Skip to main content

Ethical Considerations in Autonomous System Design

Technologies that operate without human intervention have been emerging spontaneously in diverse applications, from commercial transportation such as driverless cars, to military vehicles like remote-controlled tanks. Support is affluent from eager investors, who are perhaps looking forward to a similar market effect the IoT had. While an autonomous system [abbreviated AS for the rest of this article] has its perks of relegating tedious manual tasks towards a more comfortable lifestyle for consumers, they attract a lot of worry due to the freshness of the concept and the lack of a firm foundation in guaranteeing security. The risk of an unanticipated catastrophe is very real. There are still too few literatures that address questions on safety. Furthermore, fewer of those literatures that do address our qualms fall short of a satisfactory answer. In fact, Mr. Frenzel shares his refutations on autonomous vehicles in his blog posts. (Read “Forget this self-driving car nonsense” and “Just say no to driverless cars”). Personally, I also have a similar distaste for autonomous vehicles, because I do not find machine learning (or at least, the current state of the algorithms) sufficient for unmanned driving. I do not condemn it altogether though. Perhaps a more accurate GPS technology working congruently with a more effective learning algorithm will boost my confidence.

Aside from immediate consequences, such innovations are potently deleterious in the long run, necessitating regulation and proper training to involved designers and engineers. As the overused saying goes – an ounce of prevention is better than a pound of cure. Integration of preemptive measures with an AS against plausible long-term threats can save a company from lawsuits or bankruptcy and a consumer his/her life. After all, human well-being is always the top priority no matter what the case may be. [Have you heard of Aristotle’s Eudaimonia?] 

Let us now discuss some significant issues and considerations on designing an AS.

The Autonomous Intelligent System vs. Human Beings

The interaction between the AS and the human being is unique with respect to the situation. But how do we know that human rights are not infringed given various degrees of social and cultural norms? Obviously, it won’t be realistic to specify a universal set of constraints for everyone to follow. Rules tailoring to where and to what purpose they will be deployed for will have to be laid out. A clear delineation is a must to avoid compromising situations. The A.I. of an industrial robot that deals with equipment in a production line, for example, must differ greatly from the A.I. of a robot in healthcare, where a stringent increase in environment sensitivity is required.

There are many existing documents that define the rights of an individual, the “Universal Declaration of Human Rights” being the most well-known. Cultural diversity makes this definition particular to location. Because of a multiplicity of norms, the odds that conflict between values is not remote, thus an AS can have algorithmic biases that are disadvantageous to a particular group no matter what. In my opinion, there is an omnipotent element in defining the human right, regardless of geography or culture, and that is the safeguard against physical harm of any form. Morally officious matters, which are still unacceptable, can be dealt with solutions after discovery of the problem. Arising tensions can be pacified when both sides approach with tolerance and maturity. Unfortunately, physical harm, or at worst death, is irreversible. It can lead to unprecedented chain reactions of violence, hatred and anger. Remember what the death of an Austrian, on the fateful day of June 28, 1914, did to the world?

The Need for Methodologies to Guide the Design Process

When a robotics engineer is asked on how he designed his robot to satisfy an adequate level of trust between the people it will interact with, he/she can’t just respond with: “Oh! I just kept in mind Isaac Asimov’s three laws of robotics. I’m sure it won’t even hurt a fly. Sha-lala-lala…”. When all definitions have been identified, the actual act of merging them into the design process remains convoluted. How do I make my robot comply with this culture’s taboos? Should I also program it with the same level of sensitivity it has with people when it interacts with cattle [they are considered sacred in Hinduism]? Will my robot offend anyone when it makes this gesture or if it is shaped this way? Again, the absence of an elucidated official guide to such pressing concerns will yield answers characterized with protracted variability.

Academic institutions seldom prioritize courses that discuss ethics in AS. Maybe such a fiddling topic does not necessitate in-depth study or discussion. The underlying arguments after all, seem to be postulated from common sense. Do we really need models for intercultural education to account for specific issues in an AS?

How common do you see a news article where an unmanned aerial vehicle [UAV] has struck a wrong target? Imagine those poor victims, whose lives were unreasonably cut short because of what? A measly glitch in the AS? How about accidents involving driverless cars? Is there a need to empower awareness on such loopholes in these AS systems so that prompt solutions can be provided by the tech community? Do we need better documentation practices for such events so that the next designers will not repeat such flaws? 

In such accounts, the AS is obviously accountable. But what about cases where accountability is obscure? There is also the challenge of creating a system that can properly identify when the AS is liable to fault, so that an effective solution can be implemented.

The AS designer is also exposed to the risk of self-bias. An ostensible sense of security can be reached even though an imminent peril exists. A third party responsible for the AS’s value alignment can avert such a danger.

Finally, an AS system is bound to evolve as technology moves forward. In effect, both ethical and safety issues will change, and it is incumbent upon the manufacturers of the AS to adjust appropriately.

I usually end my articles with an interrogative sentence to provoke after-thought, but I’m afraid I’ve given enough above (In fact, my college professors once complained that I ask too many questions). So instead, I end with a declarative sentence. A resolution that any autonomous project I embark in in the future will undergo the thoughts and insights mentioned in this article. 


Popular posts from this blog

Calculator Techniques for the Casio FX-991ES and FX-991EX Unraveled

In solving engineering problems, one may not have the luxury of time. Most situations demand immediate results. The price of falling behind schedule is costly and demeaning to one's reputation. Therefore, every bit of precaution must be taken to expedite calculations. The following introduces methods to tackle these problems speedily using a Casio calculator FX-991ES and FX-991EX.

►For algebraic problems where you need to find the exact value of a dependent or independent variable, just use the CALC or [ES] Mode 5 functions or [EX] MENU A functions.

►For definite differentiation and integration problems, simply use the d/dx and integral operators in the COMP mode.

►For models that follow the differential equation: dP/dx=kt and models that follow a geometric function(i.e. A*B^x).

-Simply go to Mode 3 (STAT) (5)      e^x
-For geometric functions Mode 3 (STAT) 6 A*B^x
-(Why? Because the solution to the D.E. dP/dx=kt is an exponential function e^x.
When we know the boundary con…

Common Difficulties and Mishaps in 6.004 Computation Structures (by MITx)

May 6, 2018
VLSI Project: The Beta Layout [help needed]Current Tasks: ►Complete 32-bit ALU layout [unpipelined] in a 3-metal-layer C5 process. ►Extend Excel VBA macro to generate code for sequential instructions (machine language to actual electrical signals).
Current Obstacles/Unresolved Decisions:
►Use of complementary CMOS or pass transistor logic (do both? time expensive, will depend on sched.
►Adder selection: Brent-Kung; Kogge Stone; Ladner Fischer (brent takes up most space but seems to be fastest, consider fan-out) [do all? time expensive, will depend on sched.)
►layout requirements and DRC errors

Please leave a comment on the post below for advise. Any help is highly appreciated.

Yay or Nay? A Closer Look at AnDapt’s PMIC On-Demand Technology

Innovations on making product features customizable are recently gaining popularity. Take Andapt for example, a fabless start-up that unveiled its Multi-Rail Power Platform technology for On-Demand PMIC applications a few months back. (read all about it here: Will PMIC On-Demand Replace Catalog Power Devices?) Their online platform, WebAmp, enables the consumer to configure the PMIC based on desired specifications. Fortunately, I got a hands-on experience during the trial period (without the physical board (AmP8DB1) or adaptor (AmpLink)). In my opinion, their GUI is friendly but it lacks a verification method for tuning (i.e. the entered combination of specs). How would we know if it will perform as expected or if there are contradicting indications that yield queer behavior? Also, there is not just one IP available, but many that cater to a differing number of channels and voltage requirements (each with their own price tag).
Every new emerging technology has the potential to oversh…