Home Things What is Artificial Intelligence Whole Brain Emulation and Why You Should Care?

What is Artificial Intelligence Whole Brain Emulation and Why You Should Care?

Photo: needpix.com

If artificial intelligence still sounds like science fiction, then you are certain to miss out on the investment, educational, and vocational opportunities generated by this field of research. There is a fair amount of disagreement among scientists as to when and in what form artificial intelligence will be achieved. Some suggest as soon as 20 years while others are looking 80 years out. But all agree that artificial intelligence is coming, and it may include Whole Brain Emulation.

Whole Brain Emulation

Whole Brain Emulation is the hypothetical process of uploading the data of an entire brain into an artificial computer system. Whole Brain Emulation is hypothetical more than theoretical because the science behind it is unknown but highly sought. Whole Brain Emulation has been popularized in science fiction for decades. Dr. David “Dave” Bowman, the lead character in 2001: A Space Odyssey (1968), undergoes uploading in later books. In the movie Tron (1982), Flynn’s consciousness is uploaded by an artificial intelligence called the “Master Control Program.” More recently, in movie-flop Transcendence (2014), Dr. Will Caster (Johnny Depp) uploads his mind into a quantum computer. And in Children of Time (2016)–a novel by Adrian Tchaikovsky–both Dr. Avrana Kern and Gilgamesh Captain Vrie Guyen experiment with Whole Brain Emulation to varying degrees of success.

The Benefits of Whole Brain Emulation

The greatest benefit for society of Whole Brain Emulation comes from capturing and retaining an essence of every human being. The death of the greatest contributors to science, art, music, philosophy, religion, and technology is an irreplaceable human loss. The ability to prevent this loss through Whole Brain Emulation guarantees a continuation of knowledge, experience, expertise, memories, and ideas.

Individually, the perceived benefit is immortality. However, this comes at a cost. Most conceivable methods of Whole Brain Emulation involve the destruction of the original copy. At best, the neurocomputational structure would be a mechanically precise copy. The essence of memory would be retained, but personality would be lost. At worst, the copy would assume its own essence. That is to say, from the moment of the successful copy, the original and the copy would exist as fully, differentiated identities: an electronic doppelganger. Both would see themselves as the original. At the moment the copy “awoke,” their experiential histories would branch irreparably. When the human individual died, what remained–however extant–would be only a copy or a duplicate.

The Negative Effects of Whole Brain Emulation

Due to the invasive nature of Whole Brain Emulation, the original host will almost certainly die in the process. Some have suggested applying the Whole Brain Emulation process at the point of physical death. As mentioned, what survived would seem like the original to external observation. However, it would only be a copy of the original, not the original itself.

Should Whole Brain Emulation become possible, another foreseen negative effect would be the disequilibrium imposed upon the neurocomputational copy. Sensory input would be non-existent or at least radically altered. For example, there would be no physical sensation from touch, taste, sound, sight, or smell. Should full self-awareness be attained through Whole Brain Emulation, the extant copy may not survive due to the psychological trauma subjected in the utter loss of sensation. In response, some have suggested that by the time Whole Brain Emulation is possible, virtual reality simulations would also be possible, creating a functional model of the real world for the uploaded brain.

Finally, Whole Brain Emulation would mean a reduction in human employment. If a human original were compensated for a specific, technical task, her death would require hiring, training, and fairly compensating a replacement. However, if a copy of her brain were uploaded, she could continue the work without any of these expenses.

The Challenges of Whole Brain Emulation

In order to achieve Whole Brain Emulation, scientists would first need to be able to capture a sufficiently detailed scan of the human brain. The information or raw data from this scan would then need to be fed into a computer. Nick Bostrom, in his book Superintelligence, describes this process as a computational reconstruction of “the three-dimensional neuronal network that implemented cognition in the original brain” (30). This process would require a computer sufficiently powerful enough to contain, retain, and sustain the neurocomputational information structure.

Many biotechnical scientists believe this process is attainable, though far off in the future. Even then, degrees of success vary based on the quality of the scan. As Nick Bostom explains, a high-fidelity emulation would retain the full set of knowledge, skills, capacities, and values original in the scanned brain. However, distorted or “generic” emulations would capture only fragments of the original host.

The hardware design, structure and computational power necessary for Whole Brain Emulation are orders of magnitude greater than the fastest computers in existence today. For comparison, in 2014, the Japanese group RIKEN captured one-second of activity from one-percent of a human brain. The computer involved contained 705,024 processor cores and 1.4 million GB of RAM. It still took 40 minutes to process the resulting data. Further significant advantages will required exascale computers, capable of carrying out a quintillion (ten to the power of 18) floating point operations per second.

Because there are no sufficiently complex models for operational efficiency, it is unknown how much energy would be required to power a computer capable of capturing the entirety of a human brain that, in addition to being an exact copy, retains the capacity to think, change, and develop.

Advances in Whole Brain Emulation

Blue Brain is a Swiss-based research initiative seeking to develop the necessary underlying software tools and workflows necessary to integrate a complete process of both building and simulating a computational reconstruction of the brain. The Blue Brain project recognizes there remain necessary advances in a comprehensive understanding of specific features of the brain from “interdependencies inherent in the anatomy and physiology of the brain.”

IBM is conducting Brain Emulation research via their Watson Research Blue Gene supercomputer.

Elon Musk’s company Neuralink is working on brain-computer interfaces, a complimentary technology that will inform future emulation.

The BRAIN Initiative (Brain Research through Advancing Innovative Neurotechnologies) is a US based collaboration of public and private research agencies. It was an initiative started by the Obama Administration. While Whole Brain Emulation is outside the purview of the group’s work, its research in understanding the complexities of the brain is fundamental for Whole Brain Emulation.

What’s Next?

Whether or not Whole Brain Emulation becomes a reality, the implications of artificial intelligence on human life are increasing in magnitude and scope. For a more comprehensive understanding of the technological and human benefits and challenges facing Whole Brain Emulation, read “Whole Brain Emulation: A Roadmap,” by Anders Sandberg and Nick Bostrom.