Fail Securely

As systems become more complex, they are also likely to fail in ways of which the exact mechanism might be harder to predict and understand beforehand. With machine-learning systems, there is a bigger problem than failure, which is exhibiting a potentially dangerous behavior as and when failure occurs, as part of the system/sub-system failure. This concern is understandably further magnified in the context of lethal autonomous systems. 

Exploitation surface of a generic ML pipeline

Given the increasingly diverse exploitation routes, the prevailing ideas suggest to have a well trained human operator who is periodically assessing whether an AI is misinterpreting its environment. But conventional human-in-the-loop mechanisms are ill-suited to handle spatiotemporal complexity. Consider for example a large number of potential targets in a tough or no geography and an extremely compressed timeline for making engagement. In this situation it is simply not feasible for the AI to refer back to the human operator every time an engagement has to be made.

If adapted suitably, the old concept of kill-boxes may produce a simple socio-technical solution to the problem of how to detach from conventional human-in-the-loop and embrace independent decision-making in military AIs while still keeping the operator to monitor incoming data. It is also in line with the existing cyber-security principles of network segmentation and granting access based on user's role/location/time etc.

By not wholly depending upon autonomous systems' ability to interpret context and limiting its "full-fledged use" within a human generated spatiotemporal compartment i.e. a kill-box, we would not only impart the AI operation a human-like non-zero probability of making high risk "alpha zero" moves, but also allow more secure failures that cannot exacerbate the larger conflict while providing all the benefits of deploying advanced autonomous technology. This is especially valid for the global-common type environments like the space and the ocean, and of course the internet too; where there is further need to research and manage AI security risks as most nations in such environments are virtually in a persistent struggle with their allies and adversaries simultaneously. 

A typical conventional kill-box. (image source: WikiLeaks)

Ideally, a military should develop its own autonomous systems instead of relying on commercial off-the-shelf products or even allies' systems for those may come with their own inductive biases and are sometimes less likely to fully support complex missions. There are obvious economic, organizational and foreign policy incentives for doing this. And most importantly it would let the machine behavior policy and AI failures be much more clearly defined and adversarially trained against in a manner that does away with insecure failures while also suiting the respective country's cultural sensibilities.

Speaking of latter, complex societies require a fair amount of organized coercion, socioeconomic incentives, cultural deterrence and mutation over a course of centuries to become eligible for the "civilization badge". So it should be natural to want your AIs to reflect those civilizational ethos. Understandably culture isn't an engineer's problem but the way technology (particularly ICT) accords vastly different modes of social interactions and restructures social and even political affairs, it sets the premise for engineering which can function as 'politics by other means'. Perhaps Kaczynski was right. And that's all the more reason to develop systems that embrace failures, and fail securely.

Coalitions & AI

Following a previous article on multinational interoperability, some elementary inspection of complications was needed since the use of AI in the context of modern strategic systems exacerbates the pre-existing modalities governing the formation and functioning of coalitions.

Further to the differences in standards, cultures, and mutual trust, traditional international political-military alliances have the bulk of their operational interoperability challenges resulting from the diversity of doctrines, military technologies and soldiers’ skill-levels. Addition of Autonomous Weapons and Systems (AWS) greatly increases the already substantial operational complexity of these by introducing some major considerations into the alliance relations, chiefly:

THE DATA RIGHTS -
As the experiments with self-driving cars have shown, unlike the human soldiers, AWS will produce and collect huge amounts of multidimensional data, and there are obvious disputes about who’d have access to that data. The data will be of the external environment, also the AWS’ interaction data (with that environment and with other AWS including those of Allies and Non-Allies), the local learning and processing data, data on anomalies/lagging and security/cybersecurity incidents etc. These would present broader knowledge on the conflict and capabilities, and a picture which could be very useful to military planners and intelligence community. Because a lot of such data will be collected privately by autonomous agents, there are not just significant opportunities to achieve mutually beneficial sharing agreements between governments and organizations but also a whole new layer of interaction/information games between states. 

CONTESTED C3 - 
Related to who gets what data rights, is the matter of command arrangements. Decision-making, ethical guidelines, and operational procedures could differ significantly across allies when it comes to cyber-physical autonomous multi-agent systems. Therefore AI systems present a very different set of problems in alliances’ command & communication design. This also presents an opportunity for those nations which have better C3 infra and multilateral options of integration to take the upper-hand in setting the terms of the security cooperation.

BYZANTINE AI -               
The autonomous agents which display arbitrary, faulty, or malicious behavior are termed as Byzantine AI. These could be the result of enemy exploitation also. The management of such resources in an alliance requires special attention and creation of formal procedures, quick consensus mechanisms, and channels of risk communication to interact and intervene, including on behalf of allies, while avoiding any claims of mutual interference. For an example of bad interference, recently Tesla remotely turned-off the autopilot of a customer’s vehicle because some payment wasn’t made, such "security features" may have drastic consequences in armed conflicts, and so far we’ve been lacking a valid governance framework for managing byzantine assets. 

RESOURCE PROTECTION -
Assuming that most wars wouldn’t be a decisive all-out armageddon but follow some sort of political nash-equilibria and that today’s friends can turn into tomorrow’s foes, there emerge some concerns regarding maintaining certain elements of competitive advantage over your allies. There need to be modalities of ML process and intrasystem communication protection, development of mutually-agreed upon safety testing tools, and architectural modularity in the systems deployed, among probably many other things. Further work in modularity and resource protection mechanisms in an alliance would not just help solve the immediate interoperability challenges of a man-machine force, but also take us a step in the direction of being able to address the problem of technology/equipment life-cycle.

Clearly, one player signals too much.

There is a growing body on research in AI alignment which addresses the issues related to cooperation failures, AI’s lack of evolutionarily established signalling mechanisms, approaches for joint optimization, and open-source interactions between AI systems of asymmetric capabilities. All of which should be the top AI safety research questions for states looking to go deploy AI for security related tasks with their allies. And give the fact that most of these tasks are going to be actuated via edge devices, further expansion of security anti-patterns is needed in context of multi-agent learning systems.

Since AI integration is an inevitable and irreversible process, hopefully international alliances would churn out (or adopt) a framework which finds wider applicability across the spectrum of cooperation-competition activity. Also what promptly comes to mind is Napoleon’s somewhat ironic remark that if he must make war, he’d prefer to do it against a coalition. Which is still very relevant, even for security engineers and foreign policy wonks.  

__

Sadbuddhi in Sets

"We tend to forget that throughout history many decision-makers were delighted to accept “double or nothing” tactics if the odds looked sufficiently favorable." Kahn

Too constrained to maximize?

Alexander Kotov, the Chess grandmaster with eyebrows worth a queen sacrifice and probably also a KGB fellow, was quite on the mark while describing a disposition to make judgmental errors and go with a misplaced gut while being constrained with overanalysed choices. The agency of 'constrained maximizers', not necessarily always as rational as in the western conception, has its students in the arena of conflicts beyond the chess board, rooted in its potential for both serendipity as well as misfortune. A Machiavellian adage therefore rightly suggests that true prudence consists in knowing how to distinguish the character of troubles and in the choice to take the lesser evil. In Yogic philosophy there is a word for that, Sadbuddhi, referring to developing the much touted virtue of timely knowing the right thing to do - arguably resolving the groupthink/polythink syndromes that beset most decision making units. This writing is inclined to touch upon political culture and the decision making therein, so we'll follow the lens of intra-species aggression. And although coalationary intra-species aggression spans our non-relatives too, from lions and wolves to even numerous ant species among others, we'll stick to apes likes us.

The term "political culture" itself is recent, though the practices have existed throughout. Like humans, chimpanzees too have a coalationary psychology — that is an inclination to form closely structured groups/ecosystems with strategic hostility toward outgroups, and sometimes even outliers. Other than cooperation, reciprocity and role division - the division of meat and other resources among chimps follows a process of sharing in which power, sex, and violence tend to interact in rather human ways. So different groups of chimps in their territories end up having very different behaviours and social customs, developing different kind of tools and manifesting different communication mechanisms thusly. Reductionism it is, but a chimp order rises with its own crude geopolitics.

The human society too, going by Dunbar's logic, functions in sets of humans. The successful sets have and attract more members (Roman allies rebelled in 80s BC not for secession but for overt Roman citizenship), while losing few to others. This is compounded by the networked nature of an information society. Yet these sets and their members, like those of chimps, never function outside the laws of direct or indirect reciprocity towards each other i.e. I'll help you but when we meet again then you help me. This basic evolutionary dynamics makes Proportionality perhaps the only natural law of war, and other conflicts between sets.

Basis 'Proportionality', the nuclear MAD doctrines would appear as optimized bluffing, for all WMD may be CBRN weapons but not all CBRN weapons are WMD, but that discussion is for another day. As it is, all such interpretations, by nature of self-interest, are biased and susceptible to 'syndromes of the self' i.e. self-deception and self-centeredness. Coming back, it is safe to say that the set-membership nature of society causes, especially in the conditions of present degree of connectedness, an evolving instance of political boundary setting, from physical boundary to the ideological ones shaping member identities.

Basic Structural Coupling

It is well known, that for a system to stay "alive" it must preserve its identity. Though since the gap between perception and reality is also a socio-psychological threat surface, the system must keep maneuvering the environment around it and be fluid enough to evolve along. In the age when goal setting of AIs is termed not an engineering but a political problem, the notions of 'structural coupling' between systems and environment therefore apply equally to political cultures as well. An advice from John Boyd "the maneuverist" suggests decision makers to approach people, ideas, and machines — in that order. However the how-to of such abstractions remains subject to systemic pressures and some degrees of operational and organisational elusiveness, often threatening a conceptually devolving set.

Pursuing a Systemic Sadbuddhi thus becomes a matter of acting on and perceiving reality without letting our beliefs be colored by our desires (and fears), and as Robert Jervis has recommended, making our 'Assumptions, Beliefs, and Predictions' explicit and inquisitory, for it underlines that the predictions are always resting on some assumptions. Taking human society as a whole, the present chimp world order reflects the inertia of the predominant last great war positions. As the economic center-of-gravity shifts from Atlantic to the Pacific and technology gives chimps men equalizing effects over adversaries, how the decision makers of resource constrained competing sets battle it out remains to be seen. One thing is evident though that usually he who would pay the piper, historically speaking, will call the tune.

__

Misdirections & The Internet

"My sculptures are like no other known geometrical solid, but they look just like a cube, a pyramid or another known or absurd solid. This means that three observers at three different points would see three different solids." 

- Guido Moretti

The more intently we are occupied with something, the harder our system works to suppress distractions. Sometimes, we are skillfully distracted away from the real thing. Perception is constructive, following a predictive processing framework, which underlies why illusions occur. But our intent here is not to talk about magic, we'll do that some other time. Here we want to talk about the sometimes coercive influencing of perceptions, the skillful tunneling of our attention, the actors who may use that and the overarching cyber-enabled theater of these tricks.

Our actors may be groups/activists/hackers/guerillas/state/non-state/whatever with a political or reformist agenda. As per their situation, they may or may not use other and often more violent means of action. Our theater can be better conceptualized as a multi-agent system where each agent (actor) always maintains some local state carrying a representation of all the information which the actor has access to. Capabilities drive the actor intentions, but cultural ideologies, psychological and sometimes monetary benefits, as well as network externalities are primary drivers of behavior in this "theater".

There is emerging a lot of material on Russian and Chinese disinformation as well as hacking campaigns, it is arguable that this "material" itself is a countering mechanism of sorts. We don't know lots for sure so let us take another example, from the epic Mahabharat, day 15 of the Kurukshetra war. This because of two reasons, first that no one can object to the truth or falsehood of a mythology, and second, it makes a great example since we can see various elements of coercive influence operations in a single instance.

At the conceptual level, we can see multiple information classes and the leveraging of unique channel characteristics, also a great display of how an influential node can amplify emotions or curb resistance while building greater trust. There are different communication & influence paths, with a continuous notion of bombardment frequency and intensity. The core of disinformation was repeated thrice to Drona by different actors, while changing the nature of appeals with lots of fighting in between. This while the "celestials" were driving Drona's emotions and reasoning away from fighting, there is coinciding improvisation of inducement strategy throughout the operation. And most importantly from a cyber-driven perspective there is fake-forensics (dead elephant) to avoid inconsistency and leave a false trail backed by true evidences, which also allows for plausible deniability as well as perceptual ambiguity towards any bad self-knowledge among actors. Everything from indirect incitements to direct actions, interactions are directed towards the overall operational objective of attacking Drona's willingness to continue fighting by having him distracted and occupied with a falsehood.

The great strategist and the diplomatic finesse of neither fighting nor not-fighting. Image©watchalen

It is well known in general warfare, the general superiority of defense over attack, all things being equal. And even though military strategy is the domain of taking on the stronger opponent and taking on a weaker persistent opponent is what politicians were for, the advent of cyber-enabled operations has somewhat upended the traditional notions, including but not limited to the ideas of proportionality. David Deptula, when talking about armed drones once, noted that an actor can now project power without projecting vulnerability. This maximization of projected power while minimizing the projected vulnerability and the borderless quasi-intangible nature of cyber-enabled information operations, is markedly the defining shift in the conduct of war, arguably somewhat more threatening to the open-internet societies than the ones behind the firewalls.

Nevertheless, for big-brothers with many eyes, intelligence over internet traffic and telecommunications is only the baseline of information gathering. And with greater state capabilities, come innovations from the underground also. We are seeing ideological actors coordinating as leaderless and self-organizing trans-national networks, using less tech and even riding on international relations. Such systems (of networks) over time can become somewhat fault tolerant & resilient to a degree, if the environment is not shaped against them.

Sometimes actor actions can be attributable, but sometimes not attributable, and sometimes even misattributed. It is difficult to set expectations from future engagements if the prominent actions are a misdirection. With a leaderless system architecture comes the autonomy of peripheral attacks which puts the spotlight back over plausible deniability and perceptual ambiguity. We may see strategic shifts as stable actor behavior in such environments, even leveraging of crises and critical political events to widen any sense of distrust, doubt, disagreement and disarray among the adversary.

Come what may, our unconditional security proofs often rest over idealized assumptions. These illusions of security prevent us from making systemic changes. This may range from us not upgrading our vulnerable devices to our dear leaders overlooking contiguous societal necessities, like climate change for example. This is a pervasive human folly, we like the browsing experience so we don't turn off the JavaScript. And there is no best solution to this but akin to much of Sun Tzu tapestry is a Bruce Lee solution, be like water, LOL.

On Rewards

"Any rational society would either kill me or give me my books." 

-Hannibal

Society only tolerates a certain kind of deviance, the kind that has a great or novel expectation of rewards. AI engineers too reflect these tendencies when treating that certain kind of deviance in agents as a refreshing intelligent behavior, given that it is rewarding to the larger interests. A rudimentary example of this could be a bipedal robot figuring out the use of his hands when the walk involves a steep climb by imitating what it observes in other agents, biological and otherwise. As it is, in swarm the most critical information is always coming from the nearest agents which are also considered to be the most reliable. In humans at least, learning does not always lead to a permanent change of behavior, expectation of rewards and fear of punishments often puts us onto the path of overlooking or even misinterpreting our learnings. Yet while the discovery of which response leads to what effect goes far beyond just observations, observational learning is said to be the most prominent method of learning. It should be evident therefore that the attention control mechanism and a hyper-sensitive selection of stimulus and response can make the difference between a satisfactory outcome and the optimal one.

Unlike naturally occuring intelligence, Artificial Intelligence gets its teeth mostly from nurture, as 'nature' mostly applies to its embodiment and hardware limitations. The case of human multitasking is an interesting study in the nurturing of intelligent behavior. A productive use of downtime fetches a much higher reward value than leisure and inertia. We have effective attention mechanisms to cope with long & complex inputs, we apply controlled attention and carry out continuous switching between tasks to maximize rewards. Our perceptual motor expertise gets better with the task familiarity and an economic attentional distribution is then either always rewarded or is at least never punished. More than that, we consistently acquire new goals and build a hierarchy of goals (correlating with hierarchy of needs) which helps us to multitask better and organize our time and life towards greater reward expectancy.

Now if we are to have agents with the ability to acquire new goals on their own, operating under a predetermined hierarchy of goals based on a cause-effect relationship, AI will need a more purposeful nature of being which risks an even willful acceptance of punishments (or negative rewards) towards the greater goal. A finer sentiment sensitivity, a clear demarcation between mission-specific-sensing and non-mission-specific-sensing, and a causality based actuation that draws from dynamic sets of rules seems to be the way forward.

Any finite set of rules, as frames and boxes generally go, is bound to provide an incomplete approximation of reality. As some philosophers have noted, instinct is the most intelligent among all forms of intelligence, and it is least bound by the concerns and deliberations of rewards and punishments as compared to the non-instinctive actions. With more and more seamless and direct coupling of perception to action and an incomplete knowledge-base like humans, sometimes agents too, like humans, can run into a quagmire best described by a remark which Napoleon had made after losing the war in Russia, that, between the sublime to the ridiculous is one short step. Bonne Guerre.

---

Intelligent Money & Cognitive Economies

“We reject kings, presidents, and voting. We believe in Rough Consensus and Running Code.”  David Clark, 1992 

The above quote while originally intended to express many technologists’ view on the matter of internet standardization, also beautifully represents their political and sociotechnical values when it comes to the matter of our economic evolution: that is, rough consensus & running code. 

In An Inquiry into The Nature and Causes of The Wealth of Nations, Adam Smith overcame the constraints of his time in realistically modelling the economic complexity by attributing its emergence to an “invisible hand”. However today we have better techniques to make an accurate representation of the roles, resource flow and the contractual associations among various systems, economic agents and their networks, giving us a possibility to implement “money” as an intelligent self-governing system.

The idea is not new. In pre-cryptocurrency world (2002), Charles Goldfinger briefly wrote about it in ‘Intangible Economy & Electronic Money’, assessing that eventually our finances and money instruments et all will evolve to act intelligently, following an AI agent design paradigm. Thus the global financial system will (somewhat does but not very well) work as a finitely-generated dynamic system consisting of smaller adaptive sub-systems of autonomous money networks and rules for their engagement, each artificial agent having the capacity to form, judge, collaborate and act on their financial world-model. Interesting question is whether we will see a socially differing allocation of value then?

Complexity

While this is an active research area, there have been some recent developments in general causality based discovery models which lower the dependency on time-sequenced data as has been traditionally needed to assess markets, letting us analyze the so far hidden causality structures in economic behavior at the societal level.

Now every once in a while all dreamers have to wake up, but let us still explore this idea just a little further without getting into the nay-saying practicalities of the present-day infrastructure and governments. That being said, gradual and incremental diffusion of cognitive abilities in financial infrastructure will ultimately lead the way. 

Human-to-Machine-to-Machine-to-Human 

With an exploding ‘Internet of Things’ we now even see quantum-resistant cryptocurrencies such as IOTA, into which device maker Bosch' investment semed a good sign for things to come. It is not very hard to imagine that for a varying scale of amount, a user may want a machine to make transactions for him, or even to make the final buying or selling decisions for him. AI after all is a machine to delegate decision making to. Overtime all our personal devices also may learn to emulate and reflect our individual behavior.

So what does this imply? Having an AI agent that interacts with other AI agents and the economic environment on behalf of humans would lead to not only a more socially optimal allocation of assets (and value) but also a more advertising-resilient and intelligent market ecosystem, total market participation being a sum of human and machine activity. The big benefit though lies in the emergence of an economic swarm intelligence at the macro level.

Must overthrow the middle-man.

If we have Intelligent Agents representing buyers and sellers (and intermediaries?), we can make a market out of their supply demand conditions. Most people leave detailed traces of their online activity, so it can be argued that the Agent representing them will have some individual characteristics as well. Let us say that this agent follows a policy iteration algorithm, it has to maximize the utility function of money and the satisfaction of its master. Between its master’s satisfaction and his money’s utility, the agent will make a trade-off somewhere. The issue with utility is that in most markets, it is the intermediaries who control the information and direct the price. It is therefore again the intermediary who has the best estimate of the buyers’ and sellers’ economic potential as well, add to it various search and recommendation systems – and one can see how much markets are driven by the middle-man. The Agent’s information aggregation therefore should go beyond the information provided by the intermediary.

Humans alone are irrational economic agents, mostly making purchases with crafted and limited information. Cognification of economic activity stands to remodel the structural balance of markets, which are already to some degree intelligent, but lack the ability to signal and form internal models. In a complex world, money networks should have the ability to simulate the economy or a part of the economic environment, before proceeding with actions which could have ripple effects on other loosely connected networks. This means digital and intelligent money systems would need more space in the cloud, and evidently have a whole new set of security & architecture issues. 

Integrating The Dark Side

Untraceable & Private

Most of the economics ignores a large portion of economic activity, crime-war-politics. Also the suppression of people’s economic interests anywhere provides impetus to the creation of a black market. A lot of this money is funneled back into the governments as well. It is in the best interest of people involved with this side that their economic activity and the flow of resources is kept as opaque as possible. While currencies like Monero could be useful here, one can argue that cash is still the king and a hurdle in shifting to intelligent money systems.

Nevertheless the networks which have implicit value do thrive on anonymity.

Here I must mention Everett Rogers who gave us the criteria that a new technology must satisfy for incremental adoption:

- Complexity [Can the user understand the use and troubleshoot?]
- Observability [Is the use observed in other people, media?]
- Trialability [Is the tech available for simple non-exhaustive trials?]
- Compatibility [Does it work with the existing social and technical systems?]

The greatest hurdle crypto-currencies have to face is the Compatibility test. But digital national currencies could certainly take a lead in moving towards an intelligent agent based economic approach, if they’re going cashless that is.

It is clear that in such economies, humans will have to share some of their autonomy with the intelligent systems. What are the fundamental limits of this sharing is rather a sociological and game-theoretic question, the answering of which needs to take into account the matter of self-interest vs. public-interest – that is if we are to design such economic governance systems whose purpose is neither to serve, nor to rule. 

_\m/_

On Memory & Machines

Mathematicians argue that it is a mathematical fact, that often the most rational way of making a decision is to simply flip a coin. No wonder, most of human and societal behavior borders on irrationality. But while probability is about numbers, as Glenn Shafer says, it is also about the structure of reasoning. Humans have an ability to execute a continuous manipulation of weight distribution over past events, affecting their immediate actions. Which brings us to the workings of memory and its role in intelligent behavior. 

Traditionally computers’ memory works by recollection through an elaborate system of instructions & pointers to retrieve data. So far so good. But since our goal is designing generality in computational “thinking” - let us look a little further into how a human encodes and carries information, and how it translates into actions. I suggest the readers to have an open mind while also carrying a healthy dose of skepticism, since we’ll be making some assumptions. 

Assumption #1: Memory stored is dependent upon perceptual processes.
Assumption #2: Information is discarded more often than it is stored.
Assumption #3: Good information recall is reconstruction with least bias. 

Two humans may display varying memory of an event they experienced together, depending upon association/encoding of relevance with the input information, the health of their senses and so on. Essentially, to move towards generality, an intelligent agent must have a memory model which supports an active inference mechanism, on top of its core observe-orient-act decision cycle. 

A fundamental idea is that an autonomous agent with general intelligence should be able to fulfill surprise-attention hypothesis. While it is always best to minimize the surprise a system might face, what surprises it ( and its attentional faculties) is closely related to its memory model, how what the system retains is structured and recalled towards optimizing internal as well as external states.

Magicians and veteran criminal investigators have long known the trickeries of human mind. Given misleading information, humans tend to misremember things, so reconstructive memory has been a dangerous rope to walk upon. While generative models are getting better at producing all kinds of information, it will be interesting to see how they evolve and how much so towards reconstruction as opposed to mimicking a classification. It is therefore, a rather interesting problem in machine learning and neuroscience research, to reconstruct an event/memory from the milestones of sensory data and internal states. Eventually this will be getting into the domain of agent’s belief management and causal directionality.

Mr. Feynman The Great has one of his lectures on YouTube where he talks about the computer as a file-clerk who is getting faster and better. He also discusses, almost prophetically, the nature and future of intelligence in machines as well as the dual-use characteristic of technology among other things. Anyone interested in these subjects should spare 75 minutes of his (or her) life to listen to the man. Surely he is not joking.

____

No Explosions But

There is circulating an article on AI, based on the idea that intelligence explosion is impossible and that there is no such thing as general intelligence. Here it is, and it is rather well written, so definitely worth a read. 

"Singularity" as it is talked about is largely a buzzword, exploiting the fantasy of this mythical exponential rise of machines' intelligence. The "explosion" is happening already, the article notes that it is gradual not sudden but misses the bus at " there is no such thing as general intelligence". It is also (unfortunately) incorrect to assume that human civilization works as a single cooperative swarm. To say that those working towards AGI are merely looking for a problem-solving-master-algorithm is an incorrect definition of the problem, which isn't a good place to begin from. 

Intelligence provides the problem-solving ability, it is not the problem-solving ability itself. For example intelligence also provides the ability to delay gratification, exercise caution, predict eventualities, compete, cooperate, or even do nothing - depending upon what it recognizes as its best interests.

In purely evolutionary terms, intelligence is simply the ability to gain advantage over competition. Suppose we make two robots and train them to collect flowers, and give them a way to connect to a network (internet videos?) and learn more things about the task at hand. We then leave the robots in a beautiful field of flowers, so far both the robots are autonomous agents but not necessarily intelligent agents. Now while performing the task robot A figures out a way to pluck flowers in way that is less damaging to the plant and petals, or figures out a route which allows a faster collection than the other robot - then robot A will be considered an intelligent agent (learning & actuating) while the other guy still remains the dumb automaton.

The thing about gradual changes is, that, on a long enough timeline... 

Regarding smarter or general intelligence, it is said that "out of billions of human brains that have come and gone, none has done so". Well, there is a Chinese proverb, something on the lines that those who think something cannot be done, should not bother those who are trying to do it.
___

The Matter of Minds

Electric-Ray fish is a soft smooth-skinned animal which has evolved to grow modified muscle tissues that generate up to 30 amperes of current, the fish needs this power to defend itself in the ocean and successfully capture its prey. Somewhere in 100-200 AD a Greek physician named Galen started to beat the head of his depressed and pain-ridden patients with a live electric-ray fish. His patients immediately felt better. At the time, this treatment was seen as some sort of black magic. Today neuroscientists see it as one of the first recorded attempts to influence the human brain through an external physical stimuli. Since then we have come a long long way.

Early psychologists in their attempt to put psychology on a greater scientific footing, abandoned the study of mind and concentrated on the study of behavior, reasoning that the mind isn’t an observable phenomenon but only behavior is. However as the neurophysiology behind human behavior became more and more elucidated with time, neuroscientists took the lead in cognitive domain and presented the mind as an emergent system that could be tinkered with. Very rightly, neurotechnological interventions have been termed as the “Weapons of Perception”.

As early as in 1973, computer scientist Jacques Vidal described a system comprising of a series of computers and EEG to control “external apparatus such as prosthetic devices or spaceships”. While mind controlled prosthetics are a reality already, various research groups are also vying to operate complex machinery such as battleships and establish secure mechanisms for neural feedbacks. For example there is a futuristic NASA project for sending to space an e-crew — a group of human minds uploaded into a solid-state electronic circuitry — this will not require essentials like air, water, food, medical care or radiation shielding, withstanding extreme acceleration as well. So the size and weight of the spacecraft will also dramatically reduce.

Elliot Valenstein, a noted authority on Brain Sciences while referring to various methods of psycho-surgery and their implications, had remarked that having direct control of another human brain wasn’t possible. Though as technological advances have yielded, there is a spectrum of behavioral change which is now being observed and executed in Neurobiology research programs around the world, radically affecting the medical sciences as well through the rapidly advancing practices such as DBS (Deep Brain Stimulation) and the newer variants of TMS (Trans-cranial Magnetic Stimulation) such as tDCS/tACS etc.

The present applications of BCIs go so far as to using brain waves to train the machines, as recently a group of researchers successfully trained a neural network on P300 responses of the human brain and demonstrated a new approach to Situational Awareness and more immersive human training for uncertain environments that contains shifting stimuli such as a range of firepower, friends, foes, information blockade and non-combatants. Researchers have also successfully tried to implant information in human cognitive memory and presented their observations at platforms like DEF CON.

The human body after all is an electrochemical system that uses electromagnetic fields in everything from cellular communication to the functioning of the nervous system. In his famous analogy while discussing the electromagnetic field interactions with biological systems, Allan Frey states that just as a radio in spite of being present in an ocean of electromagnetic signals is not affected by all of them until they are of the appropriate frequency/modulation - the appropriately tuned signal therefore could interfere with the functioning of a living biological system as well. This becomes much more feasible and easily doable with the advent of Neural-Laces and their possible commercialization by companies such as Neuralink – as it would, as Elon Musk has himself noted, dramatically increase the outbound bandwidth of the biological brain, thereby providing BCIs far greater usability.

In 1879, Edward Page Mitchell had published a short story titled ‘The Ablest Man In The World’ - in that story a man becomes a genius, a political genius at that, because of an intelligence amplifying machine implanted under his skull. Evidently, the future of yesterday has come to become the present day’s reality. The military interest in Neuroscience also arises from the rise of Artificial Intelligence technologies as Neuroscience could provide an architectural roadmap around human affective functioning for modeling emotional intelligence in AI agents as well as to allow machines to inherit human ethics.

Among a plethora of use-cases, the key Neuroscience application areas which should interest the armed forces are Performance Enhancement, Performance Degradation, Espionage and Synthetic Telepathy, Special Ops, and Recruitment. There are billions of neurons in human brain and neuroscientists want to figure out what all of them are doing, and sometime make them do what they want them to do because if the neurons are showing abnormal activity, it is medical science’s duty to fix them up.

There are a some technical issues which are being worked through techniques from computational learning - such as reading and interpreting the brain’s exact output – given there is a very high signal-to-noise ratio in what emerges out of the brain, from cortical to the motor activity – and also generating inputs which can stimulate the specific neuronal firing patterns. Notwithstanding the hurdles – the cognitive domain, after land, sea, air, space and cyber, has emerged as the sixth domain of warfare – the final frontier, is the battlefield between our ears.

Alice & Bob

Alice and Bob are selfish people who should never trust each other, maybe I do not really know, but they certainly use all kinds of cryptography and masks and games to bolster their stealth and pursue their interests, so I’ll say that these two actors never truly trust or promise and always feign some trust in some reality, that is what they believe the state of affairs to be which directs their actions. As an old Hindu tradition says, the universe is falsehood, and it very certainly is for the protagonists of a cryptographic romance and the self-interested players of our interest, Alice and Bob.

Between Alice and Bob, trust does not equate to promise and it is never bidirectional. Also Alice and Bob are not the only people in the world, they interact with other Bobs and Alices too. So it can be said that they are multilateral agents. They try and make a deal that ensures at least some reciprocity in trust and promise and we call this deal a contract, defining the terms of their relationship with other agents and a computational contract automatically adjusts and executes this relationship for them.

Routing of Trust

Historically cryptography has been defined as the study and practice of communication in an adversarial environment. We’d say that for Alice and Bob a benign environment does not exist at all. While they are not the same, in a network that consists of adversaries, Alice and Bob would want to be indistinguishable from each other. That way they can ensure some anonymity from the adversaries while the lack of identification allows them to operate with stealth. So Alice and Bob may not trust each other but they do trust that the others are untrustworthy. And though the routing of value benefits from trust, the networks of value thrive on anonymity.

Distrust is a negative word, so we say that the network should be trustless. In such environments, privacy just becomes the barriers that Alice and Bob put in place for themselves, assuming that everyone, including a global adversary, is out to break those barriers and get to them. Privacy after all is also a mere contract between the individual and the society – like a funnel, at one end of which is the individual and on the other end the whole network/society – the nodes closer to the individual gain more weight and access. The nodes, Alice or Bob we may not know, could present the network with an indistinguishable or even a unique identity which is parsed differently by different nodes as per their relationship with the broadcasting nodes.

The Adversary

One aspect which is generally missed for its difficulty of measurement so far is the mental state of the multilateral agents in adversarial environments. Alice and Bob could be humans, could not be humans, or could be both, as in a team of humans and AI with common purpose acting as a single node. While node reputation can be determined, a node can act deceptively for long periods to accumulate trust of the network before beginning its sabotage operations. Traditional reputation systems do not work to mitigate the risks when Alice and Bob themselves act as adversaries of each other and use deceptive traffic to feign friendly behaviour. Mental state of multilateral agents is even more crucial to determine to understand their motivation, and such states increasingly do not apply to humans alone. Each node must assume the rest to be deceptive and inimical.

The universe is falsehood. For networks of intrinsic value, anonymity of nodes is a technical requirement for survival and resilience, and not a desired attribute. It may still allow for the “protective supervision” of law enforcement to continue but albiet in a saner way. To ponder upon whether Alice and Bob are really good and trustworthy agents is futile - the best course of action as the protagonist in motion picture “V for Vendetta” would say - is to give them a mask, and they will show their true face.
___