Many believe collaborating with them will improve human performance. But that assumes they are motivated to see value in partnerships with humans. JL
Jayshree Pandya reports in Forbes:
Technology triggered intelligence evolution in machines. The linkages between ideas, innovations and trends have brought us on the doorsteps of singularity. There is no direct evolutionary motivation for an AI to be friendly to humans. Artificial intelligence will have no inherent tendency to produce or create outcomes valued by humans-- and there is little reason to expect an outcome desired by mankind from any super intelligent machine.
As humanity stands on the brink of a technology triggered information revolution, the scale, scope and complexity of the impact of intelligence evolution in machines is unlike anything humankind has experienced before. As a result, the speed at which the ideas, innovations and inventions are emerging on the back of artificial intelligence has no historical precedent and is fundamentally disrupting everything in the human ecosystem.
In addition, the breadth, depth and impact of this intelligence evolution on furthering of ideas and innovations across cyberspace, geospace and space (CGS) herald the fundamental transformation of entire interconnected and interdependent systems of basic and applied science: research and development, concept to commercialization, politics to governance, socialization to capitalism, education to training, production to markets, survival to security and more.
The technology triggered intelligence evolution in machines and the linkages between ideas, innovations and trends have in fact brought us on the doorsteps of singularity. Irrespective of whether we believe that the singularity will happen or not, the very thought raises many concerns and critical security risk uncertainties for the future of humanity. This forces us to begin a conversation with ourselves and with others (individually and collectively) about what we want as a species.
While there is no way to calculate just how and when this intelligence evolution will unfold in machines, one thing is clear: it changes the very fundamentals of security, and the response to it must be integrated and comprehensive.
Acknowledging this emerging reality, Risk Group initiated the much-needed discussion on Artificial Intelligence (AI) Driven Technological Singularity with David Wood on Risk Roundup.
While we humans have changed our ecosystem numerous times, intelligent machines are expected to change and expand the human ecosystem in unprecedented ways. As seen across nations, the connected computers, information, communication, digitization technology and internet has fundamentally disrupted the information process created by humans, which has given rise to an information age.
The success of the rapidly progressing information age is based on data, and data and information are all around us. When we evaluate the information and intelligence evolution in machines, we realize that it has transformed human life in CGS through fundamental innovations in how we generate digital data and information, how we capture and collect the digital data, how we store it, how we retrieve and replicate the information—thereby blurring the boundaries of languages, time zones, politics, ideology, race, religion and culture.
As seen, the quantity and quality of digital data and information created and stored by humans is multiplying rapidly in cyberspace (reaching about several quintillion bytes of data already). As of now, the current human population is around 7.7 billion, and the number of humans getting connected keeps growing. And with growing human connectivity, the data and information keep growing as well, feeding machines what they need to allow them to see patterns, learn and make them more and more intelligent.
Progress in Machine Learning Algorithms and Neuromorphic Chips
On the back of the ongoing information revolution, as the machine learning algorithms improve rapidly, the power and promise of software that learn by example, patterns and models seems immense. Along with the impressive advances in software development, there also now seems to be a parallel evolution in computer hardware to enhance machine intelligence capabilities. This is clearly visible in the intensification of on-going efforts towards developing systems on a chip, focused on redesigning more efficient, lower energy consuming microprocessor chips that mimic the human brain circuitry. As these rapidly evolving neuromorphic chips are being designed to process human sensory data such as images, smell and sound and to respond to changes in that data in ways not specifically programmed, a lot is expected to change for machine intelligence and artificial intelligence evolution. This is mainly because any effort or initiative in shrinking down the power of a neural net (based on a human brain or octopus brain) onto a single semiconductor chip means that these learning, modeling and pattern recognition algorithms and technologies can now be embedded into a wider range of systems in future—thereby increasing data and information capabilities for exponential growth in machine intelligence. This is expected to fundamentally change how we gather information and intelligence.
As a result, as the computing power of the rapidly evolving computers will exceed that of even the most intelligent and evolved human brain, the exponential growth in machine intelligence will continue towards singularity. Then artificial super intelligence will be just around the corner.
There is no doubt that when a super intelligence emerges through artificial intelligence, it will bring to bear greater problem-solving and inventive skills than current humans are capable of. But would that not also mean creating another species with intelligence that may or may not have human interest at heart? What happens to human intelligence and the human race at that point of singularity?
Keeping up with Super Intelligence
This brings us to an important question: amidst the rapidly evolving and converging technologies, when the intelligence explosion seems inevitable, how will humans keep up with the super intelligent machines?
Now, there are those who believe that to overcome artificial super intelligence, emerging methods can be used to enhance human intelligence and create a superhuman with super intelligence. While in theory, creating a superhuman with super intelligence is likely possible through intelligence amplification of human brain and / or intelligence augmentation (through advances in bioengineering, genetic engineering, nootropic drugs, mind uploading and even direct brain–computer interfaces, AI assistants and more), the reality is evolution of human
interfaces, AI assistants and more), the reality is evolution of human brain and human intelligence is a very complex endeavor with too many unknowns, dependencies and variables.
While evaluating superhuman intelligence is not the focus of this article, the possibility of human brain and intelligence evolution needs to be researched further to keep up with artificial super intelligence.
Economic Impact of Singularity
Economic progress has often been driven by some sort of automation, and AI has begun to increase automation in the production of goods and services across nations and across jobs. So, the question emerges, what happens if everything can be automated — that is, if people and processes can be replaced by AI, what would economic growth look like?
While there can be no clear timeline or consensus on when super intelligence is likely to be achieved, one thing is clear: that the troubling trajectory of technological singularity forces us to think seriously about what we want as a species. Irrespective of whether the singularity is driven by artificial intelligence or any other technological means, it is bound to trigger a technological tsunami, resulting in unfathomable changes and challenges to human civilization and its ecosystem in cyberspace, geospace and space.
Singularity and Security Risks
Since there is no direct evolutionary motivation for an AI to be friendly to humans, the challenge is in evaluating whether the artificial intelligence driven singularity will -- under evolutionary pressure -- promote their own survival over ours. The reality remains that artificial intelligence evolution will have no inherent tendency to produce or create outcomes valued by humans-- and there is little reason to expect an outcome desired by mankind from any super intelligent machine.
We humans are living a paradox as the achievements of artificial intelligence advances are shaping human ecosystems with more dangerous and more valuable opportunities than ever before. Whether promise or peril prevails will define and determine the future of humanity.
This article is not to make timeline predictions to a singularity but rather to begin the discussion on the troubling trajectory of artificial intelligence evolution for the future of humanity. Whether we believe that singularity is near or not, the very thought raises crucial security and risk questions for the future of humanity, forcing us to think seriously about what we want as a species.
0 comments:
Post a Comment