Original article was published by Lukas Tassanyi (MicroHash) on Artificial Intelligence on Medium
Time to Build the next Layer of the Matrix
The COVID-19 pandemic has massively accelerated the adoption of digital technologies. We are now more aware than ever about the crucial part technology plays in our daily lifes as we come to terms with the accelerating progress of ever more sophisticated software.
Artificial intelligence is one major theme in this development. Over the last years, it has been touted as the solution to every business problem. Marketing wizards have abused the term as a buzzword to capitalize on the naivety of their customers. And science-fiction has painted a grim future of total machine-domination.
Despite most of the hype transcending rationality, artificial intelligence is real and growing. Fast. Competitive pressures are forcing businesses to adopt the most effective technologies on the market. That results in economic incentives for entrepreneurs to develop ever more sophisticated AI products, again raising the competitive pressures.
This positive feedback loop reinforces a trend that must ultimately result in an intelligence surpassing our own, in the same way Moore’s law yielded machines that can calculate more efficiently than any human could ever dream to do.
Singularity will mark not only a crucial turning point in technology or humanity but in the entire process of billions of years of evolution. It will be another step in the chain of the transformations of life. Up until today, we went from simple molecules over reproducing cells to plants and animals and just very recently arrived at the human species. From our societies, culture and technology have emerged, which have taken on a self-reinforcing evolution of their own — progressing on increasingly shorter and shorter time frames.
Now considering our many imperfections and inadequacies, wouldn’t it be arrogant to assume that we are the crown of creation? For the aforementioned economic causalities, there will come the point in time when humans have to pass on their crown to more sophisticated forms of life.
“Four polls conducted in 2012 and 2013 suggested that the median guess among experts for when they would be 50% confident AGI would arrive was 2040 to 2050”
“The thesis that AI poses an existential risk, and that this risk needs much more attention than it currently gets, has been endorsed by many public figures; perhaps the most famous are Elon Musk, Bill Gates, and Stephen Hawking.”
(Wikipedia: Artificial General Intelligence)
Despite our natural tendency to dislike these predictions due to our drive for self-preservation, we must face the reality that humanity is only a link in the story of evolution and not its final destiny. Like any form of life, we should therefore focus on our teleological obligation to advance and support the establishment of the next stage of life.
To do that, we must build an environment within which the iterational development of artificial general intelligence can take shape in the same way our own evolutionary development took shape on this planet.
That environment must allow the AGI to perceive information and choose between actions, some of which will lead to successful reproduction while others will lead to death. To do that we must simulate a virtual world that emerges from an underlying set of fundamental rules in the same way our universe is built upon mathematical waves as described in quantum theory.
Einstein’s “Special relativity replaced notions of absolute time and space with spacetime.”
“Max Planck, Erwin Schrödinger, and others introduced quantum mechanics, a probabilistic notion of particles and interactions that allowed an accurate description of atomic and subatomic scales.”
“Later, quantum field theory unified quantum mechanics and special relativity.”
“General relativity has not yet been unified with the other fundamental descriptions; several candidate theories of quantum gravity are being developed.”
Deep down there is speculated to be a world formula from which all the physical laws of our universe are derived. In recent decades, physics has been unifying fundamental theories to come ever closer to such a formulation.
In the same way, to simulate a virtual universe within which our artificial general intelligence can live, such a formula must be developed. From that a diverse, physical world can then be generated. Once we allow a basic embodied AGI with sensory and motor capabilities to enter the competitive race for resources and reproduction, computing power is all we need to let it evolve into ever more intelligent forms.
Thus we can reduce the complex problem of creating an artificial general intelligence to three smaller, more comprehensible problems:
- Providing enough computing power for simulating a sufficiently large artificial universe.
- Inventing a world formula that results in the emergence of a universe that balances chaos/diversity with order/continuity.
- Designing an AGI that is capable to reproduce in a certain niche of this universe and capable of evolution.
If we were to create a software framework that uses distributed cloud computing (later quantum computing) to simulate such a universe and put up
- a competition with a prize for the best world formula
- and a competition with financial rewards proportional to the fraction of the universe someone’s initial AGI is capable to inhabit and dominate
then researchers and entrepreneurs would compete in this virtual universe, effectively setting up the economic incentives that would ultimately result in the iterational creation of an AGI.