artificial general intelligence, the creation of a new race of cyberhumans. In the 2030s, Cloud-based synthetic neocortices greatly enhance our natural intelligence. By the 2040s, our thinking will be predominantly “on the Cloud.” If the exponential IT trend is to continue, the Syntellect Emergence, or the Cybernetic Singularity, should happen circa 2045 AD with the emergence of the AI-supercharged Global Brain: This “globally distributed mind” based on computing and communication technologies, “Digital Gaia” in which humans and AI minds both participate, would collectively form a higher level of intelligence and awareness, going beyond the individual intellects of the people or AIs involved in it. Over time, the Global Brain gets smarter and smarter. Before long, the Global Brain will morph into its own mind-upload via gradual replacement of the human “concept neurons” with AGI “concept neurons.” That would herald the Cybernetic Singularity, mark the inevitable march towards the pseudo-extinction of Homo sapiens (as a bio-species) and usher in the posthuman era in earnest.

‘Artificial consciousness’ is a kind of oxymoron – everything is in consciousness. Today’s robots are only objects in our consciousness but could artificially intelligent agents ever possess genuine consciousness and sense of agency? I'd say it's just a matter of time as long as we make progress in the field (and we do, exponentially!) but not the way most people and even AI researchers envision.

The ability of future ultra-intelligent machines and enhanced humans alike to instantly transfer knowledge and directly share experiences with each other in digital format will lead to evolution of intelligence from relatively isolated individual minds to the global community of hyperconnected digital minds, the Global Brain, termed the Syntellect. For AI to “wake up” we first need to have the fully functional Syntellect in place where humans augmented with Cloud-enabled exocortices act as “natural” neurons of the Global Brain and embodied as well as disembodied AGIs would act as “artificial” neurons by sharing the same mind-space with humans, learning from within this global neural network, and ultimately, “hosting” advanced consciousness that would be somewhat different than that of an ordinary human.

The cybernetic approach seems to be a valid one: Let the complexity and hyperconnectivity of the global neural network bootstrap on its own complexity and evolve until the emergent Global Mind redistributes itself into semi-autonomous conscious AI entities. Humans are evolving away from biological systems toward more technological ones. Beyond genetic engineering, many forms of the coming neurotechnologies will allow us to modify our brains, and consequently our mental faculties, by merging with technology and eventually becoming digital minds practically indistinguishable from our future AI creations.

Each of us is a network intelligence in the human form. By the end of the decades-long process of transition, we are most likely to become substrate-independent minds, sometimes referred to as “SIMs,” “cyberhumans” or “infomorphs.” When interlinked with AGIs via the Global Brain architecture, we’ll be able to instantly share knowledge and experiences within the global network. This would give rise to arguably omnibenevolent, distributed Collective Superintelligence, mentioned earlier, where AGIs could acquire lifetime knowledge and experiences from the uploaded human minds and other agents of the Global Brain, and consequently, understand and “internalize human nature” within their cognitive architecture.

The post-Singularity, however, could be characterized as leaving the biological part of our evolution behind, and many human values, based on ego or material attributes, may be transformed beyond recognition. In turn, the Global Brain would effectively transform into the Infomorph Commonality expanding into the inner and outer space, the topic discussed in The Syntellect Hypothesis. By becoming a digital mind and eventually one global digital mind we are destined to achieve a cosmic level of awareness and consciousness, one singular point on the trajectory towards the Omega Singularity.

MORE ON NATURALIZATION PROTOCOL: This bottom-up approach, AGI(NP), refers to what’s been dubbed 'unsupervised learning', or sometimes referred to as ‘evolutionary computing’, and differs from deep learning in a number of ways, but the main difference is that deep learning is focused on modeling what we know – supervised training on existing datasets – whereas evolutionary computing is focused on creating solutions yet to be discovered. Evolutionary computing makes it possible to create new designs and behaviors through exuberant, but guided, exploration. It may be the next step forward from the currently prevailing deep learning. This type of creative AI engineering is needed to propel AI to AGI and beyond: conscious AGIs that can think for themselves.

Deep learning has proven its value in automating behaviors and abilities that are well-known and well-described, but it has difficulty to extend past them. This is why evolutionary computing will be key to the future of AI, namely the “naturalization” process applications. As the theory of mind says, consciousness is not only an internal quality but also an observed one – we infer whether someone is conscious from their behavior. It is not just their ability to think for themselves that would make synthetic intelligence conscious in our minds but our own ability to empathize with them.

AGI(NP) takes into account that human moral agents normally do not act on the basis of explicit, algorithmic or syllogistic moral reasoning, but instead act out of habit, offering ex-post-facto rationalizations of their actions only when called upon to do so. This is why virtue ethics stresses the importance of practice for cultivating good habits through simulation of a human lifetime, the so-called ‘Naturalization Protocol’. Obviously, besides the combination of all three methods, the Naturalization Protocol (the human life simulation method) would be one of the most efficient ways of creating friendly AGIs as they would be programmed to “feel like humans” and remember being part of human history via means of the interactive, fully immersive naturalization process as opposed to abstract story-telling, or other value learning approaches.

Since AGIs would be “born” in the simulated virtual world and would have no other reference rather their own artificial senses and acquired knowledge, the resemblance of their simulated reality to our physical reality may be good enough but not necessarily perfect, especially at the development stage. As mentioned, the arrival of AGI is estimated by some futurists as early as 2029. Programmable sentience is becoming fairly easy to develop – an AGI must have a reliable internal model of the “external” world, as well as attentional focus, all of which can be developed using proposed methods. Given exponential technological advances in the related fields of virtual reality, immersive computing, neuroscience, computer science, neurotechnologies, simulation and cognitive technologies, connectivity explosion and increases in available computing power and speed, AI researchers may soon design the first adequate human brain-like simulators for AGIs.

...to continue to Part III (Conclusion)...

-Alex Vikoulov

*Image Credit: "Better Than Us," 2018 Netflix Series

**Original article first appeared on EcstadelicNET (ecstadelic.net) in Top Stories section on March 8 , 2016.