SIM Video Podcast Interview with Dr. Randal A. Koene

In December 2011, Dr. Koene gave an extensive video podcast interview to Nikola Danaylov, also known as Socrates, focused on clarifying concepts, separating fact from fiction, and laying out the concrete foundations of work towards substrate-independent minds (SIM). You can read Socrates' feature presenting the interview recordings at: http://www.singularityweblog.com/randal-koene-on-singularity-1-on-1-mind-uploading-is-not-science-fiction/

The presentation below is structured slightly differently to simplify the discovery of specific topics, questions and answers in each of the six recordings. First, we will provide an overview of the contents, followed by each of the recorded pieces and their respective highlights.

Content summary:

Part 1: On the origins of Dr. Koene's interests and the realistic expectation of SIM

Q. How did Koene become interested in nanotechnology and neural engineering?

In part 1 of the interview, Koene describes a long-time involvement with the topics that began with a personal epiphany at the age of 13: We need more time, more capabilities; we need to enhance ourselves. Koene mentions that Arthur C. Clarke's "The City and the Stars" influenced his thinking: We are, in the sense of what we consider "being", information and its processing. Originally, Koene considered storage and reconstruction at the particle level, a physics-based approach that could demand the ultimate capabilities envisioned for nanotechnology. Concentrating on the overriding importance of the mind and therefore the brain, Koene found his way into computational neuroscience and neural engineering. Organized efforts began around 1994 through the so-called "mind uploading research group" (MURG).

Q. What is fascinating about the science involved?

Koene points out that we build a model of the universe in our minds. Our lives take place in this model translation of that which is. Different species probably do this differently and to different degrees.

Q. Can we make a model of the brain, of who we are?

Koene explains that while our minds cannot hold a complete functional model of the brain, we can certainly create computational models of that. Our scientific understanding does not yet extend to the full hierarchy of strategies employed by the mind, but we do understand principles of the underlying neurophysiological mechanisms from which mind emerges.

Q. How realistic or unrealistic is mind uploading?

Koene points out that the brain is a biological machine and explains that we can make a replica of the mechanisms in another device.

Q. Are replicas that do work for us science fiction?

No. Dr. Koene clarifies that it is not SciFi in the way in which that often implies something highly improbably or extremely distant. These are scientific projects and not ambitions that will take hundreds of years to be accomplished. There is no magical ingredient that we need to wait for.

Part 2: Comparing approaches and Dr. Koene's current and past work involvements

Q. What are the main approaches to creating a whole brain emulation?

Dr. Koene describes the difference between Simulation and Emulation as the terms are used on carboncopies.org and with regard to SIM. He compares SIM with the simulation work at Dr. Henry Markram's Blue Brain project and the engineering goals of Dr. Dharmendra Modha's SyNAPSE project at IBM. He point out the crucial difference in terms of the sources of data: statistical for simulations, subject-specific for emulations. He notes that simulations can teach us about intrinsic properties of a system.

Q. How does Halcyon Molecular fit into Dr. Koene's SIM endeavors?

Koene briefly desribes how he came to work at Halcyon Molecular, and that the long-term interests of the incredible team at Halcyon go far beyond DNA sequencing. It was the strength of that team that convinced Koene to join Halcyon. Goals include curing disease and extending life, and the DNA tools being developed are directly applicable to projects that solve tool requirements for whole brain emulation. Koene gives the example of Dr. Anthony Zador's (CSHL) approach of extracting a brain's connectome by using DNA barcodes as pointers at synaptic connections, as well as the example of the Molecular Ticker Tape concept described by Dr. Konrad Kording.

Q. How is it different than Koene's work in Europe, Boston and at carboncopies?

Koene explains that in Europe he was tasked with setting up a new department of neural engineering on a long time-horizon, a time-line that shrank drastically when the financial crisis hit Spain in 2009. Carboncopies, a non-profit founded by Dr. Koene and Dr. Gildert, pulls together an extensive network of researchers and projects, maintains a roadmap and identifies fundamental resources with the aim to bring about substrate-independent minds.

Part 3: Current progress toward SIM, benchmarks, time-lines and understanding

Q. How far are we now along the route to SIM, and what are the benchmarks and time-lines?

Dr. Koene explains that carboncopies has put together a roadmap of the major routes to SIM. Concrete requirements have been identified, as have the key researchers and those pieces of the puzzle that are in great need of explicit attention and advancement. Recent progress has seen the emergence of collaborations to create the tools for large-scale high-resolution data acquisition that is needed for whole brain emulation. Koene lists milestones such as the publication of fundamental literature taking place throughout 2012, the start of projects aimed at tools to carry out functional characterization within the brain, and early prototypes of such tools expected from the end of 2012 onward. As time-lines go, Koene points out the strong correlation between effort invested and time taken to accomplish the full range of tasks for SIM. This was also explained in Dr. Koene's talk at H+ in Hong Kong.

Recent efforts by Dr. Koene have also emphasized the very concrete nature of the projects involved: We know the data that we need to acquire and the sort of protocols by which to make the necessary measurements; we know the tools required in order to get to whole brain emulation; we know which practical approaches we can use today to create those tools; and we know what we can expect to end up with once we have completed those projects. That makes SIM, WBE and mind uploading concrete and plannable, objectives that are based on existing scientific understanding and engineering capabilities.

Q. What do we really know about the brain, and what don't we?

Koene clarifies that understanding the brain fully is not the same thing as creating a whole brain emulation, even though a WBE will certainly help enormously with the exploration of the brain in a way that was never possible before, and will include the ability to back out of procedures. The technologies and results provide a virtual brain laboratory. Dr. Koene explains that these benefits are some of the main motivating factors driving work in neuroinformatics and connectomics today.

Part 4: The best case scenario, the rationale behind SIM and verifying a whole brain emulation

Q. What is the dream outcome, the best case scenario of your work?

Koene: An in-vivo method to accomplish an upload via whole brain emulation to a substrate-independent mind, with full validation of its successful accomplishment.

Q. Why should we want to upload our minds?

For the reasoning behind SIM, Dr. Koene reiterates ideas published in his paper "Pattern survival vs. Gene survival", as well as contributions to the books "The Transhumanist Reader" and "The Singularity Hypothesis" to appear in 2012. Among other things, SIM offers a way to safeguard life by enabling back-ups, fault-tolerant implementations of "being". Near-term, Koene points out that solutions offered by SIM will be needed to allow us to combine wisdom with fresh ideas, an urgent problem for extended life spans. Long-term, we have to understand for the sake of the survival of our species that we are not intrinsically evolved to handle all possible challenges that will be coming our way, or to thrive in all environments, or be able to take advantage of all types of experiences. We give our machines the ability to process information in ways that we ourselves currently cannot conceptualize. Koene imagines, for example, what it would be like to think in quantum computation. Dr. Koene explains the importance to our species that we are mentally the most capable in our environment, and that we need to continue to develop and grow in that area. He points out the kinship of WBE and brain-computer interfacing (BCI) goals in that regard.

Q. How do we verify if a WBE is an exact copy of the original?

Koene explains that it is that question where in-vivo tools become particularly important. Functional characterization at high resolution is necessary in order to provide a sufficient number of reference points to practically enable tuning and error correction for a whole brain emulation. Piece-wise, step-wise procedures to accomplish WBE are much more reliable from an engineering perspective than methods that rely on single-step data acquisition and reconstruction. Those considerations are shared by any effort to create neural prostheses.

Part 5: Ethical issues with whole brain emulation, embodiment and sources of inspiration and ideas

Q. What happens then, how do you treat the original and the WBE, are there rights, issues of hierarchy?

Koene rephrases the question, noting that it comes down to basic questions such as "Is it really you?". Consider the issue of continuity, as examined in work by Dr. David Chalmers. Dr. Koene considered himself a fence-sitter on the issue for quite a while, but recently acquired a more nuanced view of self and its maintenance, as described in dissertation work by Dr. Max More. This underlined the significance of disruptive aspects of a mind uploading procedure. The details of the process may very well matter.

Regarding ethical considerations about substrate-independent minds, legal and practical considerations, considerations involving the existence of multiple copies, Dr. Koene notes that many of these questions that initially seem entirely novel are not actually so new. They have parallels that we have been dealing with for a long time. Ethics have had to change and adapt many times, and we can learn how to tackle many of the issues by looking at examples of what we have already dealt with. It is useful to realize when issues that seem new and unique upon first consideration are in fact merely new forms of familiar questions.

Q. After achieving whole brain emulation, would we soon after be able to install that in a new body?

Koene's sense is that the types of bodies that would be of interest are hard to predict at this moment, but that having those will clearly be a high priority for a society with SIM. SIM does not imply a lack of bodies. Embodiment is essential. The connection with the environment is essential. Feedback is essential.

Q. Do you read Science Fiction that can inform your work and which authors are your favorites?

Koene reminds that Arthur C. Clarke's story was an important catalyst and crystalized some of his ideas. Dr. Koene is also partial to the work of Greg Egan, which includes some stories (e.g. "Permutation City") that are opportunities to learn about possible issues related to SIM.

Part 6: Worst case scenario, possible unintended consequences, the singularity, long-term implications, finding additional information and the main message of the interview about SIM

Q. Is there a worst case scenario that you are afraid of?

In general, Koene is worried that events in the world might bring about problems with the infrastructure of our economy so that the kind of work needed to achieve SIM could become impossible. We may be time-limited in our opportunity to get to SIM. Also, the challenges that we will face, either in the environment or through the technology we develop (e.g. biotechnology, nanotechnology, AI) could make it too difficult for our species to survive or could lead us to barely scrape by. These possibilities give Koene a sense of urgency and imply a need to hurry up.

According to Koene, a worst case scenario in terms of outcomes of work on SIM could involve problems of ownership of information, manipulation or influencing of content, challenges about staying in charge of your personal information that is your mind. He anticipates a need for privacy controls. There is also the possibility that someone could use insights that were gained from work on SIM to produce undesirable effects. The possibility of an arms race against anticipated negative developments worries Dr. Koene. It is important to take into consideration the development of other technologies around SIM. No technology develops in isolation.

Q. What about unintended consequences, does Koene have any worries about possible negatives produced?

Koene: Being worried about something bad happening does not mean that the best course of action is not to do something. The status quo is not inherently safe, and someone will be working on these technologies. It would make sense for the persons who do worry, who do take possible problems into consideration to work on the technologies. Koene explains that SIM gives us a chance to keep up with all of the capabilities of other technologies. It can give us choice.

Q. Where does mind uploading fit within the Singularity?

Koene refers to his article to be published in the book "The Singularity Hypothesis". It is important that we do not take a bystander role, where we may never be able to understand any of what goes on at the level of our creations. We should participate. Dr. Koene finds the term "Singularity" a bit problematic. Resurce requirements and physical laws need to be considered for rational prediction and planning. According to Koene, being participants turns the process into a "soft take-off" scenario. As Marvin Minsky said: "If you are riding the wave then there is no Singularity."

Q. What are the chances of us surviving if we do not take a participant role?

In the long term, Dr. Koene thinks that the chances of survival for a stagnant species are very low. On such time scales even our solar system will not persist as it is. In the short term there are very many possible scenarios.

Q. Where can viewers go to find out more?

Q. Do you have a single message for viewers to take away from this interview?

Dr. Koene: Substrate-independent minds, especially via whole brain emulation, is a concrete objective, one with a set of scientific and development projects that are based on current understanding and engineering.