Advanced Artificial Intelligence Technology

The good news is that advanced AI technology will be amazingly useful, not at all malevolent, and quite knowable. The other side of the coin is that it will be among the most difficult technical feats ever achieved by mankind.

What Artificial Intelligence (AI) will really be like
Advanced AI, no matter how intelligent, will be purely rationally controlled - and hence, knowable, non-threatening, and eminently useful.

Sigmund Freud and Artificial Intelligence (AI)
(Including audio lectures explaining Freud's concepts of human intelligence)

Paths to Advanced AI - Engineered or Brute-Forced Brain Simulation?
Why engineered AI is the only viable approach to advanced AI

Where is All the AI, Anyway?
Before we get too excited about the coming of advanced AI, we must ask a serious question - where are even the simple forms in widespread use today?

An Empirically-Derived Conceptual Architecture for Artificially Intelligent Systems of Arbitrary Complexity
Exploration of the common attributes of any software deserving of the moniker "artificially intelligent"

How Advanced Artificial Intelligence (AI) Will Learn
Advanced AI will depend as much on experience as raw intelligence in order to fulfill its design objectives well.

AI Rational Analogues for Human Emotions
There is great intelligence in human emotions, and unfortunately, great unpredictably as well. One of the keys to safe AI will be to translate human emotions into rational analogues that are far more consistent and predictable in their operation.

Advanced AI in Government
Many suboptimal decisions by political leaders are based upon the sheer vastness of the data and analysis pertaining thereto. How advanced AI could help.

Proposed Traffic Light Artificial Intelligence (AI) System as Part of a Comprehensive Strategy to Minimize Energy Use, Pollution, and Driver Time Wastage
Advanced AI will not be "one thing" - it will vary enormously, depending on the nature of the objectives for which it has been designed. This explores one possible application of a fairly simple AI application that could render tremendous value.

Microsoft Mapping Course to a Future In Line with "The Empirical Future" Blog
Microsoft's Laura is an interesting, early attempt at something like AI intended to interact with humans - and "her" process of observing her human interactants and deducing subtle cues regarding the character and attitude of these humans fits well with my ideas with regards to this technology.

Advanced AI in Government

Many suboptimal decisions by political leaders are based upon the sheer vastness of the data and analysis pertaining thereto. How advanced AI could help.

It continually amazes me how little discussion of how real, advanced forms of AI might be used in specific applications to achieve great benefits for humanity.

Let's focus on one specific scenario for advanced AI in the sphere of government.

However, let's be real here, I don't mean in a position of "leadership". When discussions of this kind are enjoined, often many instantly and in my view quite incorrectly assume that AI of "greater-than-human" intelligence will simply be handed the reins, take over, as it were. This seems deeply unlikely, for the following reasons:

1. Human beings want to be led by other human beings, not by technology. This is a deep thread of human nature; even if an advanced AI were nominally "better" at leadership, this would still be true, I believe.

2. Even if humans were to get past the obstacle above, if an advanced AI were to take the reins of leadership, and messed up in one way or another, made one or more serious mistakes, who's to blame? Interestingly, not the advanced AI - it would be the manufacturer that is on the hook. The buck will not stop with a technology product - it will stop with human beings.

3. If an advanced AI were to take over the reins of leadership, say become The President, think about it - that gives incredible power to the corporation manufacturing that advanced AI, which most people would find a repellent notion, I would suggest.

For these and other reasons, advanced AI will not be in positions of leadership, elected positions in a democracy, if you will - even if they were nominally better than their human counterparts. So I would suggest setting aside those notions, at least for now.

However, there are still vast realms of value-add that advanced AI could render to support efficient, effective governmental operations. Let's focus on just one - regulation.

Currently, there are 150,000+ regulations on the books in the USA, and these grow by some 4,000 per year. Additionally, to understand these, one must also be aware of the body of legal decisions surrounding them, which are also huge.

There are a truly vast number of regulations, covering all manner of things, such as the environment, workplace safety, etc. Some 100,000 employees in about 60 federal agencies administer these regulations.

Estimates vary, but these regulations impose a cost of some $300 billion annually on the economy, mostly in the form of costs on businesses, not out of the federal budget per se.

Now, these regulations vary tremendously in their effectiveness. A good way to examine them is in terms of cost-benefit analysis, and when discussing regulations, a valuable metric here is "cost per life saved". Some regulations are very cheap - $100 per life saved. Some, massively expensive, up to millions of dollars per life saved. Although we can get idealistic and suggest that any regulation, even if it saves one life, is worth it, no matter what the cost, this can be countered by suggesting that in a world of limited resources, a very expensive regulation that in fact saves very few lives, if those costs were allocated to regulations that save many more lives, that is in fact worth doing. It would be equally unrealistic to suggest that all these regulations are worthless, and should simply be swept away.

However, objective rationalization of existing and proposed new regulations is eminently worthwhile.

Currently, the primary responsibility for determining the cost effectiveness of various proposed regulations falls the Office of Management and Budget, the OMB. They have a budget of about $50 million, and a staff of maybe 50 or 60. They can currently assess about a quarter of the annually proposed regulation, 1,000 of the 4,000, and they presumably focus on the bigger ones, of course.

When considering the effectiveness of not just the annually proposed, new regulations, but also the existing, massive body of regulations, and the cost-effectiveness of each one, and also the objective effectiveness of each one in terms of lives saved, species protected, whatever it is, I think we begin to enter an area that exceeds the ability of the human mind to analyze effectively and well.

Enter advanced AI.

What if you could retain, in your mind, the entire body of regulations on the books, all of the legal decisions surrounding those, with additional metadata with regards to total costs to implement, total lives saved, as well as potentially other important dimensions, say, locality and time.

No human mind can do this, of course - but an AI could, because of course, it's really just an advanced computer, and retaining and processing large datasets is one of their core strengths.

When looking at the political decision process, we often blame the "corruption" of lobbyists, politicians, etc. But really, I think most politicians want to make good decisions - however, the objectivity, comprehensiveness, and consistency of the information and resulting analysis that they receive impedes good decision-making. This gap is what is exploited by lobbyists and other special interests toward their own ends.

This advanced AI application, it's just another application really, preferably within a human form factor as soon as technology allows, would work very similarly to the AI architecture I have already described.

Essentially, it's analysis, and the summarizations thereof that it would provide, would vary with the audience with which it is interacting. If it was the head of the OMB, it might be one, higher level of summarization, estimating costs over time for a proposed regulation, possibly by locale, possibly with rough estimates of legal suits that may result, based on other, similar regulations.

However, human domain experts could "drill down" with this AI - exploring each analysis result, each "assumption, all the way down to the raw data, if they so chose - individual legal rulings, specific costs, lives saved, etc.

We come back to this paradigm of advanced AI as "liaison" between human beings and what we would recognize as functional analysis software today. Very important concept, one that I think will provide almost unlimited value in terms of real, live AI potentialities.

This is just one example, of rationalizing the complex welter of government regulations, keeping the good ones, getting rid of the chaff - and possibly saving billions and billions in the process.

There are many, many more examples like this, that we could name:
1. Rationalizing investments in infrastructure, the cost-benefit analysis decisions there.
2. Investments in Research and Development.
3. The optimal way to preserve our remaining natural ecosystems (which can be considered a subset of regulations, perhaps)
4. Etc.

The point is, the sheer volume of data, the costs, lives, productivity enhancements, legal decisions, etc, are so vast that today, they necessarily must be split into literally thousands of individuals to analyze at all. And there, in an important sense, is the problem - having all of this data in one mind, being able to analyze it as one, integrated data set, there are huge advantages there, I'm sure all kinds of interesting insights and patterns could be discerned there.

But, no human can do this - but an advanced AI could, in principle. In this sense, this application could be "smarter-than-human" - but once again, we see that this is in order to assist human partners, to help them make good decisions. Each human interactant with this advanced AI would have different questions, different avenues they might want to explore. The head of OMB, or the President, for that matter, one set of perspectives. A congressman, another. A lawyer, yet another. An economist or government statistician, yet another.

All of these interactions would, in fact, help the advanced AI refine the analyses that it performs, to better assist the decision-making process and recommendations that all these humans make. It's a partnership, not a replacement paradigm, because many of these people aren't going away. Really, we're talking more about achieving much better results within an existing budgetary environment.

AI Rational Analogues for Human Emotions

There is great intelligence in human emotions, and unfortunately, great unpredictably as well. One of the keys to safe AI will be to translate human emotions into rational analogues that are far more consistent and predictable in their operation.

I would suggest that, far from being the exception, it is almost universally assumed by most people that advanced AI, as it gets closer to nominally human levels of intelligence, will start evincing all or most of the characteristics of that human intelligence. I disagree strongly with this idea in many specific, self-consistent, logical, and empirically-based ways.

However, I want to share another take on this notion that I have alluded to, but not explored in as much detail as perhaps desirable. Let's focus on emotions. Very interesting topic, have had a lot of good discussions here around that.

Humans have emotions, and we consider them an integral part of what makes us special. However, emotions are almost universally shared with most other complex creatures, so really, in important ways they are the characteristic of our minds that are the most ancient and least special of all, at least in the sense that other animals also employ them in very similar ways.

However, the key question for my purpose is, are emotions as we conceive of them really necessary for an advanced AI to successfully achieve and evince hyperintelligent (ie, greater than human intelligence) behavior? I would say, a rational understanding of the emotions of others is critical, especially the primary human interactant(s) with the AI.

Clearly, emotions serve a useful purpose, which is why we have them. Let's examine a very useful one: fear.
http://en.wikipedia.org/wiki/Fear
"Fear is an emotional response to threats and danger. It is a basic survival mechanism occurring in response to a specific stimulus, such as pain or the threat of pain. Psychologists John B. Watson, Robert Plutchik, and Paul Ekman have suggested that fear is one of a small set of basic or innate emotions. This set also includes such emotions as joy, sadness, and anger. Fear should be distinguished from the related emotional state of anxiety, which typically occurs without any external threat. Additionally, fear is related to the specific behaviors of escape and avoidance, whereas anxiety is the result of threats which are perceived to be uncontrollable or unavoidable."

As with many emotions, below a certain level of intensity fear is not incompatible with rationality. However, as with many or most emotions, above a certain level of intensity it can completely obliterate rational thought processes. This makes them unpredictable. This is critical for consideration of advanced AI. Advanced AI cannot be unpredictable in the same way we are.

Let's examine another emotion: Anger
http://en.wikipedia.org/wiki/Anger
"Anger is an emotional state that may range from minor irritation to intense rage. The physical effects of anger include increased heart rate, blood pressure,and levels of adrenaline and noradrenaline. Some view anger as part of the fight or flight brain response to the perceived threat of pain. Anger becomes the predominant feeling behaviorally, cognitively and physiologically when a person makes the conscious choice to take action to immediately stop the threatening behavior of another outside force. The English term originally comes from the term angr of Old Norse language. Anger is usually derived from sadness.

The external expression of anger can be found in facial expressions, body language, physiological responses, and at times in public acts of aggression. Animals and humans for example make loud sounds, attempt to look physically larger, bare their teeth, and stare. Anger is a behavioral pattern designed to warn aggressors to stop their threatening behavior. Rarely does a physical altercation occur without the prior expression of anger by at least one of the participants. While most of those who experience anger explain its arousal as a result of "what has happened to them," psychologists point out that an angry person can be very well mistaken because anger causes a loss in self-monitoring capacity and objective observability.
Modern psychologists view anger as a primary, natural, and mature emotion experienced by all humans at times, and as something that has functional value for survival. Anger can mobilize psychological resources for corrective action. Uncontrolled anger can however negatively affect personal or social well-being. While many philosophers and writers have warned against the spontaneous and uncontrolled fits of anger, there has been disagreement over the intrinsic value of anger. Dealing with anger has been addressed in the writings of earliest philosophers up to modern times. Modern psychologists, in contrast to the earlier writers, have also pointed out the possible harmful effects of suppression of anger. Displays of anger can be used as a manipulation strategy for social influence."

As with fear, below a certain level of intensity, anger has a useful purpose for human beings. However, above a certain level, it often precludes rational thought, and is therefore unpredictable.

A simple example of the inappropriateness of anger as a controlling mechanism can be given if we assume the form factor of a droid for an advanced AI.

If you are the human interactant, and depending on your disposition, you may strike this droid. What should be the droid's response? Well, one response that is NOT permissible is for it to strike you back and knock you through the wall (I expect that these devices will be a good deal stronger than a person, as well as faster, these will be a key part of their value proposition).

Now, maybe the human deserves to get thrown the wall, maybe he's a real a**hole, but that is irrelevant. A device that fights back in this way is unsafe, the manufacturer will get sued, and therefore this type of unpredictable control mechanism will never be designed into a device of this kind. Product safety is a huge and important trend now, and that will continue into the future.

Advanced AI, no matter what its form factor is, can never leave the zone of rational control of every action it takes. If you think about it, this makes perfect sense. Human beings that keep their rational composure, even when someone is attempting to provoke them (this board has lots of examples of this, you don't have to look far), those people that stay in control of their emotions usually stay in control of the situation.

Going back to our example, how should a droid respond in this case? What is the "rational analogue" for anger in this device, in other words? Well, it depends on the situation. If it is responding to its primary human interactant(s), the response should be neither slavish nor intimidating. It should be the response that maximizes the maintenance of respect felt by that human for that device. If this human tends to strike often, the response should perhaps be of a shaming nature, to discourage that human's propensity to strike others, which obviously is probably beneficial for that human anyway. If it is extremely rare, it depends on the circumstances, another potential response might simply be laughter, laughing it off. It is highly unlikely that a human is going to be able to actually injure one of these devices simply by striking it, so the response of the droid can be formulated entirely with the human in mind.

For essentially every emotion, I suggest a "rational analogue" can be envisioned, that fills the role of that emotion, and in fact does it better than that emotion, at least in the context of advanced AI.

How about sadness?
http://en.wikipedia.org/wiki/Sadness
"Sadness is an emotion characterized by feelings of disadvantage, loss, and helplessness. When sad, people often become quiet, less energetic, and withdrawn. Sadness is considered to be the opposite of happiness, and is similar to the emotions of sorrow, grief, misery, and melancholy. The philosopher Baruch Spinoza defined sadness as the “transfer of a person from a large perfection to a smaller one.” Sadness can be viewed as a temporary lowering of mood (colloquially called "feeling blue"), whereas depression is characterized by a persistent and intense lowered mood, as well as disruption to one's ability to function in day to day matters."

This seems very unlikely to be a valuable control mechanism for an advanced AI. Witness the personality of Marvin the Robot from Hitchhiker's guide - that behavior would be excruciating in an actual droid, and would not be successful, I would suggest.

However, here's a very interesting emotion: Empathy
http://en.wikipedia.org/wiki/Empathy
"Empathy is the capacity to share and understand another's emotion and feelings. It is often characterized as the ability to "put oneself into another's shoes", or in some way experience what the other person is feeling. Empathy does not necessarily imply compassion, sympathy or empathic concern because this capacity can be present in context of compassionate or cruel behavior."

Now we're getting somewhere. This will be very important for successful advanced AI. However, this can very easily be approached by an equivalent rational analogue, whereby observation of its human interactant(s) over time provides a rich topographical picture of that person's personality, and hence a rational strategy for empathy devised that is not too shallow, or too indulgent. I suggest a convincing evincing of empathy will be an important part of droid hyperintelligence. But again, it is not "true" empathy - it is evincing of this emotion, for the benefit of its human interactant(s), to provide comfort to them or whatever.

Here's another vital emotion: Curiosity
http://en.wikipedia.org/wiki/Curiosity
"Curiosity is an emotion that causes natural inquisitive behaviour such as exploration, investigation, and learning, evident by observation in human and many animal species. The term can also be used to denote the behavior itself being caused by the emotion of curiosity. As this emotion represents a drive to know new things, curiosity is the fuel of science and all other disciplines of human study.

Causes
Although curiosity is an innate capability of many living beings, it cannot be subsumed under category of instinct because it lacks the quality of fixed action pattern; it is rather one of innate basic emotions because it can be expressed in many flexible ways while instinct is always expressed in a fixed way. Curiosity is common to human beings at all ages from infancy to old age, and is easy to observe in many other animal species. These include apes, cats, fish, reptiles, and insects; as well as many others. Many aspects of exploration are shared among all beings, as all known terrestrial beings share similar aspects: limited size and a need to seek out food sources.

Strong curiosity is the main motivation of many scientists. In fact, in its development as wonder or admiration, it is generally curiosity that makes a human being want to become an expert in a field of knowledge. Though humans are sometimes considered particularly very curious, they sometimes seem to miss the obvious when compared to other animals. What seems to happen is that human curiosity about curiosity itself (i.e. meta-curiosity or meta-interest), combined with the ability to think in an abstract way, lead to mimesis, fantasy and imagination - eventually leading to an especially human way of thinking ("human reason"), which is abstract and self aware, or conscious.Some people have the feeling of curiosity to know what is after death.

Morbid Curiosity
A morbid curiosity is an example of addictive curiosity the object of which are death and horrible violence or any other event that may hurt you physically or emotionally (see also: snuff film), the addictive emotion being explainable by meta-emotions exercising pressure on the spontaneous curiosity itself. In a milder form, however, this can be understood as a cathartic form of behavior or as something instinctive within humans. According to Aristotle, in his Poetics we even "enjoy contemplating the most precise images of things whose sight is painful to us." (This aspect of our nature is often referred to as the 'Car Crash Syndrome' or 'Trainwreck Syndrome', derived from the notorious supposed inability of passersby to ignore such accidents.)"

Curiosity is a fundamental tenet of awareness, and forms the backbone of much scientific and technological progress. It is a very valuable emotion, and I would argue that a great deal of human rational intelligence evolved as a way to inform our curiosity more successfully. Therefore, although an "emotion", it is among the most rational of emotions, at least if not taken to extremes (as in the description of "morbid curiosity" above).

Anyway, I'm still figuring this out. Bottom line, what I'm suggesting here is that the implicit, unpredictable, somewhat mysterious operations of emotions in human and other animal brains have "rational analogues" that can be explicitly programmed, are far more predictable, and hence safer for advanced AI.

These rational analogues will form a natural extension of what I have argued elsewhere will be the rational control mechanisms for advanced AI/droids overall. The inherently rational control of these technologies argues for the knowability in their behavior patterns, even when they reach the point of "hyperintelligence", ie, demonstrably greater rational intelligence than humans. This is my central argument against the "unknowability" of the Singularity (actually, one of several, another being the continued involvement of human beings in their design objectives, for example). It is not entirely clear because it's never been explained by anyone to my satisfaction, but this notion of the "unknowability" of the Singularity seems to derive primarily from the idea that they will have essentially a biological nature to their intelligence, which I suggest is not necessary, safe, or really in any way optimal that I can imagine - except that this notion appeals to our heartstrings, which of course not a rational line of argument.

Microsoft Mapping Course to a Future In Line with "The Empirical Future" Blog

Microsoft's Laura is an interesting, early attempt at something like AI intended to interact with humans - and "her" process of observing her human interactants and deducing subtle cues regarding the character and attitude of these humans fits well with my ideas with regards to this technology.

Here's a most interesting article, discussing Microsoft and Intel's deepening investment in and vision for Artificially Intelligent applications:

Microsoft Mapping Course to a Jetsons-Style Future

As I've indicated many times, Microsoft is quite a logical place to look for widespread AI apps, not just because of their prevalence and wealth, but because that wealth has allowed them to hire tons of top-notch researchers in those fields.

In fact, the dearth of discernably "intelligent" software (as opposed to "functionally rich" software, which although not necessarily in principle different, in execution is in fact completely different) from Microsoft or elsewhere has been a major reason for my reservations about "where's all the AI?" all along. If Microsoft's, Intel's, and eventually other's plans to start to play out along the lines they hope, this could start to address my concerns in a major way.

This article is an indication that, perhaps, purely "functional" software, though still important, may be close to the end of its rapid growth trajectory, and "intelligent" software, that explicitly incorporates AI techniques along the lines below (and, eventually, countless others), are finally coming to the fore as the next big source of software and hardware growth.

And though not described in this article, Microsoft is also contributing importantly to the field of robotics, with the first "operating system" for robotic control (that's been around a while), so that not each robot has to have its own custom operating system.

This supports the point I've made repeatedly that the best place to look for sophisticated AI is not from someone's garage, but from the currently recognizable computer hardware and software industries.

And if that in fact becomes the case, you will NOT see AI that kicks your ass, tries to take over the world, or crap like that. You will see AI that is conceived, designed, and produced as consumer applications, for the purpose of making money.

Some interesting quotes from this article:

"Built by researchers at Microsoft, Laura appears as a talking head on a screen. You can speak to her and ask her to handle basic tasks like booking appointments for meetings or scheduling a flight.

More compelling, however, is Laura’s ability to make sophisticated decisions about the people in front of her, judging things like their attire, whether they seem impatient, their importance and their preferred times for appointments.

Instead of being a relatively dumb terminal, Laura represents a nuanced attempt to recreate the finer aspects of a relationship that can develop between an executive and an assistant over the course of many years."

If "Laura" can do things like the above well, she will be light-years ahead of Ramona, Microsoft's former PC "assistants" such as Bob and the paperclip, or any other chatbot I've seen.

What's Laura doing? Deducing subtle cues from being observant of its human interactant(s), a key aspect of ostensibly intelligent behavior.

I describe a far more advanced version of this same approach, in the context of droids, here:
Envisioning the Hyperintelligent Droid

If you choose to explore this link, scroll near the bottom, where I describe the droid/advanced AI skills of supersubtlety, hyperobservancy, and ultracoordination. The reason for the "super", "hyper", and "ultra" designations is because this is for an ostensibly very advanced, hyperintelligent technology. Of course, before we get to these superlatives, there will simply being able to deduce subtle cues from observant behavior, which the path that Laura, and Microsoft and Intel, are on.

An Empirically-Derived Conceptual Architecture for Artificially Intelligent Systems of Arbitrary Complexity

Exploration of the common attributes of any software deserving of the moniker "artificially intelligent"

Since I get lost in terms like "conscious", "super-intelligence", etc, and I struggle when people say they're building systems that have these characteristics, I've been thinking a lot about how to describe and model an AI system in simpler terms. Here is an early iteration of my attempts on this front.

1. Input mechanism(s)- This is simply the way(s) to get raw data into the AI system in the first place. The number and sophistication of these depends greatly on the nature and objective of the application, of course. For a chatbot , it could simply be lines of text; for a an advanced droid, the full complement of human senses will probably eventually all have a place, because each has a potential value-add for its human interactant(s), as I've already described in my blog.

2. Awareness - The ability to filter the raw data gathered by the sensory mechanisms to fulfill the design objective(s) of the system. For a chatbot, this might include colloquialisms and specific nouns used by the user that can be used to determine more customized, relevant responses to that user, while perhaps filtering out common articles such as "the", "a", "an", not dissimilar to what search engines do now.

3. Curiosity/Experimentation - the continuous "motivation" to refine or optimize the performance of an AI system "thinking" algorithms, usually through varying the input parameters that feed that algorithm in some way. Obviously, the degree of freedom with which this can be done depends on the maturity of the algorithm and the safe range of variability of the input parameters. The way that these input parameters are varied would probably be controlled another algorithm of some kind. A far more sophisticated type of machine experimentation might be the development of entirely new algorithms, though if this in time becomes possible it would seem that these might not be "closed form" algorithms, but numerical approximations developed as a result of the optimization parameters over time. An analogue of this might be how a computer determines the strength of an electric field in a certain charge geography via finite-element analysis, as opposed to solutions of Maxwell's equations in closed form. If such an approach proves viable, the advantage would be the automated or semi-automated development of algorithms for a much wider variety of optimization problems than are currently available; the disadvantage might be, these algorithms would not necessarily be describable in "closed form", and hence might defy easy human understanding of these algorithms.

4. Thinking - This is the computation of the output that a given AI system will produce, based on both its inputs, and thereby an assessment of how successful its previous output(s) were based on those inputs. This will often be algorithm-based. For an AI traffic light, this might be as simple as a minimization of cars waiting and pollution emitted, both of which are simple metrics that can be quickly gathered by sensors; for a hyper-intelligent droid, this would likely expand into hundreds or thousands of algorithms, that over time may mirror in some ways the thinking sub-systems in the human brain.

4. Output - the behavior(s) that represent the fulfillment of the AI system's design objectives. For a chatbot, this would be a line of "conversation". For an AI traffic light, the timing of its green-yellow-red pattern. For a droid, the particular service it renders, or response it makes to a query.

5. Understanding - This is the ability of the AI system to evaluate its inputs in such a way as to consistently discern the effects of its outputs on meeting the overall objectives that the AI system is designed to fulfill. This may or may not be a separate subsystem; in many or perhaps most cases, the algorithm(s) that comprise it's "thinking" module would ideally perform this function. However, it is useful to consider "Understanding" as at least a separate logical subsystem.

6. Memory - the retention of all the past interactions for the above 5 characteristics already mentioned. It amazes me how often this is forgotten; I have yet to see a chatbot, for example, that remembers any of its past conversations, or at least that gives any indication of such memory.

This model accommodates the evolution of systems that seem intelligent (or at least, less stupid) from developer skill-sets and programming languages that are functionally oriented. This is critical because functional software represents by far the biggest portion of both in the current software environment.

Despite the use of anthropomorphic terms for certain model elements, such as "Thinking", "Curiosity", and "Understanding", hopefully from their corresponding definitions it is clear that these are not necessarily equivalent to their human-brain analogues in terms of objective or complexity. As with each of the model elements that have a sophistication and complexity that spans a very wide range, and is determined primarily by the objectives that the AI system is designed for. In other words, the result of a successfully executed AI system incorporating all of these elements should be not necessarily be considered "conscious".

Considering this model in terms of more abstract terminology such as "machine consciousness" brings up some ideas. Obviously, a conscious machine that does not include all of the items above would not be valuable for most conceivable applications that deliver real value as consumer or business product. On the other hand, a machine that is conscious that includes all of the above, whatever goes beyond that I am still attempting to understand how that would be described and value-delivering in the context of a real AI system. To be clear, I am absolutely not criticizing anyone who is approaching the problem of advanced AI from a "consciousness" point of view; this could very well be a function of my own limited understanding in terms of what consciousness actually means. On this front, I am not the only one who is perhaps confused; what "consciousness" actually represents is still a hot topic of debate.

Therefore, the main way I can imagine a "conscious machine" being valuable is, if what is meant by that is a generalized "kernel", if you will, that allows the relatively rapid customization and linking together of the above AI system components for a wide variety of AI application types., of varying sophistication, from chatbots to AI traffic lights to advanced multimedia engines to advanced droids, which is my quick and dirty (though certainly not comprehensive) continuum of distinct AI types that I can quickly imagine.

Like all of the ideas in this blog, this model is empirically-based, not declarative or theoretical. In other words, it describes in a consistent way all of the various flavors of AI that I have described, and that I have seen elsewhere. Those informed the development of the model elements, not vice versa . However, it would seem that these model elements are a very useful to envision or describe new or existing types of AI systems that have not yet been considered or vetted against these components.

Why the Fear of Rogue or Dangerous Artificial Intelligence is Fundamentally Irrational

We humans are driven by primarily irrational control mechanisms to want more than it's fair share (the id), and often unrealistic expectations of oneself and others (the superego). The weakest part of the human psyche is the rational mind, the ego.

The main thing about advanced AI that scares the willies out of a great many people is that simply by achieving an ostensibly human level of intelligence, that this will automatically result in a structure exactly like our evolutionary-derived, biological brains. Really, that's what we're talking about here, that's why we're afraid. Another way of putting this is that many people make the intuitive (yet deeply flawed) assumption that human intelligence is the only kind of advanced intelligence. This is a near-automatic presumption that receives very little subsequent examination. This was the real genius of Freud, that by careful examination of a very large number of patients and others, he was able to discern this tripartite nature correctly, in particular the unconscious nature of the main drivers of human behavior. It is worth pointing out that this tripartite nature of the human mind has been recognized for millenia, and was described by Plato in a very similar way, and Marx as well. The main difference is, Plato and Marx described this tripartite nature in an idealistic, hopeful way, not an empirically-based (ie, scientific) way. Freud was the first to approach the macro-structure of the human mind in a scientific way. However, he was preceded in some ways by older thinkers such as David Hume, who echoed many of Freud's later analyses, especially the dominating (id-driven) nature of human behavior. Freud's biggest insight was the unconscious nature of these dominating drivers, which makes them unfortunately very difficult for most people to therefore understand even their own minds.

The relevant thing to note about all this is that these will NOT be the motivating drivers for AI programming or behavior, no matter how "intelligent". For one thing, to replicate unconscious motivations in a synthetic, man-made product would not only be foolish, I'm not sure it's even possible. Computer scientists don't understand Freud's concepts any better than anyone else for the most part, so the things that would make AI dangerous aren't even on their radar, they're not going to design that kind of thing. Even if they did understand these concepts, to make an irrationally controlled droid or advanced AI would be tended against by liability, ethical, and other concerns that mold the modern consumer marketplace as much as, if not more, than the technology itself.

I'm not passionate about endangering the sanctity of the 'unknowability' of the Singularity, but really, I'm not so sure there's much unknowable about it, as it will really come to pass. This is because the main thing, probably the only thing, making the Singularity unknowable is this confusion about what AI will really be like. When you make the simple assumption that AI will be just like human intelligence, sure the future is unknowable (or rather, I would suggest quite knowable - knowably, extremely dangerous).

But, we've seen that replicating irrational control in a synthetic product is not only foolish and unnecessary, it is no where suggested that this is an even remotely likely trend from any technology evidence available today - software or hardware. Not one iota of evidence supports this contention that AI will be irrational like us.

Therefore, to persist in this guttural fear is just another facet of human irrationality, and is not based on the actual nature of the technological systems that will eventually be designed and built.

I predict with great confidence that this irrational fear will vanish once hyperintelligent systems actually become available and business and consumer users have experience with them. Unfortunately, I believe any system that can be remotely described as hyperintelligent is a ways off. In all likelihood, mobile robots for the home will probably lead the way, because multi-tasking robots in the highly unstructured environment of a home seems the most demanding of the various aspects of intelligence such as sensory awareness and responsiveness, dexterity, vocal communication rather than programming, etc.

The most sophisticated robots in general use in the home now are quite simple and perform one task only - the vacuum-cleaning Roomba, the Robomow Automatic Lawn Mower, and the IRobot Security Guard. The (quite slow) extension of these single-purpose robots into multi-tasking robots that can clear the table and wash the dishes, for example, represents a large increase in intelligence of these devices, and will be the first real step toward the hyperrealistically human droid that I eventually foresee. However, even a droid that can clear the table and do the dishes may still be light years away from anything resembling hyperintelligence, to be sure. However, it will be an important first step for moving the comfort level of consumers from their irrational fear of dangerous AI to the status of any other technology product - a useful device, manufactured product, designed and created for its value in human service.

Hyperrealistically human droids, whenever they arrive, by the time they arrive will have the exact same status.

Sigmund Freud and Artificial Intelligence

Intelligence is a very interesting topic. Most smart people think that the way that they personally are smart is the best way to be smart, or even the only way. But that is a notion grounded more in personal confidence than reality. Our human intelligence is as variable as our fingerprints. Each person's intelligence is a complex mixture of natural endowment, motivation, and experience. Instead of one standardized test, if we were advanced enough to devise them, each person would have their own customized test to assess their particular style of intelligence.

However, a much bigger mistake would be to accept the common wisdom that human intelligence is the only possible model for an intelligently behaving AI system. The premise that this is the only path to a well-behaved, "moral" robot is, I suggest, deeply inaccurate, and an AI system modeled exactly on the human mind would be quite undesirable in many ways.

Our human brand of intelligence is actually 3 distinct kinds of intelligence, sometimes cooperating, sometimes competing, and real AI systems will have different mixtures of this combination that make will make them safer and more commercially successful. To understand the different facets of human intelligence, we need to go back to Sigmund Freud. I have found Freud’s characterization of the human mind into components of id, superego, and ego amazingly useful for understanding real people. And as it turns out, Freud’s model turns out to be even more valuable for understanding how AI systems will be architected.

Briefly, Freud describes the human mind in three components which perpetually struggle for dominance: The “id” is the animal brain, the most primal, and is the source of our emotions and also the interpreter of other’s emotions, as well as other creatures, such as assessing if an approaching dog is friendly or not. The id is by far the strongest of the 3 components, and influences mightily almost everything we think and do.

Ironically, the id represents many of the things we consider central to being human – this, the most ancient part of our brain, the one with which we share the most similarities with other animals. It's the wellspring of human creativity. It’s also the source of human strategic thinking. When a CEO determines he wants to double a company’s size, especially without the obvious appearance a major opportunity, that’s an "id thing”. Office politics is an id thing. All the emotions are id things. The ability to recognize faces is there too, most likely. And lots, lots more besides.

The“Superego” is the next most powerful component. Among its important functions seems to be a control mechanism for the human id, which without a control mechanism would rage largely unchecked. It is the product of the values your were raised with by your parents, church, school, and other primary influences. All the opinions that are strong in us, baked in and quite resistant to change, are superego. The important point is that this part of the brain is largely uninfluenced by the rational mind. Ideally, a given opinion is carefully vetted by the rational mind before being committed to superego, yet probably a healthy majority of our superego’s conceptions have not been through an this objective. Whenever you judge instantly and strongly, that is your superego talking, not your rational mind.

Last and perhaps unfortunately least in terms of influence is the “ego," the rational mind, the objectively analyzing and problem-solving brain. It is a distant third to the other two in terms of its power to influence thought processes and resulting human actions and thoughts. In fact, in many ways the ego isn't really a control mechanism at all, in terms of influencing our behavior - rather, it is a tool used by the id to further it's survive, thrive, and procreation objectives. The myriad manifestations of these three objectives permeate almost everything we think, say, and do; sometimes in obvious ways, sometimes very subtle. Thus, the rational ego is relegated to a far less dominant position in the scheme of human consciousness, at least when viewed from the perspective of a control mechanism.

I would suggest that many if not most complex animals have these three components as well - ids, superegos, and egos. Arguably, these are often optimized in different ways, and it seems clear that we have the most advanced egos, ie, rational analysis capability, of any creatures on the planet. This can be explained by considering that rational intelligence has been our major evolutionary strategy for the last several million years and the gray matter invested in this is the greatest of all animals. We don’t give enough respect to our animal brethren in this regard, that the fundamental nature of their intelligence is not so different from ours, but of course molded in myriad ways to make the most sense for their particular evolutionary strategies. Things like fast legs, hard shells, and sharp claws and teeth reduce the need for rational intelligence as a survival mechanism to some degree for the vast preponderance of Earth's creatures during the entire history of life on this planet.

So, loosely put, human intelligence is primarily id-driven, with the superego providing a powerful id-control mechanism as well as "instant" judgment on many topics, and the rational mind helping the id execute its desires as needed. The mistaken notion that we have 1 mind, not 3 working in synchronicity (and often competing with each other) is the reason we consider the total of all that stuff swirling around in our heads as “intelligence”. And since that is what intelligence is, the thinking goes, we assume that an AI-powered droid must also have these same constituents organized in the same way in order to behave intelligently.

This is an excerpt from a very informative lecture explaining the relevant ideas of Freud in more detail. Not only are the id, ego, and superego carefully and clearly described, these are compared and contrasted with the tripartite views of self described by Plato, and of Marx:


This lecture describes the consequences of Freud's psychological components on civilization overall, contrasting his ideas carefully with those of Marx:


Listen carefully, perhaps more than once (it took more than once for me for it to sink in). But the resonance is deep, the insights profound.

Proposed Traffic Light Artificial Intelligence System as Part of a Comprehensive Strategy to Minimize Energy Use, Pollution, and Driver Time Wastage

Advanced AI will not be "one thing" - it will vary enormously, depending on the nature of the objectives for which it has been designed. This explores one possible application of a fairly simple AI application that could render tremendous value.

Before we get too caught up in the world of droid AI, which is an exceedingly advanced version of AI, there are undoubtedly all kinds of simpler applications that might be described as reasonably proactive, doing something that is essentially a no-brainer that you almost certainly wanted to have done, without explicitly telling it.

I must admit, I have been caught up in the Kurzweilian spirit of altruism. Therefore, I will will step away from my narrow consumer focus for at least the duration of this post. What is a relatively simple AI application that within its scope of operation could accomplish the following objectives:
1. Reduce time wastage for millions of people, every day.
2. Minimize energy wastage.
3. Minimize pollution

And do all these at the same time, in a non-competitive way.

There are many examples of such low-hanging fruit all around us. Lets look at just one, the staid world of traffic lights.

How many times have you been sitting at a stoplight for 2, 3 minutes, no cars coming from any direction? It seems like an aggravating eternity, and you’re wanting to run that light, but you don’t.

You always think, what’s the deal, I'm the only one here, why doesn’t it turn green? We unreasonably expect it to display that intelligence, every time, for some reason. In my case, its because some lights seem to do this, but this could be coincidence, but that faint expectation persists in the back of my mind for every light. However, at least most the traffic lights are on a preset schedule, that doesn’t take into account individual, moment-by-moment traffic patterns. However, sitting at a traffic light unnecessarily, or even slowing to a red light and having to speed up as it turns green as you’re about to stop, wastes energy, wastes time, adds to pollution, with no benefit.

Let's project a maturity of AI technology that would enable awareness by a traffic light of its intersection. Not amazingly advanced awareness, but well enough to respond differently to different traffic flow situations.

Toward this end, if each light pointing along an intersection quadrant, each pointing along the street that it controls the flow from, could detect what was in front of it, from cars stopped right in front, to approaching cars, say, 1000 feet away, and gauge their approaching speed and acceleration/deceleration. If a car is approaching, and no other cars are coming, even if its been red for only 5 seconds turns green again, before you even start slowing down it turns green again for you, maximizing time and energy conservation.

Also, this approach would allow behaviors that are today quite uncommon. For example, often the left-turn only arrow at an intersection are paired, so that two streams of cars can turn left at once, across the intersection from each other. However, if one left-turn signal runs out of cars, then the opposite side of the street should be allowed to proceed forward.

The main thing to remember is, it's all about keeping as many cars moving through that intersection with as little delay as possible, while still giving everyone a fair chance to proceed without unduly long waits.

Of course, this is the simplest example, and things get quickly more complex with different numbers of cars coming from different directions, and closely spaced intersections, each with its own traffic light.

One light can't let too large a number get by on an individual optimal traffic flow decision, and have them pile up at the next one, a ¼ mile away. It cant be a total energy-minimization and time-minimization decision. If that were the case, one car on a less-busy street might need to wait half and hour before it gets a green, while hundreds of cars on the busy street crossing it go by. However, if that same light turned green just long enough to let that car by, and then immediately turned green for the cross-traffic again, that’s intelligent. Again, we wouldn’t say, what a smart light, but if it didn’t after doing that for a while, you’d be swearing at the incredible lack of intelligence of that light.

Obviously, the case of a single traffic light considered in isolation is the simplest scenario. Really, we want that traffic light to be communicating at least with its nearest neighbors as well, especially if they are in near proximity.

If there is a wreck on the highway nearby, the traffic light needs to become immediately aware of that, and change its behavior accordingly. So that when a large amount of traffic starts flowing off the freeway into the sphere of influence of that traffic light, there aren't ridiculously huge backups simply because that traffic light isn't responding to that changed condition.

So, it will be great when the intelligent traffic light comes along. However, consider this - this seems like a fairly simple AI app, compared to some we will describe. And there is absolutely no way that PredictionBoy is the first one to think of this, so it seems reasonable to suggest that since others almost certanly have, and maybe have even tried to design one, and found it quite tough to do so.

So, that will be a great product, look for it. But until you see it, or other products like it, with even small amounts of intelligence (or slightly lower stupidity), don't be looking for humanoid droids, or even metal men, C3-POs. Remember, we must keep our hard hats on, keep our expectations realistic. It will be a while, these are very tough problems. But get there we will, and it will be worth the wait.

Further reading:

The Application of Artificial Intelligence to Transportation System Design

How does a traffic light work? Is there an A.I. consciousness at work?

new electrosensitive traffic light using fuzzy neural network

Fascinating facts about the invention of the Traffic Light by William L. Potts in 1920

How does a traffic light detect that a car has pulled up and is waiting for the light to change?

What The Singularity Will Really Be Like

The point at which humanity is able to create an advanced AI that is more or less equal in complexity to our own fills us with both elation at the possibilities and dread at the uncertainties, as I have noted here. The Technological Singularity essentially says that as this level of advanced AI approaches that of one human intelligence, and when instructed to improve upon itself, that will lead to an open-loop, runaway intelligence explosion that will leave, as one reputable source puts it, “humanity far behind.” This common conception of being "left far behind" is due to the equally common belief that the nature of improvement of advanced AI intelligence will be upward, into greater and greater strategic planning and implied world dominance, as opposed to downward, into techniques that will help them be aware of their environment and interact with humans more effectively. The case in this blog is that it will be focused down, because that is the direction of maximum value add to human consumers and businesses.

There are Singularity conferences, and futurist and technology forums hum with what The Singularity will mean for mankind. These discussions are invariably non-specific, and center around either vague elation at a benevolent super-AI taking care of our every need, sort of a God-figure, or vague dread at a malevolent AI, that would enslave or exterminate mankind. With all that discussion, however, there is little exploration of how realistic the Singularity concept actually is, or what a real Singularity would look like on the real planet Earth.

It is important to remember that the Singularity is not an empirically derived extrapolation of current trends; it is a science-fiction concept, developed by Vernor Vinge and later appropriated by Ray Kurzweil, who in typical fashion took an armchair speculation and said, well, now it's real, now it's really going to happen, based on the progression of Moore's law primarily.

This entry will engage the Singularity in a specific and constructive way. There are several challenges with the Singularity concept which are consistently evaded by the true believers that I try to submit them to. These are:

A. It makes the implicit assumption that our future selves will simply turn our backs while this technology performs a runaway intelligence improvement loop.

B. It gives no idea what is meant by "greater intelligence".

C. The sacrosanct "unknowability" of the post-Singularity world is based on the implicit but strong assumption that this hyperintelligence will be driven as our human minds are - ie, irrationally. If this hyperintelligence is in fact rationally controlled, this unknowability conceit dissolves, there's very little evidence to support it.

I don't think the Singularity as currently envisioned is likely, for the above and other reasons, but neither can I rule it out. However, we can put some very real and reasonable boundaries around this "unbounded" event, that make the Singularity (or what I call a "steeply-sloped event") far less unknowable in absolute terms.

The perspective of this blog is that any any hyperintelligence “explosion”, should that scenario prove to be realistic, will take a finite amount of time to complete (not vanishingly small), for the following reasons:

1. Unless the hyperintelligence is infinitely great, the time to solve any problem will be finite, not instantaneous.

2. Even if the design, theoretical, and scientific problems can be resolved in an instant by one of these advanced droids, the engineering actualization will still be of finite duration. In other words, unless these droids are capable of the instantaneous transmutation of matter, factory lines will still need to be retooled, etc, to create the more improved design.

3. These will take the form of consumer and/or business products, and people and businesses have finite abilities to absorb new offerings (even in the future).

4. Humans will still be in charge. For the purposes of this unbounded improvement cycle, even assuming maximum autonomy by the advanced AI tech, humans will still at the very least select the feature sets that are improved in this way.

5. The form of the droid intelligence will be such that self-motivated autonomy will be as alien to them as it is to PCs today – regardless of the level of that intelligence, it’s nature is fundamentally alien to these type of id-driven characteristics. So not only will they not be selecting their own feature-sets to improve, they won't be selecting the problems to be solved with the resulting intelligence "product" - humans will.

6. The nature of the improvements to the identified feature sets will not be a computationally bound problem, but one requiring intense interaction between humans and droids. In particular, the deepening sophistication of hyperobservancy and/or super-subtlety skills.

There are a number of reasons to believe that a improvement loop of this kind will still be of a finite rate of advance and time to fulfill. However, to maximize the vividity of this scenario, we will assume that a significantly advanced advanced AI intelligence tech can improve upon previously identified feature sets essentially unaided, and do so quite rapidly.

Therefore, I won’t assume infinite progress, but something very substantial. Say, when a droid intelligence reaches about 1 human intelligence, it goes through a self-improvement cycle from that level to, say, 1 million times greater in a product that can be actualized in a year. Not infinite, but very substantial. The 1M number is somewhat arbitrary, it can be bigger if you like

For a starting point, I will use the AI architecture I have already described. Not only is this the most realistic one, if you read the trades carefully you will see that it is the only kind ever mentioned. This is simply the most straightforward way to achieve hyperintelligence with today’s computer hardware and software infrastructure.

The likely advanced AI arch will initially be strong in the area of rational and logical reasoning, since that is the core strength of today’s computer architectures. In fact, it is hard to get them out of that sweet spot without difficulty.

These early droids will be good enough to do many chores, basic communication, etc. However, they’re lack of id-simulation sophistication will make them limited in their imagination actualization qualities, ie, we will have no doubt that we are hanging with a droid, regardless of how human it may (or may not, initially) appear.

Therefore, this big swath of functionality will be nascent in the earliest droids, and a definite target for improvement.

Now, as I’ve described in the Envisioning the Hyperintelligent Droid article, these sensory and behavioral sophistication will be manifested as the “superpowers” of supersubltely and hyperobservancy. There will undoubtedly be others, but the value of these are so strong that they will be part of the feature profile of these devices.

I didn’t note this in that thread, but its important to say that one of the most valuable uses of hyperobservancy will be in the detection of super-subtle signals. Even more important, the accurate interpretation of these signals, not just noticing them in the first place. This doesn’t just mean that when the owner twitches his eye in a certain way, it means he wants lunch; it also means the correct identification of subtle meanings hidden in even quite dramatic and well-known events.

Hyperobservancy of super-subtle signals, especially the kind communicated by other humans, is much more a core human skill than droid skill. Even so, most of us do that poorly, I would suggest - if you ponder silently how many times you have misread someone or something, you may concur. So droid tech has quite a challenge cut out for it. It must not only be able to detect these signals, but accurately interpret and act on them correctly, almost all of the time – say, eventually with a success rate above 99%. As I noted in the droid thread, that in effect forms an infinite sink for HI multiples.

I’ll bet you can see where I’m going with this. I will assume that even if this putative, largely unsupervised intelligence self-improvement explosion were aimed at this problem (the one thing we humans get to do in this scenario, name the feature to be improved) and set off with a cycle of progress equal to the force of a nuclear explosion, the main effect would be to rapidly improve these 2 skills, along with perhaps others I haven’t named yet.

And yet, what is the effect of rapidly improving these skills? It makes them even more compatible with, complementary to humans.

In other words, the hyperintelligence explosion will not leave humanity “far behind”, rapidly diminishing specks in the rear view mirror of droid autonomy. (The droids lifting their shades slightly, to smirk at their stranded creators, while we frantically wave our arms for them to come back with our car.)
It will only make them more compatible, more complementary with us, faster.

So even in this most unknowable of all unknowables of some future models, the intelligence explosion, does not leave us behind but draws us closer together, using the architecture described in this blog. And this architecture is simply the one that will be built from the starting point of the current computer hardware and software industry.

This architecture is the one being realized, and therefore the scenario above is the most likely result of any intelligence explosion.

Where is all the AI, anyway?

Before we get too excited about the coming of advanced AI, we must ask a serious question - where are even the simple forms in widespread use today?

The thread "What AI will really be like" is the good news – that advanced AI is knowable, and will be non-malevolent. Now, for the less than good news – AI software maturity appear to be nowhere near the level needed to support many of the prognostications in this area. If you predict fantastic advances in 10 years, there should be at least some simple ones now, preferably with robust trends of consumer adoption, so the key predictive tool of "consumer behavior" can be employed to solidify understanding of what form(s) this technology will actually take.

One challenge with much of the treatment of this field is that we go from nothing to the most advanced applications imaginable – forms that seem exactly human, or even exceed human-level intelligence. There are few if any near-term, intermediate manifestations of this technology, only the ultimate culmination of it. If we can’t imagine a rich set of intermediate applications, we may be hard-pressed to make it to this ultimate objective. No other technology has gone from zero to super advanced in this way, so let’s not presume this technology is fundamentally different. Perhaps the opposite, in fact, with many more iterations that fall short of the holy grail of competitive with the human brain, but are nonetheless still quite useful in their own right.

The high-level architecture in the "What AI Will Really Be Like" entry would be more for a system that is intended to interact with and act like a human. I think most of us would consider this to be what AI means – to act and interact like a human with humans.
Although I feel simulated intelligence is a more accurate term for this field, let me use this term in a different way, so things don’t get confusing. I will use simulated intelligence to mean systems that exhibit behavior we might describe as intelligent, but not necessarily in a human way. This will usually, but not necessarily, mean a “simpler” form of intelligence than AI is usually taken to mean, but not necessarily.

How many of us interact with nonhuman systems, computers or otherwise, that we think of as “intelligent? Not functionally rich, intelligent. Even the most advanced, commercially available software, such as Windows Vista, is quite functionally rich – but does anybody, as they use this OS, or any other, think, “you know, this software sure is intelligent”? Honestly, with just about any system I've ever used, that thought hasn’t really crossed my mind, to admire its intelligence and I might wager that the same is true with most everyone else.

It's not really in our nature to admire the intelligence of others, with notable exceptions. If we have trouble doing that with each other, then manufactured products are going to find it even more difficult to have their smarts admired in this way. Because of this, it may prove fruitful to approach this from another direction - not making products "smart", but making them not be "stupid".

When we approach simulated intelligence in this way, there is a lot of insight to be gained. Though few of us have praised the smarts of any of our products, we have certainly cursed at them, been aggravated with them for this or that clueless behavior. There is more to work with when approached from this direction. We can glean a lot more in terms of business requirements, and when we achieve a minimization of stupidity in our consumer products, we have in fact found a way to increase their intelligence.

Let's turn the coin over and examine it in this way. What would it take for us to think of some of our current products, computer or otherwise, as “not stupid” (or at least as, not as stupid as they currently are)? Big question, lots of different answers, but in its simplest form, some criteria might be:
- Asks you a question once, and uses that answer to inform its behavior (at least when you’re using it) thereafter in ways that help reduce our effort to achieve our ends. As just noted above, these can be identified less by thinking of them as intelligent by the presence of this behavior, as stupid and eventually highly irritating in the absence of this behavior. A top of mind example is when you plug in an external disk drive to your computer, up comes the prompt, “what do you want to do, copy the pictures, play the music, etc, or do nothing.” You reply, saying, “do nothing,” and click the box that says, “do this action every time”. Yet, every single time you plug in that drive again, or any other, you have to answer that question again. This is a simple example, and there are already lots of confirmation and preference dialogues in many OSes, but this type of “intelligence” is really to us obvious “lack of intelligence” by its absence, or especially, when it asks, you give it the answer, and it forgets, does it a different way, or otherwise differs from your clearly stated preference. In this perhaps simplest form of intelligence, the system asks you a question, you give it your answer, and they system does it that way for you, every time, in every way that seems reasonable. Occasional reiterations of the question are OK, as long as in most ways we feel like the system "gets it" in terms of our preferences.

- To move on to a higher level of intelligence not in common use today, in combination with asking questions, the system observes your behavior over time, asks more comprehensive questions, and makes conservatively realistic preference deductions of a broader scope from that information. This can still be far from a human-like level of deduction capability. To pick an example familiar to myself, in image processing software, individuals process images in a number of different ways, sharpening, saturation, color, etc, in a way that over time that individual can see similarities in the way that they process similarly pictures with similar characteristics, low-res, high-res, orangey, red, etc. an image processing software suite that could detect these as well, possibility with confirmation from the user, could in effect proactively process large numbers of digital images or movies without the

This higher level of intelligence is more complex, and very rare in currently available systems, to the best of my knowledge. In the image processing example, after a session where you have processed a large number of images in different ways, individually, batch script, etc, at the end of the session, the software asks, remember how you processed images like these? You say yes, and perhaps an additional button, “apply my processing preferences” to an image or images thereafter, and it applies not only its functionality, but in the way you have tended apply it in images with similar characteristics, eventually in a way where after clicking that button, you say, wow, this image is prefect, I don’t need to adjust it any more.”

- Is this behavior intelligent yet? Well, its getting close...in any case, its damn convenient, saves lots of time, and makes the user much more productive, possibly much more.


Clearly, we’re not going to be regularly praising the smarts of any non-human system until they routinely show more impressive powers of deduction that we typically do. I will describe those in another thread, but may require something like the AI architecture in the previous thread. However, well before that there are an exceedingly rich set of potential SI applications that will deserve at least the consideration of implicit intelligence simply because we will not think of them as aggravatingly stupid – in other words, seemingly aware of what we’re doing, and doing something that makes sense to us, aside from a mindless, rote schedule.

It hopefully seems clear that these simple SI applications, while a bridge to more advanced AI type of behaviors, are obviously still conventional computer software and hardware type technologies. Some of the simpler forms we’re familiar with, others we’re surprised or annoyed that they don’t exist, regardless of the true reasonability of that expectation.

This is an extremely cogent point, intelligent behaviors will evolve, probably gradually over the next few decades, in ways that build on the technologies in current use, that we are already familiar with. There will be discontinuities, but they will be the exceptions to the rule that we already live with every day - technologies progress gradually, in ways that are seldom jaw-dropping, but nonetheless quite useful in the long run.

Other than the asking of simple questions, and collecting detailed preferences, each explicitly communicated by the user, hardly any of the others are apparent in any widespread system that I can think of at the moment (give me some examples, if you have, very interested).

So I ask, where’s the AI? We’ve been hearing about AI for decades, and we’ve tried different forms that had limitations of one kind of another – expert systems, for example. I’ve heard of some promising directions in this field based on advances in understanding the human brain, and we’ll see where those go. But again, where is it, in terms of widespread systems?
Where should we be looking for it, in the first place?
Naturally, droids of the kind I will describe will certainly have a strong incentive to have sophisticated, seemingly overtly intelligent, software. This is rather obvious, but this also will be among the most advanced flavors of this type of technology, so we’re looking at the finish line trying to see who wins the race, when the starting line has barely (or even) been crossed yet.

One place I would suggest would seem like a reasonable place to expect to find these early on is in computing operating systems. Ill pick on windows, since I’m the most familiar with tat, but the same comments apply to all the ones I’m aware of.
Why isn’t Vista more intelligent? I'll defer to Microsoft for the real answer, of course, but here are some reasonable attempts to answer that. One answer might be is that it isn’t necessary. That’s possible, but I would suggest unlikely. A more realistic tentative answer might be that there is so much to do in terms of developing and integrating requests from users for what are more straightforward functionality improvements. Also, things like making the underlying software infrastructure more accessible to third party developers, improved handling of graphics, etc.
Doing those things already require massive numbers of brilliant developers, working together in tightly integrated teams to produce exceedingly complex software that is backward compatible, thoroughly tested, etc. each iteration of the OS requires many years of effort by those teams to get the next version out. significant delays are not uncommon, which implies that even this traditional mode of large scale software development is becoming so complex that it is very difficult to estimate the time to complete with much accuracy.
In addition to those very sound reasons, there may another, less obvious one. It is conceivable that the techniques of developing intelligent software behaviors, such as advanced, implicit preference identification and extrapolation, are to some degree orthogonal with the techniques of traditional software developers. OSes versions, like all software versions, have a large amount of momentum, and the software developers for that software have techniques that they are comfortable with, and extend in a predictable way. Adding to the difficulty is that any intelligent behavior, if it requires significantly different developer skill sets, must be smoothly integrated into the much larger body of the functionally-oriented software.

So, there are several possibilities, which apply across the board to just about any software application from any company:
- Few behaviors were requested by the user community that could be described as qualitatively different in terms of intelligent, as opposed to simply more functionally rich.
- Intelligent behaviors were considered, deemed too complex, and deferred
- Intelligent behaviors were considered, attempted, and dropped because they were too complex (I think this happened with vista once or twice, not sure)
- Most user requests were communicated in the form of functional enhancements based on the current, known version (the menu does this, I would like it to do something additionally)
- Intelligent behaviors have been collected, and some kind of AI-supportive infrastructure is being developed or considered, and will be forthcoming at some point.

To some extent some or all of these may be true. However, it is my suspicion that for most software being developed today, neither the users in their requests, nor the development teams in their enhancement processes, are really even thinking in these terms to much extent. In other words, the vast majority of momentum across the entire software industry and almost their entire user base, conceptualize and execute software in terms of functionality enhancements or modifications.
Functionality improvements are of course highly important and exceedingly challenging , I'm not suggesting anyone’s slacking anywhere. However, I would cautiously suggest that, really, very few are even thinking in these terms in the first place, ie, intelligence enhancements, so I'm not sure we really know how valuable or not they would be if carefully selected and well executed.

A very good countering thought is, at some point does the current era of functionality enhancement software sets morph into recognizable "intelligence" incorporation? To some degree this may turn out to be true. But I would suggest not necessarily, in and of itself, without modifying the way software development is considered and executed.

In any case, the more fundamental doubt I would express is that there doesn’t seem to be a truly flexible, adaptable, and powerful AI architecture out there at the moment that would be a good starting point for incorporation into our current portfolio of conventional software apps.

So my answer to the blog titles’ question, "Where is all the AI, anyway?":
At least in terms of widespread applications, its simply not there yet. It has been a long time, 40 years or so, and has evolved through many forms of very specialized utility, but it is still not to a point of general applicability. I suggest that this is due a couple of factors. It is a very, very tough problem, and deserves great respect for that difficulty, and the most obvious application, human-like interaction, is the toughest. And to a lesser but still significant degree, we’re not really thinking in those terms, we think in terms of functionality creation and enhancement. Of course, the development of a general purpose AI architecture will most likely lead to slowly changing attention to intelligent behaviors.

That will be a long process, perhaps taking decades, or maybe just years. Its difficult to estimate, because its hard to say where we really are in terms of this technology, without at least one widespread, commercially available product to study and understand where we are now.

My point is, AI gets some breathless coverage, because it is quite exciting. Some of the current research is more promising than in past decades, based on our improved understanding of the human brain. However, it is important even for an optimist to be open-eyed and bare knuckled, in order to possess an optimism able to withstand the light of objective evidence. We should respect the immense difficulty of this space, and be patient – it may be a while before these kinds of technologies are just around the corner, in terms of being in a store, just around the corner (or on the Internet, you know what I mean – commercially available).

Because it is so tough, it is imperative that research be pointed in the directions most likely to yield promising results. The choices don't differ in their results by a year or two, but perhaps decades.


Further Reading:

Why A.I. Is Brain-Dead, interview with Marvin Minsky

The Singularity Cometh? Or Not?

What Artificial Intelligence (AI) Will Really Be Like

No matter how intelligent, advanced AI will be purely rationally controlled - and hence knowable, non-threatening, and eminently useful.

In coverage of artificial intelligence technologies whether in paper, movies, TV news, or the net, there is always this anxiety bordering on dread of the projected point when these artificial brains, presumably in a robot of some kind but not necessarily, reach and then eclipse the power of one of our own human brains. Up to that point, the coverage is excitedly optimistic. But at that point and after, it takes the tone of, we’d better be afraid, perhaps really afraid. Almost all of the coverage has either this tone of quiet dread, or remains optimistic, but with no real idea of how this superintelligent technology will actually be used in the future. The possibilities are endless, but it's almost like, they'll be the smart ones, we can't possibly imagine how they'll actually be used from our perspective here in the present.

Respectfully, I disagree entirely - these droids and the uses to which they will be put are quite knowable, because we humans will be manufacturing them within constraints, legal and otherwise, that we would recognize today.

The future will be different, but not entirely different in every single way. Corporations aren't going to turn against consumers with malevolent technology, and much of this thread will show that the droids won't get there by themselves.

Have confidence in our future selves, we are not going to lose control of our products in this way. It's funny to say that, because not only is there little risk of our products getting away from us, to get them just to their intended design behaviors will be an immense undertaking, making Vista look like DOS - or a simple "Hello, World" program. We won't have to monitor them to make sure they don't get out of hand, that's the movies. We will want them to help. In fact, some of these techs may positively require their significant help in order to be achieved - they could be too tough for humans alone. I say that to hit home the tech complexity, not to slam humans, I'm mostly referring to some of the most advanced droid techs covered later.

This general theme of things getting out of hand once AI tech gets to the human+ level is quite widespread. I just did a Google before I wrote this so that I wouldn't address something already past, but the scary or unknowable Singularity and related concepts are entrenched as common wisdom, I would venture to say. We need to address this, because these chains of perspective are deeply flawed in important ways. In fact, in just about every way, except for the prediction that computing power will some day get as powerful as a human's.

Taken together, these two perspectives represent the vast majority of opinion and coverage on this subject. These will be put to a well-deserved rest in this and following threads. Both these views are incorrect, and are therefore limiting in terms of their predictive value. Looking beyond the Singularity is kind of like saying, the sky is blue because God made it that way - and just waiting for the droids to turn on us isn’t very helpful either. These technologies will neither be spontaneously malevolent, nor are their applications beyond our ken from here in the present. We can't see everything, of course, but by carefully processing the future we can see a great deal more than is currently believed.

In regards to the potential malevolence of advanced AI systems, the thinking generally seems to be that once these devices get a little past 1 human intelligence multiple, they will look at us like the inferior life forms we are, and they will use that intelligence to enslave or kill us, as a short step to taking over the world.

Really what the proponents of that view are saying is that, because they will necessarily have to have “brains” organized very closely to our own to exhibit truly “intelligent” behavior, they will be subject to the same vagaries of intent as our own. But since they’ll be lots more smart, they win. In other words, we fear that they will be like us, because we have a history of using our intelligence aggressively, both between ourselves and with the planet's resources as a whole.

This represents the sharp horns of a most serious dilemma: advanced AI needs to be modeled more or less exactly like our own intelligence in order to behave as humans deem intelligent; however, we understand that our own intelligence is often selfish and unpredictable, so we feel inherently uneasy with replicating these control mechanisms in the intelligence of our advanced consumer products.

This approach is untenable not only because of the vast complexity of creating a droid that is driven as we are (as we will describe below). This will take decades longer than necessary, and the result will be a Pyrrhic victory for humans, because the control mechanisms for ourselves are elusive and ephemeral. To replicate these control mechanisms in a consumer product would be problematic, and perhaps ultimately a mixed bag of pros and cons that at the very least could suboptimize their market success. If I am correct in assuming from the limited evidence available that most advanced AI research dollars are currently going in this direction, then we could be spending billions and losing decades in an ultimately unproductive direction.

I predict that when advanced AI technology is finally consumer-ready it will embody a different approach. That is, it will be more a perception-management engine which is designed to respond to and interact with a given human (usually its owner) in such a way that to that person seem absolutely, realistically human. It does not matter if the droid truly feels emotion, for example, as long as it seems to a human that it does truly feel emotion. In other words, that the emotions are a high-enough fidelity simulation to be indistinguishable from an actual person to the perception of a real person .

This may seem a little vague or confusing, but not to worry - we have scenarios galore to illustrate what is meant by these concepts, in later threads.

It may be that current approaches seem in some ways quixotic in terms of pursuing an advanced AI tech that "really" thinks is because we don't really have a good handle on the business requirements, the detailed understanding of the applications we're designing these for. Things like the Turing test are good checkpoints, but hardly inform the detailed direction in which the research dollars should be putting to best use. Details are important, because when a enough details are accumulated, one can often step back and see a pattern, a cadence, that they take, which can in turn sometimes lead to shortcuts, optimizations, that produce the desired result in a shortened time with the important functionalities intact.

When a technology is too far away in time, it suffers the same effects that distant physical objects do in space, small, indistinct, fuzzy. Visionaries are those instruments, futurists provide the focus on these distant lands. If that land is going to be a real one, those visionaries need to have hard hats, with feet firmly on the ground, because much of the future is nuts and bolts, like today. Let's put on a hard hat to address this problem, this most central core of almost all interesting advanced future technologies.

Really, an architecture that seems exactly human is as good as one that is truly human, and if that architecture is more stable and predictable, it is better. If that architecture also makes it more straightforward to incorporate greater than human intelligence in ways that are valuable but nonthreatening, then that architecture could be considered superior.

Droids will eventually have everything they need to take over the world, save one – the motivation. The ambition, the will to power, the ability to feel contempt, hatred, or any other emotion, for that matter, good or bad. Without this, their other powers are not threatening at all. They will be able to simulate emotions, but not be controlled by them – a profoundly critical difference, of the first magnitude in importance. Without these emotions in the driver's seat, they will be about as likely to exhibit truly malevolent behavior as a PC today, seriously, no matter how smart they are. Much of this thread will be devoted to supporting this statement.

It's time to be fair to ourselves and acknowledge that the nature of emotions and how they affect our behavior is of monumental complexity, in absolute terms – just because we don’t think we’re thinking about our emotions does not mean they don’t require vast computational resources, both in our own brains and any man-made system. If for no other reason, their great complexity means that there is almost a nil probability of their spontaneous, unintended appearance in a droid or related consumer product of any kind. The main source of unintended behaviors today in computers are software bugs, and bugs create problems, occasionally even make your computer freeze up, but they don’t exhibit highly coherent, complex behaviors. That’s why it’s called a bug, usually localized in effect, very rarely if ever coherently working with another bug to produce a substantively different software functionality.

The nature of emotions and their control mechanisms are so complex that I doubt science understands them enough to provide a detailed description, much less the conceptual framework for replicating our emotion drivers in a manufactured product. Remember, scientific knowledge leads engineering knowledge, sometimes by very substantial timeframes; and if our scientific knowledge is still tackling the basic how do emotions work, how do they control us, then it will be reasonable to expect a very long wait for droids driven in this way.

On the other hand, with the architecture suggested in this blog, emotions are less critical instruments of behavioral control as they are displays for the benefit of its owner. What proportion of whatever id-like behavior is simulated will be in response to the cues from its owner - so the deepest mysteries of emotion and the id don't have to be really solved, just solved enough to make them successful consumer offerings.

Don't get me wrong, I'm not suggesting that this architecture will be easy - it will still take decades to realize. However, it will be realizable in a way that an emotion-driven droid will not, for the forseeable future.

From studying present and past technologies the lesson is abundantly clear: you don’t get something for nothing – complex and unintended behaviors do not appear spontaneously in a human built system, no matter how advanced the technology gets. Any desired system behavior requires work, usually a lot of work, by lots of smart people who have no reason to design a malevolent product, and every reason not to.

You don’t get something for nothing - that would fall into the “we should be so lucky” category. This kind of spontaneous creation of new and unexpected behaviors in technology is almost like believing in magic, and spontaneous creation of highly complex system characteristics are about as likely as real magic – they simply don’t happen, and are not likely to happen, even in the far future and with incredibly more advanced technologies. Not only will that fact remain intact, it could deepen considerably because of the much greater complexity of those systems. We seem to think that spontaneous, complex behaviors are more likely in a more complex system, but actually the reverse is true – they are even less likely, because the spontaneous behavior would have to coherently and significantly affect even more intricately interconnected systems in more and more improbable ways.

The malevolence concern is so extremely implausible that the onus is really on those who claim it could happen, and exactly how if so. If a solid, detailed, logically consistent case can not be made, then we must acknowledge that this is a fear unsupported by any real evidence, and wisely set it aside for the time being, until actual evidence supporting that belief does surface. Until then, it's magical behavior and therefore not useful for predicting the future.

In any case, I predict that the concerns about AI systems turning malignant will evaporate almost as soon as real products are introduced. Much of the anxiety is simply because we don’t have experience with those advanced AI products. If we look at the most formidable computing machines today, supercomputers, they are amazingly fast and powerful, but we don’t consider them likely to spontaneously develop motivations and take over the world – at least I haven’t seen serious concerns of that nature recently.

Far from being malevolent, the exact opposite will be true: advanced AI systems built in the future, even those with vast multiples of human intelligence, will by their very nature and architecture not be competitive with human intelligence, but amazingly complementary. The PC today has that role, and these future techs will feel complementary in the same sort of way, but orders of magnitude more sophisticated. These devices will provide immense value to us, into the far future.

To go further, we need to explore the nature of intelligence more clearly. In particular, let's see if the widely accepted belief that any advanced artificial intelligence must be architected just like ours holds water or not.

Intelligence is a very interesting topic.

I don’t believe this almost universal conception is correct, and completely disagree with Kurzweil and others that these systems will require an organization essentially identical to our own brain. In fact, it will be almost completely the opposite, and it turns out that the type of intelligence exhibited is not threatening and even quite complementary to humans, amazingly so in fact.

So, we have 3 intelligence subsystems, id, superego, and rational, to work with. What mix of the 3 does an optimal AI system require? For example, a droid that must interact smoothly and effectively with one or more people, while executing many other tasks as well. I will devote a thread to explaining in detail how and why, but for now accept that the optimal mixture of the 3 will be to make the rational engine as the true control center, with id-interpretive and id-simulation engines, which are used as necessary to convince humans of their naturalness, or humanness. Their emotions are purely for our consumption, they don't control the droid's behavior at all. The rational control center is orchestrating the simulated emotions, and will be completely in control at all times.

This is an important distinction from almost all efforts I hear of today in terms of ultimate objective: current approaches are seeking to have the "emotions" of these droids actually run things in the droid's behavior and actions. This is a far more complex, inherently unstable, and ultimately unnecessary holy grail to try to attain.

Think about it. These are consumer products, we are not intentionally creating a new, autonomous lifeform for its benefit. We are not God, we are consumers, and we want a creation that is as much tuned to what we want as any technology developed today. Remember, these people will be real people, think about what you would do if the products were available - what would you want? This blog will describe a fair attempt at answering this and many other questions, discarding the future's halo and just walking around in it like the place it is, our place now, but with a little more treadware on it. Probably one or two pretty bad things might happen in the next 100 years, but in the big scheme, probably not stop man's trajectory of progress. Lots, lots more knowledge, both scientific and engineering. At some point, the knowledge itself becomes difficult to dislodge, so it would take a pretty big calamity to come near stopping or even significantly slowing that juggernaut.

Despite our amazing knowledge gained, we must acknowledge that these are very tough problems. Emulating the human brain - the most complex object in the universe, that we know of - pretty big boots to fill. Let's attack this as consumers rather than semi-Gods and do it right, pragmatically - 100 years seems like a long time, but it is nothing. The eons stretch ahead, and for our civilization to weather those eons, we'll need to get sustainable very quickly, and these technologies should provide immensely powerful tools to help us get those equally big challenges under control.

This brain will be a synthetic creation (at least for quite a while), an immensely advanced version of a microprocessor computer system today. Our current systems are very challenging to make, but once made, there is no evidence that any computer circuitry has ever rearranged itself in spontaneous ways. To proactively address some concerns out there like, "well, the id-simulators might take over, then you're in trouble again" - trying to get that exciting malevolence back into the picture, but please refrain - there is absolutely no reason to believe that the probability of this occurring is anything but vanishgly small.

The potential for malevolence in our droids is kind of attractive in weird ways -sexy, dangerous. But there is not the slightest support for fears of this kind, so even though they seem a little less cool because they won't get medieval on you, once you set that preconception aside much progress can be made in divining the amazing value proposition these consumer products will provide.

The advanced AI/droid brain will be much advanced, but should be able to leverage the chip and software industries infrastructure of today fairly well. It will be like a super-PC, very different but alike in one important way: its degree of emotion in terms of the decisions it makes and actions it performs. Our current computer systems do not seem to be getting more emotional as a parameter to consider in their design and production.

Superego-type simulators may be needed as well in some form, also primarily for our benefit. Consider their superego to be the legal code that its owner is operating under, customized for locale. This is for its owner's interest almost entirely: to keep its owner and itself out of legal trouble, because being in jail means that owner's weal is way down. A droid convicted or implicated in any legal trouble could mean big problems for its manufacturer, which they will want to avoid as a high priority. The very best way in almost every case to avoid trouble of this kind is to prevent its occurrence in the first place, and most likely will be exclusively applied to social situations, when someone in addition to its owner is involved.

So, a rational control center runs the show, and interacts with id-interpretive and id-simulating learning engines as needed to seem human in character, with the droid's interactions bounded by the human legal code.

It may not be apparent, but this organization is a huge difference from humans, almost as different as it can get. Of central importance is that in the droid brain, the id-simulators aren’t driving the droid’s behavior – they are being used to understand its environment to some degree, especially interpreting the gestures and actions of human beings. It is also being used to simulate id-like responses appropriate to the situation. These behaviors will seem absolutely indistinguishable to us from “real” emotions at some point, but that’s because simulations that are sufficiently sophisticated seem to be real, they suspend our disbelief. That will be the point, to suspend our disbelief with high-fidielity simulations, but generated differently from how our real emotions are.

This difference severely undermines almost all of the worries around rogue systems of this kind. This is because when you think about it, our ids are the source of almost all our bad behaviors. This is primarily due to the fact that it developed in a time far before civilization, in a more rough and tumble world. Many of these “primitive” motivations are incompatible with civilization, or at least must be repressed more of the time. Droids will be designed and built with primary directives that are as important to them as our survive and sex drives are for us. However, these will be different in ways that will are non-threatening to and much more compatible with human beings than we often are among ourselves.

For example, our id-driver of survive and thrive is unbounded. This is especially true for the survival instinct – it is usually absolute, and we will react as vigorously to a friend trying to kill us as a stranger. There are exceptions of course, such as a hero giving his life to save others, or a parent giving his life to save his child. These are exceptions to what is generally a very powerful, in effect absolute, will. In droids and other AI systems, this survival instinct, instead of being absolute, will be carefully bounded by hardware and software to be appropriate to its design purpse. That is, it might resist being turned off by a stranger, but be fine with being turned off by its owner, for example. A more extreme example might be a battle droid that is a fierce killing machine on the battlefield, but docile as a kitten when being turned off by its field commander.

The idea of a bounded or limited survival drive seems strange to us, but that’s only because we don’t have that. We think of it as part of intelligence overall, but that isn’t correct either – it is far more basic than intelligence, a core component of biological creatures competing with each other and many other creatures in a complex, violent world. However, there is absolutely no reason to believe that a technological system of this kind could not be developed as easily as one with absolute survival drivers, like humans do. In fact, for droids, when engineering design decisions are made for these devices, a bounded survival driver will be far more preferable in almost all cases I can think of. This is another key difference that I suggest delivers a major body blow to fears of rogue AI systems. Carefully bounded in essence means carefully controlled. It may seem that the line between bounded and absolute survival is thin, and that a droid could switch from one to the other unpredictably. To the extent that this ever seems to happen to any degree, this will be a bug, not a spontaneous, magical behavior. This is a good thing – bugs are time-consuming to find and eliminate, BUT they do not generally in isolation or combined with other bugs produce highly coherent, complex behaviors. In others words, a bug isn’t going to produce a robot that suddenly has true emotions, which would be required to do things like take over the world.

The second big id-driver of sexual procreation for humans will be replaced by design in droids by something like “maximize my owner’s weal and happiness”. This completely selfless driver seems strange to us, but that’s only because we don’t have that, at least as an all-consuming motivation. But, again, there is absolutely no reason to believe this driver cannot be designed and manufactured in a system of this kind. The drive for sex is not a basic ingredient of intelligence – it is to fill this need that much of our intelligence has been organized by the demands of biological evolution. And of course, lots of other animals have this need as well, whether or not we consider them intelligent.

So, the advanced AI systems controlling rational-core engine will collect data, passing appropriate information to its id-interpretive engine, and responding with a mixture of rational feedback, mixed with id-like simulations of human behavior as appropriate.

At some point we’ll be able to interact with a droid or other AI system of this kind, and it will behave exactly like a real human being – to the extent that we as humans can actually suspend our disbelief that they are not. They may talk, respond, laugh, whatever, like real human beings, but they are totally different inside. Remember, its all pretend, it’s all a simulation. In fact, I think that “simulated intelligence” is a much better moniker for this field. It's less threatening, and more accurate in terms of what’s actually going on.

In addition to the above, there are additional reasons that this AI architecture will be preferable to one more closely modeled on our human brain. The biggest one is market efficiency. I predict it will prove far easier to build a rational-driver brain than an artificial brain that is truly id-driven while being also stable and predictable. Our id is a deep and mysterious place, filled with lots of things that could be hard to justify in a manufactured product.

Although it will still be a huge undertaking, creating a rational-engine driven AI brain with id-interpretive and id-simulating components will be a far easier and straightforward effort than one that is id-driven. The comparison of the two efforts may warrant the Vista-DOS comparison again - it would be much harder to drive a droid with emotions, and more risky, than the rational-centric architecture described here.

So multiple forces – stability, cost, as well as performing all the functionality required – strongly suggest this will be some flavor of the preferred AI architecture for the majority of advanced AI systems. The relative importance of each may vary, and as we’ll discuss, things like the id-interpretive engine may grow quite large in later models in order to detect and interpret subtler and subtler human signals. Regardless of the relative size or importance any of these components attains, these will still be logical, rational-engine driven. We think we are rational driven – these will be rational-driven in a much truer way. To the extent that they can be described as having id-like drivers (which will, by the way, be controlled by the rational engine), those id drivers of “bounded survival” and “maximizing its owners weal” are nonthreatening. And because we’ve seen that highly complex behaviors don’t spontaneously appear, even in systems of this sophistication, there’s really not much to dread in this technology. This also has important implications for the idea that their intelligence can be measured in terms of multiples of human intelligence – that they will get to 1, then 10, and someday immense multiples of human intelligence – in the millions, perhaps, and higher. It goes almost without saying in every other source I’ve seen that a droid with 1M times a human’s intelligence would be as far above us as we are a cockroach, which is why we think that sooner or later, we’re going to be prime targets for squishing by these droids. However, the discussion above reveals the fatal flaw in fears of this kind. The architecture of their mind is so different, that to truly compare their total mental capabilities with our mental capabilities is not apples to apples, but apples to oranges. To make a thorough, reasonably accurate side-by-side comparison would require not one number, like the universally accepted multiple of human intelligence, but hundreds, perhaps thousands, of metrics, many of which we probably aren’t even skilled enough to devise at the moment. Our ids drive our brains, but in ways that are far more elusive to describe thoroughly than their reasoning control centers.

But of course, companies that will make these are not going to try to communicate a large number of comparison metrics to the future’s consumer base. Then as now, they will simplify these into a small number of metrics that help guide buying decisions, but are far from a complete picture.

For example, one metric might be logical deductions per second. It should be clear that this is the sweet spot for this AI architecture, and they will outshine us greatly, in the same way computers today are far faster at processing numbers than a human could ever try. Our computers today are millions of times faster than us at many types of calculation, but those are in ways that we appreciate, that help us, that complement us.

Another potential metric might be a fairer comparison to our own natural strengths, say, emotional cues interpreted per second. I can say with confidence that it will be a long, long time before an AI brain, or any other technology we care to ponder, will be better than we humans at that, that is one of our most basic skills, deep within our id, optimized as an important driver by evolution.

I have no idea which metric or small set of metrics will eventually be used to compare droid and human intelligence. However, if forced to venture a guess, I would say that companies then will do what companies today do - select the metric or metrics which makes their product seem most impressive. This would imply that multiples which stress the droid's natural abilities will be the most commonly used. If this is the case, if for example logical deductions rather than emotional cues are utilized, then my high-multiple being a counterintuitive or misleading metric is even more true. Be that as it may, at some point the complexity of these droid brains will equal and then eventually far exceed the complexity of our own. But this doesn't mean they will be any more potentially malevolent than earlier models, because the same considerations apply almost irrespective of complexity.

Here’s the non-intutive, common wisdom-violating result of comparing two radically different types of intelligence with one or two metrics. It may take 100s, 1000s, or even higher multiples of “human intelligence” as represented in these droids to approach the level of skills at reading cues that a human reads and interacts with another person now instantly. Our “animal,” id brains are in fact immensely powerful and perform impressive feats of computing without even our conscious thought, and what is simple for us will require very advanced AI brains to even approach our skill there.

We don’t think much about these computationally demanding skills that the id performs much of the time, or tend to discount them, but that’s because we can do these without really “thinking” – so we make the deeply incorrect error that they are “easy”, and will be covered by these droids early on, with a few multiples of human intelligence. However, then as now, what is easy for them will be hard for us; what is hard for them, will be easy for us. That is the basic fulcrum of their complementarity to us. Over time, much of their computational investment will be in doing things that are second nature to us.

However, even given these apples-to-oranges, imperfect set of whatever human-to-artificial intelligence metrics are adopted, I will in a later post describe how even truly immense multiples of these metrics, in the millions or higher, of some of these future systems will still have plenty to focus that intelligence in ways that are not competitive, but complementary, used in ways that its individual human owner appreciates deeply, but is not threatened by in any way. Fundamentally, instead of going up, into higher and higher levels of strategic, world-conquering ambition or what-not, these multiples will be focused down, into observing and analyzing more and more subtle signals communicated by its owner, other humans, and its environment in general. As we’ll see, going down is as much if not more computationally demanding than going the other way. And the design and manufacture of these will keep them within their design boundaries, no matter how complex they get, in the same way that even though microprocessors today are rapidly becoming faster and denser, but without a growing suspicion that they are becoming more likely to exhibit truly autonomous, non-bug related motivations. These future droids will be as likely to feel contempt for our human brains as a microprocessor does today.

Malignant robots are prevalent in movies and such, but that it is only because it would be a boring movie without the robots getting out of line in some way, using their spontaneously generated ids in ways that in reality continue to be the sole domain of human being’s capabilities, but completely alien to droids, for a long, long time – even when those droids have giant multiples of human intelligence and are absolutely indistinguishable from humans in appearance and behavior.

The fact that our droids won't turn on us doesn't mean this is a utopian future, however. There will be real challenges, but they're not what you think they are. These may be as serious as some of the traditional fears, but of an entirely different nature. They will come from within us, not the droids per se. Wherever we go, humans will lead the way, our will never overthrown by our creations. Those future challenges won't trigger a rush of adrenaline, but something in the pit of your stomach, somewhere between butterflies and a small rock there, as we realize we are heading into things that our current technologies barely hint at.

Further Reading:

Applications of artificial intelligence (Wikipedia)

Bayesian theory in New Scientist

Why Robots Won't Rule the World

AI and Scare Tactics, a Tale of Two Species

Ray Kurzweil On IT And The Future of Technology

Searching for Intelligence in Edinburgh
Artificial Intelligence celebrates new advances (Aug 10, 2005)

The Id, Ego, and Superego

Sigmund Freud, Id, Ego, Superego, Biography

Dictionary: superego

Origins of Values