5
Potential Intelligence and Military Applications of Cognitive Neuroscience and Related Technologies

INTRODUCTION

Chapter 2 discusses the basis of neuroscience and neurotechnologies, and Chapter 3 describes apparent trends in neuroscience and neurotechnologies. This chapter applies the technology-warning method developed in Avoiding Surprise in an Era of Global Technology Advances (NRC, 2005) to particular fields of science and technology that potentially have important military applications. In some cases, the committee amplifies the technology-warning method to illustrate how neurophysiological research and cognitive and neural research conducted in other countries might affect assessments.

Appendix C (Chapter 2 from Avoiding Surprise) sets forth in great detail the background and purpose of the technology-warning method and defines the key terms listed in technology-assessment Charts 5-1 through 5-8 below in this chapter. Readers will find it helpful to read Appendix C before reading this chapter. In keeping with the main subjects of interest identified in Chapter 2, the examples below are meant to be representational depictions of future military applications. Figure 5-1 illustrates cross-functional information flow and outlines one possible solution to analyzing cognitive neuroscience research by intelligence analysts.

At the base of Figure 5-1 are examples of sources and types of information that must be collected. Sources of neuroscience information should include the public sector, the private sector, and the intelligence community. That information needs to be analyzed and applied to the technology-warning methodology. Afterward, the information will need to be analyzed, packaged, and delivered to military decision makers. Next, a feedback mechanism should exist whereby the decision maker can formulate questions to address to the producers of the information. The decision makers could ask for more information (deeper dive)



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 124
5 Potential Intelligence and Military Applications of Cognitive Neuroscience and Related Technologies INTRODUCTION Chapter 2 discusses the basis of neuroscience and neurotechnologies, and Chapter 3 describes apparent trends in neuroscience and neurotechnologies. This chapter applies the technology-warning method developed in Aoiding Surprise in an Era of Global Technology Adances (NRC, 2005) to particular fields of science and technology that potentially have important military applications. In some cases, the committee amplifies the technology-warning method to illustrate how neurophysiological research and cognitive and neural research conducted in other countries might affect assessments. Appendix C (Chapter 2 from Aoiding Surprise) sets forth in great detail the background and purpose of the technology-warning method and defines the key terms listed in technology-assessment Charts 5-1 through 5-8 below in this chap- ter. Readers will find it helpful to read Appendix C before reading this chapter. In keeping with the main subjects of interest identified in Chapter 2, the examples below are meant to be representational depictions of future military applications. Figure 5-1 illustrates cross-functional information flow and outlines one possible solution to analyzing cognitive neuroscience research by intelligence analysts. At the base of Figure 5-1 are examples of sources and types of information that must be collected. Sources of neuroscience information should include the public sector, the private sector, and the intelligence community. That informa- tion needs to be analyzed and applied to the technology-warning methodology. Afterward, the information will need to be analyzed, packaged, and delivered to military decision makers. Next, a feedback mechanism should exist whereby the decision maker can formulate questions to address to the producers of the information. The decision makers could ask for more information (deeper dive) 

OCR for page 124
 pOTENTIAL INTELLIGENCE AND MILITArY AppLICATIONS Ideal Global Cognitive Neuroscience Information Flow for the Intelligence Community Review Decision New Requirements Decision Maker Report Intelligence Review Community Review Peer Critique Analysis* Fusion Gathering Refinement Academia Industry Sources Intelligence Medicine Community *Technology warning methodology detailed in the 2005 National Research Council report Avoiding Surprise in an Era of Global Technology Advances. FIGURE 5-1 TIGER technology-warning method as one part of an overarching analytical 5-1.eps information flow. or clarification of specific issues presented in reporting. Finally, decision makers must identify topical areas requiring collection and analysis of information by the intelligence community. As with weather forecasting, technology forecasting is measured in terms of probabilities, not certainties. Figure 5-2 is a notional diagram that illustrates some of the possibilities of technology-warning maturation. The abscissa of the diagram is time. It may be obvious, but it must be stated, that as analysts try to forecast farther into the future prediction becomes less accurate. By analogy, the afternoon weather forecast is more accurate than a 7-day forecast, which is more accurate than a 1-year forecast. The same difficulties will be apparent with neuro- science forecasting as analysts try to predict developments over the next 20 years. Figure 5-2 attempts to illustrate some considerations required in watching the development of cognitive neuroscience and related technologies. In Figure 5-2, technology-warning levels are plotted on the ordinate of the diagram. The advancement of knowledge and technologies may have sev- eral trajectories, including linear and exponential, incremental, and nonlinear or crossover. Intelligence officers are often concerned with developing “trip wires” to avoid technology surprise. Trip wires are described as “observables” and “indicators” in Aoiding Surprise (NRC, 2005). To monitor these indicators of

OCR for page 124
 EMErGING COGNITIVE NEUrOSCIENCE AND rELATED TECHNOLOGIES Technology A Technology B Technology C Arrow is crossover technology Technology Warning Time FIGURE 5-2 Technology warning over5-2.eps time. The graph depicts ways in which technology may be developed. Understanding the trajectory (incremental, exponential, and so on) in which specific types of technologies may develop could assist analysts in developing trig- gers and watching for crossovers. technology warning (forecast changes or maximal potential) more correctly, some thought must be brought to bear on the journey that information may undergo on the way to increasing technology-warning levels; that is, the path must be studied, as well as the end point. Some things may be on a linear trajectory (Technology A in Figure 5-2) or an exponential trajectory (Technology B in Figure 5-2) simi- lar to Moore’s law of transistors doubling or increasing in magnetic resonance imaging (MRI) magnet size (Lloyd, 2000). Other types of science and technol- ogy warning concerns will develop in quantum or step functions (Technology C in Figure 5-2). Still other technologies will be crossover technologies (indicated by the yellow arrow in Figure 5-2). These were called innovation or integration events in Aoiding Surprise (NRC, 2005). Crossover technologies, such as the use of a cellular telephone in an improvised explosive device (IED), are difficult to predict and often represent extreme force multipliers.1 1A force multiplier is a capability that, when added to and used by a combat force, substan- tially increases the combat potential of that force and thus enhances the probability of mission accomplishment.

OCR for page 124
 pOTENTIAL INTELLIGENCE AND MILITArY AppLICATIONS Most advances in cognitive neuroscience and related technologies will have dual uses. Pressures leading to advances will come from the pull of medical neces- sity and the push of science and technology advancement. Large state-funded programs are possible, but so is the use of new (or existing) technology in an innovation or integration. As cognitive neuroscience and related technologies become more pervasive, using technology for nefarious purposes becomes easier. Moreover, the triggers and observables become less obvious to the analyst and the collector. For example, a program using specialized equipment, such as that outlined on the Australia Group List of Biological Agents for Export Control, may or may not be applicable.2 However, the types of experiments being done may be more telling than the type of equipment needed. The same equipment might be needed for medical and for disruptive neuropsychopharmacological experiments. It could be asked, What types of experiments are being done? How are the experi- ments being controlled and monitored, and why were they chosen? How would human experimentation be conducted outside accepted informed-consent limits? MARKET DRIVERS OF COGNITIVE NEUROSCIENCE AND RELATED TECHNOLOGIES AS INDICATORS OF THE DEMAND FOR COTS TECHNOLOGIES Overview The military and intelligence community can forecast some aspects of how applied cognitive neuroscience technology might threaten the national security of the United States by tracking market drivers. Market drivers in this context are specific groups of neurotechnology consumers that create demand for devel- opment and delivery of applied cognitive neuroscience technology goods and services.3 The applied cognitive neuroscience technology consumer landscape, for the purposes of this report, is categorized into three segments: • Health. Customers are seeking help in addressing mental illness, brain disease, and injury; • Enhancement. Customers do not possess a diagnosable neurological dis- order but are seeking some cognitive performance advantage or want to prevent a probable decline; and 2 For additional information, see Australia Group Web site, http://www.australiagroup.net/en/ biological_agents.html. Last accessed on March 25, 2008. 3The report entitled The Neurotechnology Industry 00 report: Drugs, Deices and Diagnostics for the Brain and Nerous System: Market Analysis and Strategic Inestment Guide of the Global Neurological Disease and psychiatric Illness Markets from NeuroInsights provides a comprehensive breakdown of the current market drivers. For additional information, see Neuroinsights’ Web site at http://www.neuroinsights.com/marketreports.html. Last accessed on January 24, 2008. The approach taken in this report segments the marketplace differently to comply with the statement of task.

OCR for page 124
 EMErGING COGNITIVE NEUrOSCIENCE AND rELATED TECHNOLOGIES • Degradation. Customers seek advantage by degrading, temporarily or permanently, the cognitive abilities of others. There are dramatic differences in the motivations and sizes of these customer segments. The health market represents all the customers that would benefit from diagnosis, treatment, and cures for a mental illness or brain disorder. This market has been documented to be dramatically underserved in the United States and the world. Only about one-third of the potential market is actually served, leading to the two related conclusions: (1) the market has not reached its full potential and (2) the direct and indirect cost burden to the customers (and their governments) might change dramatically.4 Estimates vary on the potential revenue of the neurotechnology health market, but a reasonable estimate can be bounded by current revenue reports (as the low estimate) and total impact or cost burden (as the high estimate). Cost burden estimates include the cost of products and services as well as the indirect cost of no treatment. Total neurotechnology company revenue is about $120 billion; whereas total economic burden in the United States is on the order of $1 trillion per year and the world burden is about twice that.5 NIMH has stated that over 50 percent of Americans will suffer from a mental disorder in their lifetime. But the ability of this consumer group to pay for or get access to diagnosis, treatment, and cures constrains the health market segment. In other words, the demand exceeds the market’s ability to provide goods and services in this segment. The enhancement segment is largely underground, but growing. For example, healthy consumers cannot legally get a prescription to improve their attention or memory. However, there is significant anecdotal information that healthy people do take Alzheimer medication (Aricept), medication to treat narcolepsy (modafinil), and medication to treat attention deficit disorder (Ritalin) to improve performance. Sleep and wake medications, e.g., zolpidem and armodafinil, respec- tively, as products for enhancement represent a pharmaceutical market of over $5 billion. This market is generally more acceptable because one can be otherwise healthy and get a prescription. In many cases, the enhancement market is served by the off-label use of products and services created for the health market. There are not good estimates for the potential revenue of the enhancement segment. Dramatic policy, ethical, and cultural changes will have to occur before this market becomes transparent and can grow quickly in the United States. But it is clear that the aging U.S. population will be a significant driver for this market. 4 For detailed statistics, see the National Institute of Mental Health Web site at http://www.nimh. nih.gov/health/topics/statistics/index.shtml. Last accessed on January 18, 2008. 5 For additional information, see the World Health Organization Web site at http://www.who.int/ healthinfo/bod/en/index.html for additional information. Last accessed on January 18, 2008.

OCR for page 124
 pOTENTIAL INTELLIGENCE AND MILITArY AppLICATIONS The neurotechnology degradation market segment is completely underground with only speculative information available.6 This cognitive weapons market does exist (e.g. the fentanyl derivative used in the 2002 Moscow theatre hostage crisis (Wax et al., 2003), and there is demand in the market from government agencies. There is potential for products that are developed for the health and enhancement markets to be misused in the degradation market—for example, using TMS to cause seizures (Pascual-Leone et al., 2002; NRC, 2003). Finding 5-1. International market forces and global public demand have created an impetus for neuropsychopharmacology and neurotechnology research that will lead to new technologies and drugs, particularly in areas of cognition and performance, that will include off-label uses. Off-label drug use can alert intel- ligence analysts to compounds, methods of administration, or risk factors that may be unknown in civilian or military medicine and can help identify profiles of unanticipated effects. The Search for New and Patentable Neurophysiological Agents There are important U.S. and international market incentives for research activities in neuropsychopharmacology. No currently marketed drugs, psycho- therapeutic or otherwise, are curative for mental illnesses; all have limited thera- peutic efficacy, and all have undesired side effects. But the fact that many currently marketed drugs are or have been major sources of profit for the ethical pharma- ceutical industry creates a market emphasis on finding new patentable entities. Additionally, analysts in a number of disciplines have predicted large increases in the prevalence of some disorders (e.g., Alzheimer’s disease) that currently have few treatment options. This creates another strong market impetus. One perhaps underappreciated feature of neuropsychopharmacology men- tioned above further strengthens the market impetus. Few, if any, current psycho- therapeutic drugs were designed or developed on the basis of an understanding of the psychopathology or the pathoetiology of the target disorder. Achievement of molecular- or organism-level understanding of pathophysiology of brain dis- eases will create enormous opportunities for drug discovery and development. Furthermore, the likelihood that some of these new drugs will exploit heretofore unknown mechanisms or brain functions creates opportunities for patentable pharmaceuticals. Any real advance in understanding of human behavior or brain disease could lead rapidly to truly new types of drugs, which could present poten- tial positive or negative threats. 6 For listings of neurotoxic agents and effects, see the Federation of American Scientists’ Web site at http://www.fas.org/biosecurity/resource/lists.htm. Last accessed on January 24, 2008.

OCR for page 124
0 EMErGING COGNITIVE NEUrOSCIENCE AND rELATED TECHNOLOGIES Market Barriers and How the Drivers May Change Consumer demand drives the market, but markets always have constraints, both internal (consumer driven) and external (ethics, culture and government). The expansion and reduction of a market can be altered by changes such as new products, changes in regulations, or consumer awareness to name a few. The neurotechnology markets are highly influenced by the ethical, regulatory, politi- cal, and cultural environment. This section looks at potential changes in the three categories listed above (health, enhancement and degradation) and then introduces two additional markets that could appear in strength in the future, the illegal and forced treatment neuro- technology markets. Health Market As stated earlier, there is a great deal of pent-up demand in the health market because of access and cost barriers. And overall demand for pharmaceuticals to treat mental illness and many physical diseases is moderated by the associated ineffectiveness or the side effects associated with many of the treatments. In the NIH CATIE schizophrenia study, approximately 70 percent of the patients did not do well on their first drug and had to be switched to a new drug. Customers seek- ing better health are often unsatisfied by the existing products—there is demand for improved and more cost-effective prevention and treatments. One barrier to the expansion of the health market is the stigma associated with mental illness.7 Social attitudes, ethics, and government policy remain major limiting factors. It has been a difficult battle to approve the Mental Health Parity Act of 2007 so that mental illness is viewed as a disease of the brain instead of a lifestyle choice or poor parenting. As awareness improves and health insurance broadens its payments for mental illness/brain disease, there could be as much as a doubling in treatment and associated revenue. A significant barrier to market growth is the substantial investment of money and time needed to develop a drug for brain disease that is effective. On aver- age, it takes almost 10 years and $500 million to take a neuropharmaceutical to market; however, new drug development could be aided by new technologies and processes, the market could advance more rapidly. Changes in related fields, such as genetics and neuroinformatics may be able to impact this process. There is promise that once neuroimaging is used effectively in the develop- ment and evaluation of drug treatments, there could be radical changes in the health market. First, diagnostic imaging tools may be able to provide an objective, quantifiable diagnosis of a disease or illness and then second, they may allow for some level of prevention if diagnosis can occur earlier and then third, the treat- 7 For additional information, see the Surgeon General’s Web site at http://www.surgeongeneral. gov/library/mentalhealth/chapter1/sec1.html#roots_stigma. Last accessed on January 18, 2008.

OCR for page 124
 pOTENTIAL INTELLIGENCE AND MILITArY AppLICATIONS ment could be more specific to the individual. Early and accurate diagnosis could lead to treatments with fewer side effects. A factor that can potentially grow the neurotechnology health market is the advent of neuroinformatics. Diagnosis, prevention, treatment and then cure of brain injury and disease will benefit from better use of the complex array of information associated with a given health problem. Patient information, behav- ior, genetics, and image data can be combined to better understand the individual aspects of the affliction. As mentioned before, neuroinformatics may well enable a shortened development cycle for new drugs and treatment regimens. Finding 5-2. Neurotechnology products may be dual use. Products intended for the health market can be used in the enhancement and degradation markets. Additionally, the product life cycle can be shunted because of the nature of the enhancement and degradation markets. Enhancement Market The neurotechnology enhancement market is analogous to the athletic per- formance enhancement market. People will make the choice to take illegal and off-label prescription neuropharmaceuticals even if they do not know the side effects or believe that the side effects are worth the potential enhancement. This controversial market will grow dramatically if evidence becomes available that a specific drug is consistently effective in improving performance. However, if this market expansion does occur, it is likely that current attitudes will prevail and result in use being squelched through policy and law. History shows that prohibition will not work to stop consumption of a banned product and that the government that takes the path of prohibition will invest in methods to diagnose the abuse. In addition to off-label use of FDA approved drugs and devices, there are enhancement products available that are not subject to FDA approval. These include software products and vitamin supplements. In almost all cases, the effectiveness of these products has not been established with a clinical trial- like process, but their popularity is increasing (Vollset and Ueland, 2005). The software products are focused on enhancing neural plasticity and delaying the effects of aging. In some cases, they include direct neural feedback using EEG. This nascent market could grow rapidly because of the ease of developing new products and the lack of clear regulations. An interesting off-label use of an inexpensive iontophoresis dose controller is as a transcranial direct current stimu- lation system, or tDCS.8 tDCS enhances certain cognitive functions (Antal et al., 2006). 8 For additional information, see the IOMED Web site at http://www.iomed.com. Last accessed on January 18, 2008.

OCR for page 124
 EMErGING COGNITIVE NEUrOSCIENCE AND rELATED TECHNOLOGIES TABLE 5-1 Current Drugs Used in New Ways or For New Targets and New Types of Drug Entities Current Agents That Might Achieve Aim or Target of Agent for Change Part of Aim Memory, learning, cognitive speed Various drugs, stem cell reparative therapy, nanoneural networks Alertness and impulse control Stimulants, modafinil, ampakines, antidepressants Mood, anxiety, and self perception Beta blockers, selective serotonin re-uptake inhibitors Creativity Transcranial magnetic stimulation Trust, empathy, and moral decision making Oxytocin, testosterone suppression Sleep and wake medications Zolpidem, armodafinil SOURCE: Hughes (2007). ©2007 by Elsevier. Reprinted with permission. The 65 and over population is expected to double in the next 25 years. This shift to an older population, with 20 percent of the population over the age of 65 by 2020, will increase demand for enhancement neurotechnology. Much of the market growth could be limited to non-FDA controlled products because of the associated policy issues. Public and medical demand for enhancement or a competitive edge may create market impetus in many areas. This public climate is illustrated by sources as respectable as a 2006 report in the Harvard Crimson about student demand for stimulants,9 a 2008 NYTimes.com report about profes- sional baseball players claiming medical exemptions for stimulant use, 10 and a 2003 Fortune magazine report suggesting memory pills for a competitive edge.11 Development areas that may respond to such market incentives include cognition and mentation in all their forms (arousal, attention, cognition, learning, memory, motivation, among others), movement and performance (strength, speed, stamina, motor learning, among others), and mood (anxiety, depression, craving, emo- tional memories, among others). As noted earlier, development of new agents for such uses could result in drug entities that change cognition, performance, or mood in unexpected, cur- rently unavailable ways. The examples in Table 5-1 drawn from Hughes (2007) suggest both current drugs used in new ways or for new targets and new types of drug entities. 9 For additional information, see http://www.thecrimson.com/article.aspx?ref=513261. Last accessed on January 23, 2008. 10 For additional information, see http://www.nytimes.com/2008/01/19/sports/baseball/19stimulants. html?_r=1&ref=sports&oref=slogin. Last accessed on January 23, 2008. 11 For additional information, see http://money.cnn.com/magazines/fortune/fortune_archive/2003/ 05/12/342287/index.htm. Last accessed on January 23, 2008.

OCR for page 124
 pOTENTIAL INTELLIGENCE AND MILITArY AppLICATIONS Actual progress, or at least actual changes, in such development programs may accelerate as drug developers attempt to define drugs with more targeted psychological or cognitive effects. As discussed in another context, much research on “cognition enhancers” has sought broadly acting drugs. Programs that target narrower, more specific, effects on cognition may be more interesting and warrant attention. For example, drugs that appear (at least initially) to have effects con- centrated in one area, such as speed of learning, accuracy of retention, selective retention or selective forgetting (as of traumatic memories), expanded attention span, and social cognition, might deserve particular attention. Such agents might warrant attention for both their risks and their ability to enhance performance and safety of the individual warfighter. Drug entities developed to have increased selectivity of effects on cognition or performance might conceivably be integrated into enemy military programs. Their adoption by a military force, for example, might be facilitated both by their targeted effects and by possible reductions in side-effect profiles. Agents now used to enhance cognition or performance can pose safety risks to warfighters that include abuse, misuse (e.g., excessive doses or too frequent use), unintended consequences, unexpected interactions with situations, and alterations in social behavior. Indeed, drug entities with reduced side-effect profiles could achieve greater use, independent of increased selectivity of action. Adoption of such agents by medical practitioners, patients, and/or portions of the general public will be influenced by important social and cultural forces, as discussed elsewhere in this report. One force that might warrant particular analy- sis is inflated or frankly bogus marketing claims for specificity, effectiveness, or safety made by sources in private, government and/or academic sectors. Degradation Market The development of a more humane way of fighting a war by using cognitive weapons could dramatically change the degradation neurotechnology market. Pills instead of bullets. The fear that this approach to fighting war might be developed will be justification for developing countermeasures to possible cog- nitive weapons. This escalation might lead to innovations that could cause this market area to expand rapidly. Tests would need to be developed to determine if a soldier had been harmed by a cognitive weapon. And there would be a need for a prophylactic of some sort. If a particularly effective degradation product is developed that has few side effects, escalation of this market will be self- fulfilling. The concept of torture could also be altered by products in this market. It is possible that someday there could be a technique developed to extract infor- mation from a prisoner that does not have any lasting side effects. Recently, it has been documented in a small study that tDCS delays a person’s ability to tell a lie (Priori et al., 2007).

OCR for page 124
 EMErGING COGNITIVE NEUrOSCIENCE AND rELATED TECHNOLOGIES Finding 5-3. Neurotechnology products developed specifically for the enhance- ment and degradation markets may not be subject to the same development and distribution/access constraints that apply to the health market. Safety, social, ethical, and legal constraints may not apply in countries where regulation is less rigorous. forced Treatment Market There are instances where it might be possible in the future to control a person’s behavior through the application of drugs or electromagnetic devices. It is estimated that about 25 percent of prisoners are mentally ill. There are currently significant efforts to identify the brain/biological basis of certain types of crimi- nal behavior. What if certain types of criminal behavior had a biological/brain basis and could be prevented with treatment? Some cultures or governments may determine that forced treatment gives them a competitive advantage by reducing the cost burden of disease and illness. This has occurred in the United States for things like tuberculosis. The limiting factor for this market to grow is resolution of the related ethical issues. For example, the Mind Research Network has col- lected the neuroimaging and genetic data on over 300 prisoners using a mobile MRI facility and has a goal of collecting a thousand subjects by 2009. As stated in Chapter 4, current guidelines for ethical recruitment and participation of human volunteers in research generally prohibit participation by prisoners, due to actual or potential coercion risk (IOM, 2006). TECHNOLOGY ASSESSMENTS: NEUROPSYCHOPHARMACOLOGY Current scientific understanding of neuropsychopharmacology suggests types of drugs whose appearance in national or international research or commercial arenas might pose predictable threats to U.S. warfighters. One type of identifi- able threat might be development of antagonists for drug entities that currently have no antagonists, such as an antagonist for ethanol. Such an antagonist could allow adversaries to protect their own warfighters against an agent that is widely dispersed (for example, in gas, in aerosols, in drinking water, or by high-altitude delivery). Another type would be the coupling of known agents with unexpected routes of administration, doses, or personnel protective agents. A known example is the aerosol use of fentanyl during the Russian hostage-taking (an instance of all three surprises). Other threats could arise from training of warfighters to operate under new pharmacological conditions. An adversary could train warfighters to operate under the influence of chemical agents that ordinarily disrupt performance or could modify warfighters to resist such agents. Such resistance could be conferred

OCR for page 124
 EMErGING COGNITIVE NEUrOSCIENCE AND rELATED TECHNOLOGIES CHART 5-1 Use of Neuropsychological Agents as Incapacitants Technology Triggers and Observables Aerosols of opioids serve as excellent Future military deployments and/or incapacitants; reversal agents and training by other countries should premedications can be administered to be observed. In addition, medicinal protect soldiers. Russia deployed this chemical developments of other technology in the Moscow Dubrovka countries should be observed for Theater in 2002. The agents were military applications of medical probably fentanyl derivatives and may pharmacology, with particular attention have included inhalation anesthetics to more potent fentanyl derivatives and (Wax et al., 2003). inhalation anesthetics. Attention should also be given to technologies that would allow aerosol release in building ventilation systems. Accessibility Maturity Consequence Level 3 Warning This technology is asymmetrical in that the U.S. military is not believed to have it, whereas Russia is known to have it and to have deployed it. The consequence of releasing material would be nonlethal incapacitation of U.S. operators or assets. A small number of enemy operators could rapidly incapacitate a larger number of U.S. forces without engaging in combat. Once incapacitated, the blue forces could be killed or captured by the red forces that had been pretreated with antidote.

OCR for page 124
 pOTENTIAL INTELLIGENCE AND MILITArY AppLICATIONS CHART 5-2 Nanotechnologies or Gas-Phase Technologies That Allow Dispersal of Highly Potent Chemicals over Wide Areas Technology Triggers and Observables Pharmacological agents are not used Triggers to watch for would be as weapons of mass effect, because development of such technologies their large­scale deployment is for agricultural or pesticide purposes. impractical; it is currently impossible to Precedents exist for using toxins get an effective dose to a combatant. as weapons delivered by ingestion, However, technologies that could be transdermally, or by inhalation. Analysts available in the next 20 years would should watch for development of allow dispersal of agents in delivery standardized delivery systems that can vehicles that would be analogous to a distribute small­molecule “payloads” pharmacological cluster bomb or a land over large areas, such as crop fields. mine. Those delivery systems would protect agents from meteorological conditions and then release agents on contact with a soldier. Of particular concern would be a single delivery system that could be easily loaded with different agents, as warheads are switched in an artillery round. This type of system would allow easy crossover to nefarious purposes. Accessibility Maturity Consequence Level 2 Futures The consequence of such a system would be an off­the­shelf small­molecule delivery system that could be leveraged against U.S. forces. Highly potent compounds could cause troop incapacitation or performance degradation. Coupling off­the­shelf agents with an off­the­ shelf delivery system would create new threat opportunities against U.S. forces. Spreading agents that would be released on contact would deny U.S. military operations some strategic infrastructure or geographic positions.

OCR for page 124
 EMErGING COGNITIVE NEUrOSCIENCE AND rELATED TECHNOLOGIES CHART 5-3 Technologies for Highly Potent Blood-Pressure Agents or Sensory-Specific Pharmacological Targeting Technology Triggers and Observables Existing pharmacological agents The literature should be monitored for could be used in a nefarious way. development of sympathomimetic and An example would be currently used parasympatholytic drugs. Development agents, such as alpha blockers, that may be revealed in basic chemical would work quickly to drop blood discovery and synthesis pathways. pressure if delivered in high doses. In Antihypertensives continue to be of addition, anticholinergic agents could great pharmacological research and cause molecular changes that lead to development interest. Clinical trials temporary blindness. should be monitored closely, and reports of unknown or unrecognized incapacitants should be carefully analyzed. Accessibility Maturity Consequence Level 2 Futures The consequence of development of these technologies would be an asymmetrical threat to U.S. troops. The consequence of releasing material would be nonlethal incapacitation of U.S. operators or assets. A small number of enemy operators could rapidly incapacitate a larger number of U.S. forces without engaging in combat. Once incapacitated, the U.S. forces could be killed or captured by enemy forces that had been pretreated with an antidote.

OCR for page 124
 pOTENTIAL INTELLIGENCE AND MILITArY AppLICATIONS CHART 5-4 Drug-Delivery Systems Applied to the Blood-Brain Barrier Technology Triggers and Observables New nanotechnologies have allowed Triggers and observables include molecular conjugation or encapsulation open­source publications and that may permit unprecedented access collaboration between pharmaceutical to the brain. companies and industrial fabricators of nanotechnologies. Research and development will continue because of the need for new neuropharmaceuticals. This is a subject of broad international interest and should be monitored to maintain U.S. awareness of these technologies as related to the blood­brain barrier. Human trials with regard to drug delivery to the central nervous system should also be monitored closely because it will be required for technology maturation. Accessibility Maturity Consequence Level 2 Warning Increased access to the brain by drugs that cross the blood­brain barrier would allow increased potency of therapeutics. Brain injury remains a scourge of the warfighter; better treatment by military medical personnel would be of extreme benefit to our troops. Not only would successful treatment for injuries increase morale, but it would allow for return of highly trained personnel to active duty. On the disruptive side, if an adversary were able to breach the blood­brain barrier, warfighters could be rapidly incapacitated. That could be done by covertly assaulting troops, and the result could range from severe incapacitation to death.

OCR for page 124
0 EMErGING COGNITIVE NEUrOSCIENCE AND rELATED TECHNOLOGIES TECHNOLOGY ASSESSMENTS: DISTRIBUTED HUMAN-MACHINE SYSTEMS AND COMPUTATIONAL BIOLOGY One example of a potential threat in the category of distributed human- machine systems and computational biology is a computational cognitive frame- work. Computers are used widely and successfully to store, recall, and process information, but they do those things very differently from humans. Although modeling the whole brain in the next 2 decades is highly unlikely, it is not unrea- sonable to imagine that substantial subsystems could be modeled. Moreover, increasingly sophisticated cognitive systems will probably be constructed in that period that, although not aiming to mimic processes in the brain, could perform similar tasks adequately enough to be useful, especially in constrained situa- tions. Making up in part for the lack of years of “experience” that constitutes an important basis of much of human intelligence, intelligent systems are increas- ingly using the Internet to assist in their training. In the next 2 decades, one can easily imagine artificial cognitive systems that will be increasingly capable of complex reasoning that, in combination with human capabilities, can exploit an incredibly large amount of knowledge and substantially advance the ability of human analysts to monitor large amounts of information and to formulate and test credible hypotheses. A second example of a potential threat in the category of distributed human-machine systems and computational biology stems from physiomimetic computing hardware. With the exception of a few small-scale endeavors, the computational-cognition community has essentially been confined to using large- scale general-purpose computers for almost all its modeling and simulation efforts. However, the digital computers have evolved through a history of solving computational problems that lend themselves well to the notion of precise logic or well-structured probabilistic or statistical models. The brain is inherently differ- ent, with its neurons and other components acting more as analog devices with a continuous set of values and a complex set of connections. It is possible that one could make important advances in artificial cognitive modeling by constructing a computer based on analog circuits that would address many of today’s modeling and simulation challenges. Such a computer could perhaps be much smaller than today’s digital computers and be easy for the United States or its adversaries to construct (Watson, 1997; Konar; 2000; Von Neumann, 2000; Giles, 2001; Arbib, 2003; Siegelmann, 2003; Schemmel et al., 2004; Trautteur and Tamburrini, 2007; Mills, 2008). A third example stems from microscopic magnetic detectors. Precision detection of magnetic fields in atomic devices is quickly becoming a mainstream technology. Previously, superconducting quantum interference devices (SQUIDs) were the primary means of precision magnetic-field detection, but the necessity of having ultralow temperature has prevented SQUIDs from becoming an inexpen- sive, simple-to-use technology. New microscience and nanoscience technologies that will enable small precision magnetometers (measured in millimeters) are

OCR for page 124
 pOTENTIAL INTELLIGENCE AND MILITArY AppLICATIONS emerging. They could enable dramatic increases in the number of sensors (and, similarly, spatial resolution) and the portability of measurement devices. Such changes could revolutionize neuroimaging. A fourth example is a brain-machine interface (BMI) for the control of weapons systems. With breakthroughs in training and the detection of brain activity for BMIs to the required level of complexity and specificity, virtually any conceivable weapons system could be controlled by this means. Basic research would need to be performed to determine the ultimate range of physiological potential and the limitations of such interfaces. Even if this range of performance were found to be feasible, it would remain to be demonstrated that BMIs are superior to conventional methods for controlling computing functions and robotic vehicles. With respect to the use of brain scanning in combat risk-assessment devices, it has been suggested that a weapons system could indicate potentially hostile forces as candidate targets while passing over noncombatants. Functional neuroimaging technology can already reliably discriminate between men and women, and it seems possible that this kind of capability could be extended to take other simple discriminators into account. However, it would require excep- tional advances in technology to have the capability to discriminate reliably between combatants with potentially hostile intent and noncombatants without hostile intent, and such a capability may ultimately be infeasible given the wide range of individual differences and the high consequences of misidentification. A final example would be the development of sophisticated distributed human-machine systems that are capable of greatly enhancing the cognitive and physical performance of human warfighters or allowing them to coordinate the actions of autonomous systems effectively. See Charts 5-5 to 5-8 for examples of the committee’s application of the technology warning methodology to the areas of computational biology and distributed human-machine systems. FINDINGS AND RECOMMENDATION Finding 5-4. Rapid advances in cognitive neuroscience, as in science and tech- nology in general, represent a major challenge to the IC. The IC does not have the internal capability to warn against scientific developments that could lead to major—even catastrophic—intelligence failures in the years ahead. An effective warning model must depend on continuous input from strong internal science and technology programs, strong interactive networks with outside scientific experts, and government decision makers who engage in the process and take it seriously as a driver of resources. All that remains a work in progress for the IC. Technology warning in the IC today is hampered by several factors including the low priority it has among senior leaders; the paucity of resources invested by the community in internal science and technology capability; the continuing inade- quate management attention to the needs of IC analysts; and the need to establish

OCR for page 124
 EMErGING COGNITIVE NEUrOSCIENCE AND rELATED TECHNOLOGIES CHART 5-5 Computational Cognitive Framework Technology Triggers and Observables A computational cognitive framework • Research papers. that, in combination with human • Commercial products. capabilities, can monitor and perform • Increased or unusual Web crawling. complex reasoning on large amounts of knowledge from the Internet. Accessibility Maturity Consequence Level 1 Watch This technology would lead to important advances in the ability of human analysts to monitor large amounts of information and to formulate and test credible hypotheses. CHART 5-6 Physiomimetic Computing Hardware Technology Triggers and Observables Analogue “brain­like” computer that • Substantial investment in analogue­ would enable a different approach to computing hardware. computational cognition. • Revolutionary new algorithms that take advantage of the new hardware. Accessibility Maturity Consequence Level 3 Futures Access to an artificial cognitive system at a very low price. This is a low­probability event, but such a system could provide both inexpensive intelligence and potentially an offensive­ cyberwarfare capability.

OCR for page 124
 pOTENTIAL INTELLIGENCE AND MILITArY AppLICATIONS CHART 5-7 Microscopic Magnetic Detectors Technology Triggers and Observables Millimeter­sized magnetic detectors New commercial products or research that require volume under 0.5 cm3 and devices that would enable functional ultralow power. imaging in traditional application environments. Accessibility Maturity Consequence Level 2 Alert Imaging device with extreme portability and many more channels of data. It would give the developer a revolutionary advantage in understanding of neural behavior in realistic situations. CHART 5-8 Sophisticated Distributed Human-Machine Systems Technology Triggers and Observables Software or robotic assistants for • Research papers. advanced sensor grids, control and • Commercial products. coordination of unmanned autonomous systems, advanced command posts and intelligence analyst workbenches, coordination of joint or coalition operations, logistics, and information assurance. Accessibility Maturity Consequence Level 2 Alert This technology would greatly enhance the cognitive or physical performance of warfighters and decision­ makers or allow them to coordinate the actions of autonomous systems with much­improved effectiveness.

OCR for page 124
 EMErGING COGNITIVE NEUrOSCIENCE AND rELATED TECHNOLOGIES close ongoing collaborations with analysts in other agencies, in the scientific community at large, the corporate world, and academia, where the IC can find the most advanced understanding of scientific trends and their implications. Finding 5-5. The recommendations in this report to improve technology warning for cognitive neuroscience and related technologies are unlikely to succeed unless the following issues are addressed: • Emphasizing science and technology as a priority for intelligence collec- tion and analysis. • Appointing and retaining accomplished IC professionals with advanced scientific and technical training to aid in the development of S&T collection strategies. • Increasing external collaboration by the IC with the academic community. It should be noted that some components of the IC have made great strides in reaching out to the academic community. Recommendation 5-1. The intelligence community should use a more central- ized indication and warning process that involves analysis, requirement genera- tion, and reporting. Engagement with the academic community is required and is good, but it is not now systematically targeted against foreign research. REFERENCES Arbib, Michael. 2003. The Handbook of Brain Theory and Neural Networks: Second Edition. MIT Press, Cambridge, MA. Antal, A., M.A. Nitsche, and W. Paulus. 2006. Transcranial direct current stimulation and the visual cortex. Brain resolution Bulletin 68(6):459-463. Bliss, T.V.P. 1999. Neurobiology: Young receptors make smart mice. Nature 401(6748):25-27. Giles, Jim. 2001. Think like a bee. Nature 410(6828):510-512. Hughes, James. 2007. The struggle for a smarter world. futures 39(8):942-954. IOM (Institute of Medicine). 2006. Ethical Considerations for research Inoling prisoners. Washington, D.C.: The National Academies Press. Available from http://www.iom.edu/ CMS/3740/24594/35792.aspx. Konar, Amit. 2000. Artificial Intelligence and Soft Computing: Behaioral and Cognitie Modeling of the Human Brain. Boca Raton, FL: CRC Press. Lloyd, Seth. 2000. Ultimate physical limits to computation. Nature 406 (6788):1047-1054. Mills, Jonathan. 2008. The nature of the extended analog computer. physica D: Nonlinear phenomena 237(9):1235-1256. NRC (National Research Council). 2003. The polygraph and Lie Detection. Washington, DC: The National Academies Press. Available from http://www.nap.edu/catalog.php?record_id=10420. NRC. 2005. Aoiding Surprise in an Era of Global Technology Adances. Washington, DC: The National Academies Press. Pascual-Leone, A., N.J. Davey, J. Rothwell, E.M. Wasseran, and B.K. Puri. 2002. Handbook of Transcranial Magnetic Stimulation. London: Arnold.

OCR for page 124
 pOTENTIAL INTELLIGENCE AND MILITArY AppLICATIONS Priori, A., F. Mameli, F. Cogiamanian, S. Marceglia, M. Tiriticco, S. Mrakic-Sposta, R. Ferrucci, S. Zago, D. Polezzi, and G. Sartori. 2007. Lie specific involvement of dorsolateral prefrontal cortex in deception. Cerebral Cortex 18(2):451-455. Routtenberg, Aryeh, Isabel Cantallops, Sal Zaffuto, Peter Serrano, and Uk Namgung. 2000. Enhanced learning after genetic overexpression of a brain growth protein. proceedings of the National Academy of Sciences U.S.A. 97(13):7657-7662. Schemmel, J., S. Hohmann, K. Meier, and F. Schurmann. 2004. A mixed-mode analog neural network using current-steering synapses. Analog Integrated Circuits and Signal processing 38(2):233-244. Siegelmann, H. 2003. Neural and super-Turing computing. Minds and Machines 13(1):103-114. Tang, Ya-Ping, Eiji Shimizu, Gilles R. Dube, Claire Rampon, Geoffrey A. Kerchner, Min Zhuo, Guosong Liu, and Joe Z. Tsien. 1999. Genetic enhancement of learning and memory in mice. Nature 401(6748):25-27. Trautteur, G., and G. Tamburrini. 2007. A note on discreteness and virtuality in analog computing. Theoretical Computer Science 371(1-2):106-114. Vollset, Stein Emil, and Per Magne Ueland. 2005. B vitamins and cognitive function: Do we need more and larger trials? American Journal of Clinical Nutrition 81(5):951-952. Von Neumann, John (updated by Paul M. Churchland and Patricia S. Churchland). 2000. The Com­ puter and the Brain. New Haven, CT: Yale University Press. Wang, Hongbing, Gregory D. Ferguson, Victor V. Pineda, Paige E. Cundiff, and Daniel R. Storm. 2004. Overexpression of type-1 adenylyl cyclase in mouse forebrain enhances recognition memory and LTP. Nature Neuroscience 7(6):635-642. Watson, Andrew. 1997. Neuromorphic engineering: Why can’t a computer be more like a brain? Science 277(5334):1934-1936. Wax, P.M., C.E. Becker, and S.C. Curry. (2003). Unexpected “gas” casualties in Moscow: A medical toxicology perspective. Annals of Emergency Medicine 41(5):700-705.

OCR for page 124