Category: Ethics

Horty’s Defaults with Priorities for Artificial Moral Agents

Can we build robots that act morally? Wallach and Allen’s book titled Moral Machines investigates a variety of approaches for creating artificial moral agents (AMAs) capable of making appropriate ethical decisions, one of which I find somewhat interesting. On page 34, they briefly mention “deontic logic” which is a version of modal logic that uses concepts of obligation and permission rather than necessity and possibility. This formal system is able derive conclusions about how one ought to act given certain conditions; for example, if one is permitted to perform some act α, then it follows that they are under no obligation to not do, or avoid doing, that actα. Problems arise, however, when agents are faced with conflicting obligations (McNamara). For example, if Bill sets an appointment for noon he is obligated to arrive at the appropriate time, and if Bill’s child were to suddenly have a medical emergency ten minutes prior, in that moment he would be faced with conflicting obligations. Though the right course of action may be fairly obvious in this case, the problem itself still requires some consideration before a decision can be reached. One way to approach this dilemma is to create a framework which is capable of overriding specific commitments if warranted by the situation. As such, John Horty’s Defaults with Priorities may be useful for developing AMAs as it enables the agent to adjust its behaviour based on contextual information.

Roughly speaking, a default rule can be considered similarly to a logical implication, where some antecedent A leads to some consequent B if A obtains. A fairly straightforward example of a default rule may suggest that if a robot detects an obstacle, it must retreat and reorient itself in an effort to avoid the obstruction. There may be cases, however, where this action is not ideal, suggesting the robot needs a way to dynamically switch behaviours based on the type of object it runs into. Horty’s approach suggests that by adding a defeating set with conflicting rules, the default implication can be essentially cancelled out and new conclusions can be derived about a scenario S (373). Unfortunately, the example Horty uses to demonstrate this move stipulates that Tweety bird a penguin, and it seems the reason for this is merely to show how adding rules leads to the nullification of the default implication. I will attempt to capture the essence of Horty’s awkward example by replacing ‘Tweety’ with ‘Pingu’ as it saves the reader cognitive energy. Let’s suppose then that we can program a robot to conclude that, by default, birds fly (B→F). If the robot also knew that penguins are birds which do not fly (P→B ˄ P→¬F), it would be able to determine that Pingu is a bird that does not fly based on the defeating set. According to Horty, this process can be considered similarly to acts of justification where individuals provide reasons for their beliefs, an idea I thought would be pertinent for AMAs. Operational constraints aside, systems could provide log files detailing the rules and information used when making decisions surrounding some situation S. Moreover, rather than hard-coding rules and information, machine learning may be able to provide the algorithm with the inferences it needs to respond appropriately to environmental stimuli. Using simpler examples than categories of birds and their attributes, it seems feasible that we could test this approach to determine whether it may be useful for building AMAs one day.

Now, I have a feeling that this approach is neither special nor unique within logic and computer science more generally, but for some reason the thought of framing robot actions from the perspective of deontic logic seems like it might be useful somehow. Maybe it’s due to the way deontic terminology is applied to modal logic, acting like an interface between moral theory and computer code. I just found the connection to be thought-provoking, and after reading Horty’s paper, began wondering whether approaches like these may be useful for developing systems that are capable of justifying their actions by listing the reasons used within the decision-making process.

Works Cited

Horty, John. “Defaults with priorities.” Journal of Philosophical Logic 36.4 (2007): 367-413.

McNamara, Paul, “Deontic Logic”, The Stanford Encyclopedia of Philosophy (Summer 2019 Edition), Edward N. Zalta (ed.), URL = <https://plato.stanford.edu/archives/sum2019/entries/logic-deontic/>.

Epistemic Responsibility Today

Section 6 of Miller and Record’s Justified Belief in a Digital Age provides suggestions for responsible belief formation given the role and influence algorithms possess in today’s society. The notions they present, however, are vague and appear to be shortsighted. They suggest “subjects can use existing competencies for gaining information from traditional media such as newspapers to supplement internet-filtered information and therefore at least partly satisfy the responsibility to determine whether it is biased or incomplete” (130), except the nature of ‘traditional media’ (TM) has shifted. Since the widespread adoption of social media platforms and online news streaming, TM has seen an increase in competition as small and independent news websites are also shared between users. Importantly, expectations for endless novel content has pressured TM to keep up by increasingly producing editorials, commentary, and speculation. Pundits receive as much airtime as journalists due to the nature of consumer demand, subsequently influencing belief formation. The notion of political bias in TM is also a large concern, where journalistic integrity and credibility ranges drastically between companies. Additionally, TM is more likely to be subsumed under an umbrella corporation with an agenda of its own, whether political, financial, or religious. Deference to TM has always been associated with epistemic risks, and reasons to be sceptical of stories and information are growing as technology modifies our consumption habits.

Further down on page 130, it is recommended one explore outside their personalized feed by investigating others’ posting history: “Instead he can casually visit their Facebook profiles and see whether they have posted an interesting story that the automatically generated news feed missed.”. While this does improve chances of being exposed to diverse content, it is most effective when one reads the feeds of contrasting personalities. Close friends and family members may hold similar attitudes, values, or perspectives which do not adequately challenge one’s suspicions or beliefs. Opposing views, however, may not be justified or well-formed, and ‘opposing’ is open for interpretation. On page 131 the authors state: “… suggests, internet sites, such as political blogs, may refer their readers to alternative views, for example, by linking to opposing sites, out of a commitment to pluralism.” If this program were to be followed, it would suggest religious individuals with dogmatic beliefs are epistemically irresponsible. This may be an unexciting verdict to a philosopher, but it is difficult to determine whether this normative approach to belief formation is suitable for all humans.

Epistemic justification is complicated in the digital age, and it is unclear how much research is required to fulfill one’s epistemic responsibilities. If one stumbles across a scientific claim, it seems reasonable that one ought to determine whether the news headline matches the outcome of the study. Considering the replication crisis has further complicated this process, how much scientific scrutiny is required at this point? If a reader has an understanding of scientific methodology and access to the article, is it irresponsible if one does not examine the methods section? As ideal as epistemic responsibility seems, it might be unattainable due to the nature of the internet and human emotion. Our ability to access such a wealth of knowledge, even when curtailed by algorithms, generates an infinite regress of duties and uncertainty, a fact unlikely to sit well with the average voter.

Works Cited

Miller, Boaz, and Isaac Record. “Justified belief in a digital age: On the epistemic implications of secret Internet technologies.” Episteme 10.2 (2013): 117-134.

AI and the Responsibility Gap

This week we are talking about the responsibility gap that arises from deep learning systems. We read Mind the gap: responsible robotics and the problem of responsibility by David Gunkel along with Andreas Matthias’ article The responsibility gap: Ascribing responsibility for the actions of learning automata.

It seems the mixture of excitement and fear surrounding the rise of autonomous agents may be the result of challenges to our intuitions on the distinction between objects and subjects. This new philosophical realm can be analyzed from a theoretical level, involving ontological and epistemological questions, but these issues can also be examined through a practical lens as well. Considering there may be a substantial amount of debate on the ontological status of various robots and AIs, it might be helpful to consider issues on morality and responsibility as separate to the theoretical questions, at least for now. The reason for this differentiation is to remain focused on protecting users and consumers as new applications of deep learning continue to modify our ontological foundations and daily life. Although legislative details will depend on the answers to theoretical questions to some degree, there may be existing approaches to determining responsibility that can be altered and adopted. Just as research and development firms are responsible for the outcomes of their products and testing procedures (Gunkel 12), AI companies too will likely shoulder the responsibility for unintended and unpredictable side-effects of their endeavours. The degree to which the organization can accurately determine the responsible individual(s) or components will be less straightforward than it may have been historically, but this is due to the complexity of the tools we are currently developing. We are no longer mere labourers using tools for improved efficiency (Gunkel 2); humans are generating technologies which are on the verge of possessing capacities for subjectivity. Even today, the relationship between a DCNN and its creators seems to have more in common with a child-parent relationship than an object-subject relationship. This implies companies are responsible for their products even when they misbehave, as the debacle surrounding Tay.ai demonstrates (Gunkel 5). It won’t be long, however, before we outgrow these concepts and our laws and regulations are challenged yet again. In spite of this, it is not in our best interest to wait until theoretical questions are answered before drafting policies aimed at protecting the public.

Works Cited

Gunkel, David J. “Mind the gap: responsible robotics and the problem of responsibility.” Ethics and Information Technology (2017): 1-14.

Addiction by Design: Candy Crush et al.

For class this week, we read the first four chapters of Natasha Schull’s book Addition by Design. I think the goal was to consider the similarities and differences between slot machines and gaming applications on handheld devices.

While the two addictions are comparable despite their differences in gameplay format, apps like Candy Crush have found profitable solutions to their unique problems. Developers expect players to “leave their seats” as cellphone use generally orbits around other aspects of daily life. While “time on device” (58) is surely an important part of app design, creating incentives for users to return are also significant. Though this may be accomplished in a number of ways, a common strategy is to generate frequent notifications to both remind and seduce users back to their flow state (49). Overall, the approach may seem less inviting than sounds and lights but its ability to display explicit directions may be effective. Text has the ability to specify rewards if the user opens the app right then and there. A pay structure involving varying wait times may also push users to pay for the ability to return to “the zone” (2). This may take the form of watching an advertisement or being disallowed to play for intervals from an hour to a day, sufficiently frustrating users to pay to continue playing. Similarly to embedding ATMs in slot machines (72), app stores with saved credit card information allow developers to seamlessly lead users to the ‘purchase’ button, quickly increasing revenue. Financial transactions thinly disguised as a part of the game offer a new way to siphon money from vulnerable individuals, especially parents of children with access to connected devices. Additionally, gaming apps are typically weakly associated with physical money like bills and coins, unlike slot machines from mid 20th century (62), perhaps making it easier for consumers to pay without drawing their attention to the movement of money. This brief analysis suggests the nature of gambling is evolving by modifying existing modes of persuasion and adapting to new technological environments.

One large concern, however, arises from where this money goes; while governmental agencies oversee regulations (91) and collect revenue (5) to fund programs and projects, private companies simply collect capital. This carries severe implications for individuals, communities and economies as this alternative stream of income dries up. Therefore, it could be suggested that state and provincial legislators should consider addressing this issue sooner than later.

Works Cited

Schüll, Natasha Dow. Addiction by design: Machine gambling in Las Vegas. Princeton University Press, 2014.

Algorithmic Transparency and Social Power

This term I’m taking the course Science and Ethics, and this week we read Langdon Winner’s 1980 article “Do Artifacts have Politics?” along with a paper from 2016 published by Brent Daniel Mittelstadt and colleagues titled “The ethics of algorithms: Mapping the debate.” We are encouraged to do weekly responses, and considering the concerning nature of what these articles are discussing, thought it should be presented here. There is definitely a lot that could be expanded upon, which I might consider doing at a later time.

Overall, the two articles suggested risks of discriminatory outcomes are an aspect of technological advancements, especially when power imbalances are present or inherent. The paper The ethics of algorithms: Mapping the debate focused particularly on algorithmic design and its current lack of transparency (Mittelstadt 6). The authors mention how this is an epistemic concern, as developers are unable to determine how a decision is reached, which leads to normative problems. Algorithmic outcomes potentially generate discriminatory practices which may generalize and treat groups of people erroneously (Mittelstadt 5). Thus, given the elusive epistemic nature of current algorithmic design, individuals throughout the entire organization can truthfully claim ignorance of their own business practices. Some may take advantage of this fact. Today, corporations that manage to successfully integrate their software into the daily life of many millions of users have little incentive to change, due to shareholder desires for financial growth. Until the system which implicitly suggests companies can simply pay a fee, in the form of legal settlements outside of court, to act unethically, this problem is likely to continue to manifest. This indeed does not inspire confidence for the future of AI as we hand over our personal information to companies and governments (Mittelstadt 6).

Langdon Winner’s on whether artifacts have politics provides a compelling argument for the inherently political nature of our technological objects. While this paper may have been published in 1980, its wisdom and relevance can be readily applied to contemporary contexts. Internet memes even pick up on this parallel; one example poses as a message from Microsoft stating those who program open-source software are communists. While roles of leadership are required for many projects or organizations (Winner 130), inherently political technologies have the hierarchy of social functioning as part of their conceptual foundations, according to Winner (133). The point the author aims to stress surrounds technological effects which impede social functioning (Winner 131), a direction we have yet to move away from considering the events leading up to and following the 2016 American presidential election. If we don’t strive for better epistemic and normative transparency, we will be met with authoritarian outcomes. As neural networks continue to creep into various sectors of society, like law, healthcare, and education, ensuring the protection of individual rights remains at risk.

Works Cited

Mittelstadt, Brent Daniel, et al. “The ethics of algorithms: Mapping the debate.” Big Data & Society 3.2 (2016): 1-21.

Winner, Langdon. “Do artifacts have politics?.” Daedalus 109.1 (1980): 121-36.