Hey Siri, Who's Your Daddy?
“Don’t be evil.” Googlers generally apply those words to how we serve our users. But “Don’t be evil” is much more than that. Yes, it’s about providing our users unbiased access to information, focusing on their needs and giving them the best products and services that we can. But it’s also about doing the right thing more generally – following the law, acting honorably, and treating co-workers with courtesy and respect.
— From Google’s Code of Conduct (withdrawn in 2018)
* * * * *
Geoffrey Hinton recently resigned from his job at Google, so that he could speak freely. Called by some “the Godfather of AI,” Dr. Hinton’s work began in the 1970’s as a graduate student at the University of Edinburgh. There, he began work on what is now known as “neural networks.”
A neural network is a mathematical system which can learn by analyzing data. For example, in 2012, at the University of Toronto, Dr. Hinton built a neural network which could analyze photographs and teach itself to identify common objects, such as human faces and bodies. In that same era, other researchers built similar networks based on digital text.
Today, these systems use much larger amounts of data, and in some ways, are becoming more intelligent than humans. Dr. Hinton cautions that the tremendous progress of the past five years portends a great deal of risk over the next few years.
In leaving Google, Dr. Hinton stated that, “It is hard to see how you can prevent bad actors from using [AI] for bad things.” AI can already be used to flood the internet with misinformation, such as fake photos and videos, so that it is already possible to create an environment in which the average person is “not ... able to know what is true.”
The larger concern is that eventually these neural networks will generate and also run their own computer code, and become, for example, autonomous weapons, as fictionalized in “The Terminator.”
* * * * *
IBM defines “Artificial Intelligence” as “The science and engineering of making intelligent machines, especially intelligent computer programs. It is related to the similar task of using computers to understand human intelligence, but Artificial Intelligence does not have to confine itself to methods that are biologically observable.”
The seminal work in the field is Alan Turing’s paper entitled, “Computing Machinery and Intelligence.” (Mind 49: 433-460 (1950).) This paper documents a study on the question, “Can machines think?” The test was whether a human interrogator could distinguish between a human and a computer text response to various questions.
AI now has four potential goals:
systems that think like humans,
systems that act like humans,
systems that think rationally, and
systems that act rationally.
Further, there are two types of Artificial Intelligence:
weak, and
strong.
Weak AI, which is also called “Narrow AI,” is trained to perform specific tasks. It is not weak in the sense of being the opposite of “robust,” it is weak only in the sense that it has a narrow focus. This is the type of AI which is already almost ubiquitous, for example, in applications from SIRI and Alexa to speech recognition, customer service, and automated stock trading to self-driving cars.
Strong AI is still theoretical. This would comprise Artificial General Intelligence and Artificial Super Intelligence. If this form of AI comes to fruition, it will have a self-aware consciousness, and it will have the ability to solve problems, to learn, and to plan for the future. The best example of this type of AI is HAL, the superhuman computer in “2001: A Space Odyssey.”
* * * * *
The ethics of AI research is informed by the ethics of biomedical research, and in particular, the National Research Act. This act is intended to provide guidelines for research on human subjects and to protect humans from unethical research. It was enacted in 1974 in response to the outrage over the Tuskegee syphilis experiments, and built upon earlier models, including the Nuremberg Code, the Declaration of Helsinki, and the Belmont Report.
AI researchers are expected to have respect for other people, just as the medical industry is concerned with the informed consent of people who are involved in experiments. Similarly, they are expected to be motivated by beneficence, just as doctors pledge an oath to “first, do no harm.”
Some of the ethical implications of AI research are that:
AI may replace human intellect in scientific creativity,
AI likely will replace humans in many different jobs from routine and mundane tasks to more complicated matters, such as medical diagnosis and legal briefing,
AI will further erode privacy, and
in making rational decisions, AI may discriminate in unintended, socially-condemned ways.
A further problem is that at present, there is no real accountability in AI research, so that there are no particularized consequences for causing harm – other than proscriptions and remedies in the general law.
* * * * *
I have wondered for a long time (a) whether capitalism is a system in the sense of an economic process designed and implemented, and then periodically or constantly refined to increase efficiencies, and to respond to changes in the markets and in government regulations, or (b) whether it is simply the name for the way people will take advantage of each other, if given the chance, subject to such constraints as financial limitations, morality, and the law.
The dictionary doesn’t help.
Capitalism is an economic and political system in which a country’s trade and industry are controlled by private owners for profit, rather than by the state.
— New Oxford American Dictionary, Third Edition
Nor does Adam Smith help much. His (correct) assumption is that people are naturally self-serving, and he argues that as long as everyone seeks the fulfillment of his own self-interest, the material needs of the whole society will be met. There is no need for government or any other outside force to interfere with or regulate the market(s).
Adam Smith was, however, opposed to monopolies. He argued that competition, which is guided by an “invisible hand” towards the general welfare, is the key to advancing the best interests of the society. Conjoined to this ideal of free market competition is the proposition that the division of labor will promote the growth of wealth and prosperity.
Frankly, all of this “classical” economics seems most like an idealism, if not a naivete, born of smoke-filled club room fantasy. Nowhere does Adam Smith account for the hugest elephant which has ever been in any room. That elephant is the inherent corruption of mankind; that people will, more often than not, lie, cheat, and steal, if given the slightest opening. Adam Smith does acknowledge that his capitalist ideal depends upon the virtues of prudence and justice, but these virtues are particularly difficult to enforce in the absence of government.
Ultimately, it has always struck me that capitalism is much more a system in its effect and adaptations, than it is a system by design, except, of course, that the thesis is based on the premise that private ownership of property is a good. Which, in general principle, it is. History has shown that a system of incentives is vital to human achievement and progress.
* * * * *
I mention this for two reasons. First, capitalism, in contrast to money, which is merely one mechanism for the transfer of capital, can be thought of as the OG of AI.
Capital has only one objective: the concentration of capital; the concentration of ownership of property, real and personal. Capital is a relentless force, and as far as I can tell, it is wholly amoral in its pursuit of its objective. Capitalism may have served as the model for a terminator:
It can’t be bargained with. It can’t be reasoned with. It doesn’t feel pity, or remorse, or fear, and it absolutely will not stop. Ever. Until you are dead.
— Kyle Reese
Further, capital has only one feature, aside from its amorality: the more concentrated it becomes, the more risk averse it becomes. The more concentrated it becomes, the more it will seek out rentier activities to generate profits with less risk. The more concentrated it becomes, the less incentive it will offer to the market for innovation, since offering such incentives involves more risk of loss.
Second, we only have to open our eyes to see the devastation which capital will leave in its wake if insufficiently managed. On the one hand, while capital itself is beyond good and evil, those in its service, for their personal benefit, perceived or genuine, will, and often do, commit acts and make omissions which can or will be fairly adjudged by the society as evil. In the end, fulfillment of self-interest, or perceived self-interest, can, and often will, hurt, rather than benefit, the society.
On the other hand, capital proves remarkably effective in meeting its objective, often with the aid of evil. As of 2016, the top ten percent of the people owned 78% of the wealth.
Most likely this share has increased over the past seven years. It was reported, for example, that the ten richest people in the United States doubled their wealth during the economic oppression imposed by the government purportedly in response to the SARS-CoV-2 virus, which may have been the real point of that fascist exercise.
Unfortunately, this kind of concentration is a recurring problem. The last time we had such dramatic income inequality was right before the depression in the 1930’s.
The bottom line is that the richest people own a great deal of wealth, but the country itself is in poor condition. We are lagging in manufacturing capacity and factory output, we are not training enough people in the trades, and we lack realism in the use of natural resources. In addition, the government is under corporate capture, and is making free trade agreements, enacting right to work laws, deregulating industry, providing preferential tax treatment of capital gains and corporate profits, and pursuing monetary policies which are facilitating the concentration of capital.
The fact is that all of this, and more, and worse, is capital doing exactly what capital does, as if by program or an embedded code. Moreover, the fact is that all of this is the exact opposite of what capitalism’s loudest proponents promise.
* * * * *
I’m not sure how meaningfully, or even whether, capitalism, a system by design or in effect, serves as a warning about the need for carefully thought-through restraints in the development and implementation of Artificial Intelligence. However, capitalism looks like a self-actualizing intelligence which operates independently of, or apart from, humanity, and we have history books filled with evidence that when the “animal spirits” of capitalism are not carefully managed, capitalism leads to economic instability and human misery.
The last thing we need is another such beast, no matter what its proponents promise.