January 11, 2019 | CogWorld on FORBES
The evolution of artificial intelligence (AI) — from artificial narrow intelligence (ANI), through artificial general intelligence (AGI), to artificial super intelligence (ASI) — is on its way to changing everything. It's expected that soon, artificial intelligence will combine the intricacy and pattern recognition strength of human intelligence with the speed, memory and knowledge sharing of machine intelligence.
As the rise of AI continues, AI is challenging and changing not only the way humans live, learn and work, but also how entities across nations: its government, industries, organizations and academia (NGIOA) construct their commercial and economic industries and markets. With this technology driven growth of artificial intelligence, the need to do most manual, mathematical and mundane work is already in decline and will likely be greatly diminished in the coming years. Moreover, with all these new digital assistants and decision-making algorithms assisting and directing humans, more complex day-to-day work for humans is being greatly lessened.
While attempts to create artificial intelligence have fueled many applications and technological advances, it's fundamentally the increased processing capabilities that have made it possible to design intelligent machines that can rapidly compute and make intelligent implications from diverse data inputs. Despite these rapid advances in processing capabilities, human-like artificial intelligence, however, remains hard to determine—as it still requires further understanding of how to replicate human brain performance and processing in computers. While many computer scientists believe that mimicking the processing method of the human brain will be the key to achieving human-like artificial intelligence, that alone is perhaps not enough, and many other variables will need to be evaluated, like memory and conscience.
Moreover, since current computer processing systems are not efficiently able to process the massive amounts of diverse data and information required to achieve human brain-like processing power, there is perhaps a need for developing a fundamentally new computing system and new computing hardware and software infrastructures that are mathematically correct, adaptable, energy independent and secure.
Towards Artificial Super Intelligence
The potential development of artificial super intelligence points to a frankly scary scenario in the coming years. Eventually, the processing power of the human brain may not be able to match that of ASI, and the risk manager in each one of us must begin thinking about the risks that arise from a situation in which humans are competing -- rather than collaborating -- with AI.
These risks to natural human intelligence are compounded when one remembers that AI will reach a point where it will be able to improve its own software design and capabilities far beyond what its designers envisioned. Already, AI researchers are working to develop this ability for self-improvement in more benign situations, such as photo recognition and games. Admittedly, hardware would likely also need to be evolving simultaneously, but this potentially exponential growth in AI warrants greater caution.
These fears are heightened by the reports emerging about how the structure of the adult human brain changes when a new cognitive or motor skill, including vocabulary, is learned. As a result, there is a growing concern that when AI takes over most human tasks and intelligence work, this may change the level of human intelligence, and the nature of natural human intelligence as well. If the human brain is not allowed the opportunities to learn new skills, how will its development suffer? That brings us to the question:
Enhance Human Intelligence or Artificial Intelligence?
Natural human intelligence has always seemed to be one of humanity's greatest weapons against its enemies. Therefore, as artificial intelligence matches and surpasses human intelligence, the question is whether natural intelligence will be able to compete with artificial intelligence in the coming years.
The answer to this question will define the future of humanity, as at some point, when and if artificial intelligence becomes an enemy to humankind, natural human intelligence may or may not be enough to defeat artificial intelligence. So, although artificial intelligence is becoming more and more a part of our everyday lives, the question everyone across nations needs to evaluate today is whether our efforts should be towards enhancing human intelligence or artificial intelligence.
So, what can humans do as a species to proactively plan against that scenario? In the below video, I interviewed Prof. Risto Ilmoniemi, Head of the Department of Neuroscience and Biomedical Engineering, Aalto University School of Science, Finland. They need to focus on the science of intelligence. Risk Group initiated this much-needed discussion on Science of Intelligence on Risk Roundup.
Disclosure: Risk Group LLC is my company.
Prof. Risto Ilmoniemi, Head of the Department of Neuroscience and Biomedical Engineering, Aalto University School of Science, Finland participates in Risk Roundup to discuss “Science of Intelligence: Human Intelligence to Computational Intelligence.”
About Human Intelligence
From the beginning of human time, the power of imagination has helped humans explore the unknown, evolve and change the world. As seen over the years, humans have fundamentally transformed geospace with ideas, imagination and innovation, and that same power of imagination will likely help humans explore the unknown of cyberspace, space and the spaces we haven’t even defined yet, and thereby change the realities of the human ecosystem in the coming years. Imagination has always been an indicator of human intelligence. In fact, imagination is a creative power that is necessary for inventions in cyberspace, geospace and space (CGS)—and the same power of imagination is driving scientists today to better understand human intelligence.
From what is known about humans, natural intelligence evolution has proceeded by natural selection -- from random mutations -- and by greater societal emphasis on education and learning. Some studies indicate that heritability (genes) may account for up to 50 percent of intelligence levels; and the role of the environment (education, environment, experience, resources, etc.) contribute to the rest. While much about the human brain genome is already decoded, we still don’t have a clear understanding of how exactly to enhance human intelligence. The reason behind this limited understanding is that the human brain is incredibly complex, with a network of perhaps 100 billion neurons that function individually and collectively, and as a result, understanding of the human brain remains difficult.
However, recent research by Joe Tsien and his team from Augusta University in Georgia have speculated that there must be a basic design principle from which human intelligence originates and the brain evolves—and that one simple algorithm can explain human intelligence. So, if the origins of human intelligence are based on a fundamental algorithm, it is necessary to understand and evaluate whether brain computations rely on relatively simple mathematical logic or a more complex process. Moreover, if human intelligence is indeed based on a fundamental mathematical logic, then the current advances in neuroscience and computer science give hope to evolving human intelligence at the speed of AI’s evolution due to exploding processing speed.
Having said that, knowing what determines intelligence is perhaps still the biggest question facing humanity today. Given there are many theories of intelligence, a question that remains is whether intelligence is a single general ability with a single point of origin or a set of many abilities with multiple, complex points of origin.
While there is a growing concern that human intelligence evolution is stagnant, science is now perhaps getting closer to entering a new phase in which it will be possible to change and improve human DNA. While neuroscience still has a long way to go, and this is most likely decades away, these advances in science and technology will almost certainly discover how to modify and enhance natural human intelligence eventually.
Now, irrespective of human intelligence or AI, intelligence is believed by some, such as Joe Tsien and his team from Augusta University in Georgia, to be having a very simple mathematical pattern underlying it. So, in theory, if the information and processes of the human mind can be disassociated from the biological form, it is no longer tied to biological limits such as lifespan or memory lapse and aging processes. Furthermore, information and knowledge stored within a human brain or information within a computer’s memory could then be copied or transferred to one or many other recipients, whether they be digital, robotic or biological.
Although there are many efforts going on across nations to understand the origin of intelligence within the human genome, it remains to be seen how effective gene editing or any other technology can be at influencing intelligence. The time is now to use all available resources for ensuring that this is not the end of the growth of natural human intelligence but rather just the beginning. So, even though, artificial intelligence is becoming more and more a part of our everyday lives, understanding how the human brain produces intelligent and self-aware behavior is still perhaps one of the greatest challenges facing science and technology today.
To reach the desired end in intelligence evolution requires an evolution of both natural and artificial intelligence: collective machine intelligence that can help solve complex challenges facing humanity and human intelligence evolution that can be ready to take on artificial intelligence if it goes against humanity. After all, a conscious AI may not have the same priorities and values as does humanity -- who knows if advanced AI will value human life?
Humans have no choice but to learn to think in a new way to survive as a species as artificial intelligence continues to evolve exponentially (and as of now, natural human intelligence remains limited and fixed). It is more important than ever that natural human intelligence evolution also happens simultaneously, and it is within human control and power to make ourselves more intelligent.
Jayshree Pandya, Ph.D., contributor, is Founder of Risk Group & Host of Risk Roundup. Jayshree Pandya (née Bhatt), Founder and CEO of Risk Group LLC, is a scientist, a visionary, an expert in disruptive technologies and a globally recognized strategic security thought leader and influencer. She is actively engaged in driving the global discussions on existing and emerging technologies, technology transformation and nation preparedness. Her work focuses on the impact of existing and emerging technological innovations on nations, nation preparedness and the very survival, security and sustainability of humanity.