Transcending Biology: Reverse Engineering the Brain
Summary of the Roadmap session of the first 2018 Carboncopies workshop on whole brain emulation
Keith Wiley, 2/13/2018
Will we one day be able to construct a computational model of an individual’s brain such that we can say the working model re-instantiates that person’s distinct mental functions? Will we be able to further interpret that model as a preservation of personal identity and even of life? These two concepts, known respectively as whole brain emulation (WBE) and mind uploading (MU), were the focus of the nonprofit organization Carboncopies’ first workshop of 2018, held on January 28th, titled Transcending Biology: Reverse Engineering the Brain. Carboncopies’ stated purpose is ‘making whole brain emulation possible’. The workshop was freely viewable online by the general public, as well as in person on site in the Bay area. URLs for videos of the individual talks are provided at the bottom of this article. An impressive group of professional scientists, all leaders in their respective fields, gathered on this day to present the state of the art in technologies that represent the nascent stages of eventual WBE.
The workshop opened with an introduction by Dr. Randal Koene, founder and Chairman of Carboncopies. Koene began with a brief explanation of the feasibility of WBE, describing how it is the cognitive experience that matters when prioritizing aspects of the brain, not the underlying neural structure and function of which we are not directly aware. By analogy, internet browsers operate at the level of code and design patterns, not transistors and electrical properties. Brain emulation is, in this way, analogous to emulation of a web browser across varying computing architectures. As to why humanity should desire and pursue WBE and MU, Koene argues that in the future, humans are likely to play a decreasing role in the ongoings of society, government, and the overall range of mental experiences. For humanity to keep up with the inevitable advance of artificial intelligence, we must expand our own mental capacity to access further realms of experience. As to how to achieve WBE from a technical perspective, many of the other speakers addressed current work, but Koene gave an overview of the larger picture and how it may pan out as contemporary methods continue to evolve. It all comes down to structural, but more importantly, functional, modeling of the brain by means comparable to nascent neural prostheses. Such devices, both current and futuristic, operate by modeling, replicating, and ultimately replacing corresponding neural function, one brain region at a time, until in the limit none of the organic structure remains. Note that for the development of a neural prosthesis, functional recording alone may not be enough; an understanding of the underlying circuit structure may be essential to make the modeling process feasible. Structure scanning to discover such neural circuit layout was discussed by subsequent speakers, as presented below. Eventually, this expanding knowledge may reach the point where we can seriously consider emulations of large portions of the brain to be within grasp, and even emulations of the whole brain in the extent of such reasoning. Koene described the goals of Carboncopies as first seeking and proposing methods toward WBE, and second, enabling researchers to pursue projects toward those proposed methods. Finally, Koene announced that Carboncopies was launching, as part of the workshop, a tech review project with the goal of maintaining a living roadmap of projects and progress. This tech review will seek peer-reviewed research into the technical aspects of WBE. Koene and Carboncopies hope that this workshop will be a call to arms for researchers to take up WBE research in earnest in the years to come.
Following Koene’s introduction, the first session of the workshop brought in external speakers to present on various aspects of neuroscience relevant to WBE. The session began with Dr. Kenneth Hayworth from Howard Hughes Medical Institute at Janelia. Hayworth runs the Brain Preservation Foundation (BPF), whose goal is to promote and advance long-term whole brain preservation. The BPF aims for the eventual establishment of brain preservation as a standard medical practice offered in hospitals for life-preserving (or life-stasis) purposes. The BPF is motivated by the likelihood that brain preservation might be achievable sooner than mind uploading, and can therefore be used to bridge the intervening timespan until mind uploading is possible. Early success toward the BPF’s goals has been encouraging, the most noteworthy event being that, in 2016, the BPF awarded its first cash prize for achieving new milestones in brain preservation. This prize was won by Robert McIntyre’s team at 21st Century Medicine for the electron-microscopy-verified (EM) preservation of a whole rabbit brain. Such a feat of peer-reviewed and verified, high-quality preservation of macro-scale neurophil (a whole brain in fact) had never been achieved previously. Not only does the BPF motivate research by others through its cash prizes, but Hayworth has himself pioneered the technological development of slicing and imaging machines for the subsequent scanning of preserved brains. Such imaging is an obvious critical step in one of the more likely mind uploading procedures, a procedure in which a stasis-preserved brain’s structural scan is used as the basis for a subsequent WBE.
Prof. Tony Zador of Cold Spring Harbor Laboratory then presented work using the most recent advances in DNA barcoding to map a brain’s projectome. The projectome differs from the more widely recognized connectome in that the former traces the brain’s general neurite extent, while the latter explicitly captures the individual synaptic connections that underlie the brain’s interconnected network. Zador’s work with DNA barcoding injects randomly generated DNA snippets from a vast library of possible base-pair combinations into localized populations of neurons. The DNA is replicated with a virus so as to fill a given neuron wherever that neuron’s dendrites and axons may stretch, including far across the brain. The brain is then imaged such that each barcode is uniquely exposed via color or direct sequencing, thereby revealing the entire projected structure throughout the brain of the population of neurons hosted locally around the injection site. This technique has already been utilized by the well known Allen Brain Atlas. Zador further presented a nascent extension of this work that offers the possibility of using the same basic technique to map connectomes.
Dr. Adam Marblestone, of MIT and Kernel, presented optical methods for molecular connectomics. This technique combines expansion microscopy, DNA barcoding, and in-situ fluorescent sequencing for the Rosetta Brain Project. Expansion microscopy is a clever method for increasing spatial resolution. It literally swells the brain so as to make the smallest features more resolvable than they would be at their original scale. While electron microscopy can achieve nanometer resolution, it cannot penetrate deeply into tissue, nor can it benefit from various optical imaging advantages, such as color labeling. Alternatively, optical imaging (classical microscope imaging) penetrates deeply and can utilize color, but cannot achieve the same resolution as electron microscopy. Expansion combines the best of both, bringing small features within reach of conventional optical microscopes such that DNA barcoding can be utilized. Another application of DNA barcoding is to assist the error-correction stage of neurite tracing, in which a neuropil volume is slice and scanned, and neurite structures are reconstructed by tracing through the scanned layers, classically with electron microscopy. This practice suffers in regions where neurite details get too small or neurites from multiple neurons become excessively intertwined. Appending such a practice with barcoding could greatly assist in resolving the errors that occur in such regions.
Prof. Theodore Berger, from the University of Southern California, gave a wonderful presentation of the current state of the art in neural prosthetics. Berger’s work focuses on the hippocampus, a deviously challenging region since it resides near the center of the brain, inside and beneath other regions. The hippocampus mediates the conversion of short term memories into long term memories, and damage here is implicated in some of our most problematic brain maladies, such as Alzheimer’s, dementia, blunt force trauma, and stroke. Recent work has implicated depression in hippocampus defects as well. Affected patients have functional short term memories and retain earlier long term memories, but are weakened or prevented from forming new long term memories. Berger’s prosthesis applies electrodes to two regions of the hippocampus, loosely labeled as the input and output areas. It learns the neural spatio-temporal firing code from which it is possible to learn the transformation function performed by these regions of the hippocampus and can actually take over or strengthen this transformation for a damaged hippocampus. In this way, the prosthesis revitalizes a patient’s ability to form new long term memories. The work is still in the research stages, but has proceeded with great promise. Further advancements are expected, as is eventual standardization of a medical application. The implications are profound to say the least. This is precisely the sort of work that gives us a glimpse into the possible future of neural prostheses. One needs only a decent imagination to extrapolate such to the wilder possibilities. Namely, taking this concept to completion by replacing every brain region with an equivalent prosthesis is, in fact, whole brain emulation.
Dr. Shawn Mikula, from the National Institute for Physiological Sciences in Okazaki, Japan,presented his work on taking serial section EM further than its historical applications. EM is generally limited to volumes on the order of one millimeter cubed, but Mikula has extended the technique to rodent brain scales with impressive results. Using a machine based on Hayworth’s design, he has sliced and image a whole mouse brain using EM techniques. Mikula hopes to be able to image macroscale whole brains in a matter of weeks, which represents orders of magnitude speed up over the current estimates of years.
The impressive work these researchers have done gives a good survey of the most advanced techniques currently available. It was a great pleasure not only to see presentations on what is possible today, but to hear these experts speak on the likely paths of advancement over the coming decades. The expectations are universally optimistic; we should expect to see these methods expand in capability, accessibility, and generality throughout the twenty-first century. The conclusion is clear. Preserving, scanning, imaging, mapping, and functionally reproducing the brain, better summarized as whole brain emulation, is a technology that is practically destined to come to fruition given sufficient interest and funding, and commitment.
Keith wiley serves on the board of Carboncopies as director of communications. His book, A Taxonomy and Metaphysics of Mind-Uploading, is available on Amazon.
The workshop’s archived URL is https://www.carboncopies.org/workshop-2018-jan and includes links to the videos. The same URLs are offered below in chronological order:
- Dr. Randal Koene — https://www.youtube.com/watch?v=AGIkkXIvBy0
- Dr. Kenneth Hayworth — https://www.youtube.com/watch?v=yD1mivVKRrs
- Prof. Tony Zador — https://www.youtube.com/watch?v=LIAMYMJFHPk
- Dr. Adam Marblestone — https://www.youtube.com/watch?v=D9EvSK--b04
- Dr. Shawn Mikula — https://www.youtube.com/watch?v=AhYzsnxEv0U