Thursday, June 21, 2018

Shan Gao Two Essays: (1) "Comments on Human Rights and Artificial Intelligence: An Urgently Needed Agenda" and (2) "Business AND Human Rights: will it be a lasting happy marriage?"

(Pix © Larry Catá Backer 2018)

This post highlights two short essays produced by Shan Gao, a valued member of the research team at the Coalition for Peace and Ethics.  The first essay, Business AND Human Rights: will it be a lasting happy marriage?, looks generally at the context of human rights discourse. The second essay, "Comments on Human Rights and Artificial Intelligence: An Urgently Needed Agenda," considers some of the implications of Mathias Risse’s short essay on Human Rights and Artificial Intelligence: An Urgently Needed Agenda (HKS Faculty Research Working Paper Series; May 2018 RWP18-015) .

Comments on Human Rights and Artificial Intelligence: An Urgently Needed Agenda
Gao Shan

Mathias Risse’s short essay on Human Rights and Artificial Intelligence provides an excellent and comprehensive overview on the contemporary social anxiety toward the advancement of highly sophisticated and powerful technology development: the Artificial Intelligence and the implication to the human rights. The essay honestly reflects a sentiment from our society toward this new but less understood technology by the public. In echoing popular narratives developed by sci-fiction community between the 60s-80s, this essay revived a classic hypothetical tale about the conflict between humans and human like or superhuman like machines.

In particular, the essay unfolded interesting discussions focus on the consequences of this technology such as: machine consciousness, emotional attachment toward machine by humans, and domination of human by the machines, etc. Before a response to refute these discussions, this commentary will first consider what is the nature of AI. At the end of this response, it will provide a short opinion on what artificial intelligence means for human rights context.

What is the nature of AI?

It is worth to mention that, the author rightfully titled the essay by putting Human Rights before Artificial Intelligence—a detail I noticed when I first typed artificial intelligence ahead of human rights—which reminded us that, the problem we are exploring is not an artificial intelligence issue but a human rights issue.

We are living in an Artificial Intelligence hyper society. As Mathias Risse’s short essay illustrated, there is a sentiment of anxiety in the air fueling a narrative of unsettling future that AI developed super-human like intelligent and make human become obsolete. But what is artificial intelligence? What type of answer are we seeking?

A technological answer about this subject is less relevant because such answer reframe the question as an artificial intelligence issue rather than a human rights issue—although it is a necessary and important discussion in implementing beneficial and sustainable social and industrial development policy.

From starter, Artificial Intelligence is business. It is an entity of business developed as a profit generating vehicle to provide service for societal needs. Although precise relationship between the technology and the provider, manager, supervisor and patent owner within the context of human rights remained undefined for robust discussions, the conventional Business and Human Rights, and even civil law (directing into AI developed by and for non-commercial context) framework should be applicable for related discussion.

There is a sense of anthropomorphism from the society that treats artificial intelligent technology as unprecedented intelligent creature. It cultivates a belief that prospective entity would equipped with superhuman computational power competent for tasks require humanly-like intelligent. However, this might be wrong.

It is a false belief to consider all of this as unprecedented. Artificial intelligent long existed before the buzzword of AI, in the form of Multinational Corporation. The creation of limited corporate liability entity and later evolved global trade and more complex supply chain—designed to deliver a more efficient supply-demand social regime—is pre-AI artificial intelligence.

The MNC and intentional civil society is a form of ‘artificial intelligence’ without chips and semiconductors. These entities equipped with super-human computational power through its superior financial and intelligent resources. Engineering and managerial brainpower was invested in the process of developing sophisticated institutional governing mechanism and behavior pattern to ensure the success of global production by collecting and analyzing data, producing intelligence for decision-making. From institutional perspective, the same working mechanism applies to AI.

The comparison between MNC institution or global production network and artificial intelligence is uncanny: both entities direct superior computational power to optimize the process of collecting, analyzing information for decision-making. In addition, both entities use such power to define the standard of value and expectation of consumers. For certain task and under certain conditions, artificial intelligence can implement this process in a more flexible, economical manner, essentially, makes it more accessible and affordable for more people.

The mechanism of artificial intelligence in the process of social production is parallel to the MNC and international organizations, thus the much developed Business and Human Rights is a useful reference point for developing boundaries and limitations with the application of the technology.

Questions raised in the essay and why they might be mischaracterized?

It is easier to response to some of the issues raised by the essay after we consider artificial intelligence as transitional business entities. The short essay cited few examples about the thematic conflicts between the machine and humans, in general, it can be described as follows:
● What if AI provides better service than humans?
● Moral and value issue:What about developing consciousness for AI?
● Moral and value issue: How about humans developing emotional attachment to AI powered robot machine?
● Ethical issue: AI dominate over human?

Provides better service than humans?

(Albert Birkle,The Telegraph Operator, 1927)

The fear that AI provides better service than human and thus render human obsolete exposed some subtle psychological issue about the self. In fact, the fear essentially directed to a over-generalized question about self-value.

The author cites the example that AI might provide more “accurate” recommendation for which book to read or where to go for vacation. This hypothetical suffers from two flaws:

First, the judgement of accuracy lack the “subject.” The author may refer to such recommendation is more commercially accurate for the business owner. Or the author may refer to the recommendation is more accurate for consumer but not necessarily for business owner. In either case, the preference of AI over human should be considered as a progress of civilization rather than an issue. Second, even if the assumption that AI is more accurate in providing recommendation service, which I strongly disagree, that fact along cannot be a prediction for the obsolete of human service. It is worth to mention that online booking service did not replace the needs for travel agency. Another example would be the fact that advertising industry using AI generated background music to save production cost in shooting commercials does not diminish societal needs for good music. In this sense, technology providing powerful tools to make the work more efficient.


(Randolf Rogers, the Lost Pleiad, 1874)

Humanlike consciousness AI is eye catching, but also tends to be over-romanticized by the society. Asking the consciousness of AI is equally asking the consciousness of MNC. I do not see how the development of AI could possess moral and value issues about human like consciousness, unless we are creating a consciousness-based commercial market, maybe Ethical class thought by AI? In fact, I think this question is less exciting because it is similar to asking if your realtor has consciousness or your municipal government has consciousness or your university has consciousness. The way institution handling ethical, moral, cultural or political controversies are very different than an individual does. I doubt it is appropriate to adopt conventional analysis about consciousness on the AI. However, empirical study on business and human rights can provides some reference for this issue.

Emotional attachment?

(Minor Araki, Snow Monkeys at Play in Autumn and Winter, 1992)

I do not see any ethical or moral dilemma here, for the prospect that human develops emotional attachment toward AI powered machines. It is not unprecedented and it is very common for humans to develop emotional attachment toward an (life-form or non-life-form) object. The existed laws and cultural norms recognize and protect such human behavior towards objects, such as pets, toys, photographs, etc.

It is interesting to mention the example cited by the author about robot nanny because the hypothetical scenario of developing emotional attachment toward machine may against our own life experience: technology industry tends to automatically obsolete machines for newer products, we can all relate to this issue, who is using a cellphone bought 10 years ago and has any cell phone accessory last more than 3 years ? In addition, how could human develop emotional attachment to a product or service when such product or service is developed by a business model requires frequent and disruptive updates? Emotional attachment is harder to attach when the lifespan of machine is less than a fish.

Domination over human?

(Philip Guston, Halloween Party, 1942)

The domination over human argument is interesting but paradoxical. There is a fear that machine surpasses humans in intelligence. However, such fear might be unnecessary.

The key here is to ask what is the purpose of intelligence? If the purpose of intelligence is to serve human needs, then such fear is unfounded because the complexity of human needs would always develop new demanding that require more intelligent to supply the demand. For example, our cellphone has more computational power than 3 years ago, but we still develop new functions to meet our needs. The assumption that AI surpass human intelligence in serving human needs in paradoxical because human would develop new needs making such intelligence sounds not intelligent at all. In fact, even the author realized this issue by citing the example of AI powered bot and artificial stupidity.

The real fear about AI surpassing human intelligence can be found in cases if such intelligence is not used to serve the needs of human but the AI itself. Or in other words, AI compete with humans for natural resources without any reliance on humans. That could be a real threat to the survivor of the human being that requires special attention.

The author also made following assertions “we must try to make sure their(AI) values are aligned with ours to render as unlikely as possible any complications from the fact that a super intelligence might have value commitments very different from ours”

The above statement projected a sincerely felt societal fear that  control of our life that being threatened by the AI. However, this fear might be unnecessary because the control one fears to lose may not actually exit as posited.

The fear of losing control is well rooted in our mental attitude toward the rise of transnational frameworks, values, and orders. Artificial intelligence is a derivative of transnational institution, developed by universal programming language to serve local and specialized needs. This institution, like MNCs and international organizations, is maintained, operated by a group of professionals sharing same sets of skills and training, to tailor the power of the institution for very local demands and needs.

Essentially, the fear of AI surpass humans is the same as the fear of transnational institution intrudes local society and local orders with decentralized values and norms. Thus, a question about the AI vs humans is a question about the boundaries and limitations of globalization in communicating with local community.

What it means for artificial intelligence under human rights?

The artificial intelligence under human rights context might be an intimacy issue—as a social norm, behavior or maybe ideologically, to what extent and on what level we human being wish to be connected with each other, economically, culturally, physically, and emotionally though the technology of artificial intelligence? Will sensible boundaries and limitations be developed during the transition to a more algorithm-based future when the human beings are more connected than ever for the past 300 years? Or, as I argued above, this is an issue about globalization vs. local community not man vs. machine.

Mathias Rise’s short essay identified certain well known fear from the public toward new technology, which provides an excellent avenue to ask questions and collect answers to explore the truth through facts. This process is not artificial, but it is truly intelligent and humane.

Business AND Human Rights: will it be a lasting happy marriage?
GAO Shan

The conception of “Business and Human Rights” is fascinating; it is a marriage of two concepts: “business” and “human rights.” Is there a term that can capture the essence of these two separate words without unstylish conjunction “and”? The conjunction “and” implying a form of relationship between two separated entities with independent meanings. In other words, our naming suggested an established social structure treats “business” and “Human rights” as an exclusive entity with different references and functions. Like a marriage, the union of these two entities reinforces the notion of a new structure. Thus, what does this new concept mean?

A mechanic approach to the meaning of Business and Human Rights can read it as follows:
A. It is more of business, business is the dominant one in this union.
B. It is more of Human Rights, Human Rights is the dominant one in this union.
C. Neither, it is a new idea that requires further exploration and study.

More Business Approach:
Business and Human Rights is a conception embodies the essence and principle of Human Rights with a belief that Human Rights is part of an obligation for business establishments to benefit from a healthy and sustainable growth and profit scheme. It is an fashionable definition that is well recognized from most MNC webpage. However, critics of this approach, such as the second approach described below, consider the above definition as insincere because it is a carefully designed and tailored information campaign produced by PR firms, to cater growing social anxiety toward injustice and inequality.

Hospitality toward this approach suggested a sentiment from the critics that because business approach is motivated by defending instructional interest, it tends to use human rights to cosmetically fixing public image crisis resulted from questionable business practice. In other words, business approach suffers from critics questioning their movie action and intention in promoting the human rights agenda.
More Human Rights Approach:
Human rights approach reasserted the supremacy of human rights in their discourse. It considers that human rights cannot be an ostentatious factor in producing business judgement, but a determining one that veto the vice and harm done by the traditional business practice.

Human rights approach suffers from the criticism for being hypocritical and naïve. It is hypocrisy because some human rights groups behave similar as business establishments under same financial or legal pressure. Some groups are just as budget conscious and attention hungry as big business institutions. It is not unusual that some groups would deploy mass media to manufacture social anxiety against the business world. It is naïve because some groups acted as idealist and single minded without a function framework to resolve highly complex and nuanced issues with a historical, political and cultural context. It has became clear that when their immature approach on these issues produce damages and harms, these groups would not hesitate to claim legal immunity.

A third approach?
Then what is Business and Human Rights? The first approach emphasizes more on the business component, tailors human rights for a more conscionable business development model. It grounded as the code of conduct or professional ethics that is being developed and applied by the business for the business. Human rights is a factor in formulating legally and socially acceptable business judgement. The pragmatism of this practice subjects to criticisms directing toward the motivation and intention.

The second approach highlights more on the human rights component. It denounces the pragmatism of efficiency regime and questioning the legitimacy of cost-benefit analysis under the first framework. However, this approach may sufferer from the same type of hypocrisy and corruption when it cannot escape from the larger historical and social structure that requires an obedience toward financial and legal constraints. In addition, the second approach may overlook the nuance and context of the issue in formulating their attacks to the first approach without offering practical alternatives.

Then what all of this leave us to the conception of Business and Human Rights? Will the marriage between pragmatic business mind and idealist Human Rights be a realist and sustainable framework to address the conflicts between the reality and idealistic?

The first issue for Business and Human rights is to identify realist principle to establish boundaries and limitations of each approach. Business and Human Rights framework can develop solutions to minimize the risks and limitations of each approach for a better integration of the two under the one union. A tendency of dominance by one side is normal in a union, however, an actual one-sided dominance may not be healthy and sustainable.

No comments: