Capital Punishment: A foolish response to serious crime

[Reproduced from my article in The Island, published on 2nd July: http://www.island.lk/index.php?page_cat=article-details&page=article-details&code_title=206811]

After a few false starts, the incumbent President of Sri Lanka seems to have acted upon his well-known belief in the utility of Capital Punishment [1]. This institutionalised crime is now due to take place in our land of ahimsa, after a hiatus of 43 years. Whether The President sprang into action because of the purported rise in drug trafficking, as he publicly espouses, or whether it was an attempt at gaining ignoble popularity prior the upcoming elections (Capital Punishment is quite a popular “solution” to serious crime in some ignorant quarters of Sri Lanka [2]), is not our concern here. Our purpose here is to discuss the value of Capital Punishment as a social instrument for mitigating the spread of serious crime, and examine its moral underpinning.

There are three popular arguments in favor of Capital Punishment:

1. It allows society to punish a wrongdoer, thereby balancing “the celestial scales of justice”.

2. It serves as a deterrent for persons who may be contemplating violent or serious crimes.

3. It serves as a redress for the victims of violent or serious crimes.

A mere century and a half ago, judicial experts and the intellectual community at large were in favour of leveraging Capital Punishment. Even that great rationalist luminary of the 19th Century, John Stuart Mill, famously argued in parliament in favor of capital punishment, albeit for the most extreme of cases: “…when the attendant circumstances suggest no palliation of the guilt, no hope that the culprit may even yet not be unworthy to live among mankind, nothing to make it probable that the crime was an exception to his general character rather than a consequence of it, then I confess it appears to me that to deprive the criminal of the life of which he has proved himself to be unworthy, solemnly to blot him out from the fellowship of mankind and from the catalogue of the living is most appropriate.” [3]

However, times have changed since the days of JS Mill. We saw two entire new branches of science emerge, which has something definitive to say about the efficacy of Capital Punishment; namely psychology and sociology.

We know today that argument #2 (i.e. Capital Punishment is a deterrent) is empirically false [4] [5] [6] [7] [8], and we know that arguments #1 (“punishment”) & #3 (“redress”) are a mere window dressing of a primitive, Darwinian instinct that was useful in stone-age tribal societies. Contrary to this vengeful instinct, many a moral philosopher, both ancient [9] and modern [10], has rejected capital punishment, and the use of retribution as a solace for victims, as an uncivilized way of conducting human affairs.

Albert Camus, that outstanding French libertarian and writer, highlighted the concern pointedly:

“But what is capital punishment if not the most premeditated of murders, to which no criminal act, no matter how calculated, can be compared? If there were to be a real equivalence, the death penalty would have to be pronounced upon a criminal who had forewarned his victim of the very moment he would put him to a horrible death, and who, from that time on, had kept him confined at his own discretion for a period of months. It is not in private life that one meets such monsters”.

The utopian human being with perfect mannerisms and an unfailing character is an imaginary socio-psychological construct, a conceptual role-model for children. The uneasy truth is that human intent is fickle, governed by a nervous system whose structure and function is fraught with aberrations, which cannot be eliminated through nurture alone. A psychopathic personality, for example, could be the direct consequence of a poor endowment of mirror neurons, as a result of generic mutations that attenuate empathy from birth [11] [12] [13]. Those with such subtle birth “defects”, no matter how peaceable their childhood influences may have been, may be saddled with a fundamental inability to empathize with other living creatures. They may not even be able to empathize with themselves, in a self-reflective manner. Not being able to feel for someone (or even for one’s own self) makes it easy for one to cause injury or distress to others. When children with such genetic predispositions for low empathy experience trauma in early childhood, psychopathy and “Cluster-B” personality disorders emerge.

An Iowa Supreme Court Justice made this observation as far back as 1840:

“Crime indicates a diseased mind in the same manner that sickness and pain do a diseased body. And as in the one case we provide hospitals for the treatment of severe and contagious diseases, so in the other, prisons and asylums should be provided for similar reasons.”

If society ends up killing every such person who yields to their natural instinct (to strike, rape or unscrupulously exploit), rather than finding ways to curb or neutralize their behavior, we then get into a fascinatingly diabolical downward spiral. The more we kill those who lack empathy, in order to better the lives of those who have it, the more we lower the empathy of the empathetic. We know that a taste for judicial killing brutalizes society [14], as was the case in Victorian England, where public hanging made life cheap, and people more and more violent. We find such a brutal society today in Saudi Arabia, where domestic workers are abused [15], and where murder and sex crimes are rampant. The executioner hacks away to no avail.

That is precisely why more enlightened nations, including Sri Lanka, aspire to practice Restorative or Preventative Justice [16].

We should also not make any mistake on the legitimacy of the actual act; Capital Punishment is a premeditated violent crime committed by the state, according to modern jurisprudence. It is not an act of self-defence (as Camus and others have clearly pointed out). Perpetrators are often executed years after their bad acts were committed, and by that time their attitudes have changed dramatically for the better, or they at least have been neutralized as a damaging force. There are ample such cases widely publicized in the media [17].

There probably are a dozen other reasons [18] for permanently abolishing capital punishment and resorting to a lengthy prison sentence, ranging from the danger of punishing the innocent, to the cost of the entire procedure outweighing the cost of a life sentence. To quote Jeffrey A. Fagan, Professor of Law at Colombia Law School:

“As states across the country adopt reforms to reduce the pandemic of errors in capital punishment, we wonder whether such necessary and admirable efforts to avoid error and the horror of the execution of the innocent won’t—after many hundreds of millions of dollars of trying—burden the country with a death penalty that will be ineffective, unreasonably expensive, and politically corrosive to the broader search for justice.” There is one very special reason why Sri Lanka should not take this extreme measure. Sri Lanka is the Asian poster child for a country operating a genuinely restorative system of justice, supposedly drawing inspiration from the compassionate philosophy of Gautama Buddha [19]. Our restorative system of justice serves as a beacon for sociocultural progress, in comparison with our neighboring countries. It is disappointing to see our President succumb to a belief in Capital Punishment, rather than stand upright and explain to people the hard truth that we cannot win the war against drugs, crimes or deviance, through mere attrition.

Advocates of Capital Punishment (including our President) often hide behind the fact that the USA and China leverage it. Let us quote that preeminent American moral philosopher Sam Harris, who refutes the pride in this assertion:

“Especially in the United States, is a barbaric system of imprisonment—to say nothing of capital punishment—that should make all citizens ashamed.”

Our own experts are unequivocal on the matter. Dr. L.B. L. De Alwis, ex-Chief JMO, had published an excellent analysis of the Lankan situation [20] in The Sri Lanka Journal of Forensic Medicine, Dec-2011, where he strikes at the core of the problem, along with a superb background analysis. Let us quote.

“In my opinion it is not the Non-implementation of the death penalty that has contributed to the rise of grave crime, especially murder, in Sri Lanka, but the release of murderers, rapists, drug barons, extortionists, highway robbers etc. sentenced to death or to long term rigorous imprisonment by the Judiciary, but later released by the executive in the shortest possible time for petty political advantage”.

Yasantha Kodagoda and other Lankan legal luminaries have held similar views [21] over the years.

It’s simply awful when drug barons rule the roost, or when terrible crimes happen. Our hearts go out to the victims and their kin, and our minds scream for justice. The state owes four things to society in instances of serious crime:

• Dispense justice swiftly and accurately, where the perpetrators are correctly identified, tried fairly and sentenced appropriately.

• Provide the next of kin of the victims with psychological counseling and support, to the utmost degree possible.

• Learn lessons from the crime and share them for the broader education of the general public.

• Firmly discourage lawlessness and mob-justice, which would interfere with criminal investigations.

The third point above is important and not to be underestimated in its value. Education, awareness and vigilance are the real weapons against serious crimes. Profiling of violent or deviant persons, cautioning parents and children about how to stay safe in ungated, low-income neighborhoods where dangers lurk, enforcing better policing etc. are all steps to be facilitated by the state. Furthermore, in counterpoint to The President’s current reasoning, a “war on drugs” is arguably a futile one [22]. There is ample expert opinion which shows that drug abuse is best mitigated through education, counseling, better parenting and social awareness. Of course we must proactively dismantle drug smuggling operations and imprison heavy traffickers for life. But laws alone are insufficient. We must “raise consciousnesses” against the use, manufacture and import of hard drugs, like we did against cigarets, and like we do for the accommodation of multiple sexualities. We must have a social awakening that makes the use of hard drugs unfashionable, and shun traffickers, not reward them (some of the largest importers were linked to Lankan politicians in the recent past [23]).

Poor parenting (and associated personality disorders in children) is not to be underestimated as a causal factor for the consumption of hard drugs. Lack of self-confidence in teenagers due to the indifference, over-control or violence of parents, is still fairly commonplace in Sri Lanka, even amongst upper middle income and wealthy homes. Such teenagers and young adults are far more susceptible to easy escape routs, like drugs, gangs or cults. We must also differentiate between hard drugs and marijuana. Experts tell us that the latter is less addictive and much less damaging to the body than alcohol or tobacco [24], and could reasonably be consumed in a safe form by psychologically balanced adults, to enjoy its soothing or psychedelic properties. Understanding this fact would help to focus the war, and not waste police resources and public money.

What the state does not owe society is a reactionary, “quick fix”, which would prejudice or pervert the broader course of justice in our country, and create an unhealthy punitive culture amongst our children. We leave the reader with these two quotes:

“I have never heard a murderer say they thought about the death penalty as a consequence of their actions prior to committing their crimes.” – Gregory Ruff, police lieutenant in Kansas

“I do not believe that any of the hundreds of executions I carried out in any way acted as a deterrent against murder.” – Albert Pierrepoint, Hangman, UK (1931-1956)

References:

1. President Sirisena on Capital Punishment (listen to the clapping as he expresses his personal approval): http://www.adaderana.lk/news/32381/capital-punishment-if-parliament-approves-president

2. Political motivation reported behind Sirisena’s move on reintroducing Capital Punishment: https://www.bbc.com/news/world-asia-48812576

3. Mill on Capital Punishment: http://ethics.sandiego.edu/books/Mill/Punishment/

4. Capital Punishment: Deterrent Effects & Capital Costs: https://www.law.columbia.edu/law_school/communications/reports/summer06/capitalpunish 5. 88% of criminologists do not believe the death penalty is an effective deterrent: http://www.deathpenaltyinfo.org/study-88-criminologists-do-not-believe-death-penalty-effective-deterrent

6. The Death Penalty and Deterrence: http://www.amnestyusa.org/our-work/issues/death-penalty/us-death-penalty-facts/the-death-penalty-and-deterrence

7. The death penalty is ineffective and indefensible: http://www.usnews.com/opinion/articles/2014/09/29/the-death-penalty-is-ineffective-and-indefensible

8. Failure to Deter Crime: http://nccadp.org/issues/deterrence/

9. Buddhism on Capital Punishment: http://www.bbc.co.uk/religion/religions/buddhism/buddhistethics/capitalpunishment.shtml

10. Sam Harris on capital punishment: “The result, especially in the United States, is a barbaric system of imprisonment—to say nothing of capital punishment—that should make all citizens ashamed”. http://www.samharris.org/blog/item/reflections-on-free-will

11. Psychopathic criminals have empathy switch: http://www.bbc.com/news/science-environment-23431793

12. Do Mirror Neurons Give Us Empathy? http://greatergood.berkeley.edu/article/item/do_mirror_neurons_give_empathy

13. How Empathy Can Be a Luxury: http://bengraves.hubpages.com/hub/Psychopaths-and-Mirror-Neurons-How-Empathy-can-be-Absent

14. Brutalizing society: http://www.bbc.co.uk/ethics/capitalpunishment/against_1.shtml

15. GCC declares war on domestic violence: http://www.arabnews.com/featured/news/657236

16. Restorative Justice: https://en.wikipedia.org/wiki/Restorative_justice

17. The case of Karla Faye Tucker: https://en.wikipedia.org/wiki/Karla_Faye_Tucker

18. 13 Reasons to Oppose the Death Penalty: http://www.oadp.org/facts/13-reasons

19. Buddhism and capital punishment: http://www.bbc.co.uk/religion/religions/buddhism/buddhistethics/capitalpunishment.shtml

20. DEATH PENALTY versus LIFE IMPRISONMENT: http://www.pdn.ac.lk/med/departments/forensic/Vol2No2.pdf

21. Death penalty not the answer, expediting dispensation of justice is: http://www.island.lk/index.php?page_cat=article-details&page=article-details&code_title=69674

22. The “War on Drugs” is lost: https://samharris.org/a-war-well-lost/

23. Political links to Drug Traffickers: http://www.ft.lk/article/255408/At-sea-over-drugs- 24. Alcohol vs Cannabis: https://www.livescience.com/61786-marijuana-versus-alcohol-brain.html

Advertisements

Opinion: masquerading behind the so-called “constitutional crisis” is an incompetent UNP leadership

There is a popular belief in Sri Lanka that is well supported by a kind of postmodern pseudo-intelligentsia, that “all politicians are rogues”. This fascinating meme – a viral idea that fits snugly in our consciences – is an intellectually vacuous concept. Yet time and time again it is craftily touted by people with vested interests, who often are on the loosing side of a major political struggle, and wish to detract support for the winners. Aside from the ostentatious and hypocritical nature of this presumption (especially when politicians themselves speak to it), there are good practical reasons for caring citizens to steer clear of this type of shallow, counterproductive analysis.

How on earth can we improve over time if we are unwilling to spot subtle differences in the choices that are made available to us? Sociologists tell us that progress is made through consistent, small wins. Like in nature, where tiny genetic mutations aggregate over many generations to produce entire new life forms, it is the seemingly small differences in the attitudes and skills of our representatives which ultimately amount to vast sociopolitical progress (or regress) over time.

The transition from a chaotic, war-torn country into a peaceful and more prosperous one is an excellent example from our own country’s recent past. This transition was led by an administrative team that was explicitly chosen by some of us; a team that succeeded where others had failed miserably for over three decades. They clearly had something better in them that suited the purpose, that some of us saw. So we absolutely must pay close attention and make definitive choices, or risk drifting into some clandestine political agenda that has little to do with mainstream interests like economic development.

In the present context of the so-called constitutional crisis, the better administrative team would be the one that has a clearer top-level agenda for responsible and purposeful government producing practical results, and not metaphysical rhetoric. Facilitation of economic growth and individual financial autonomy, skilled defense of an economy against global churn or downturn, liberalization of human values and improvements to the system of justice, prevention of terrorism, the enhancement of public services and utilities, and the facilitation of better lifelong education are obvious competency areas to watch out for amongst the two factions. Proof of even a marginal difference between the two factions is worthy of decisive support in favor of one side over the other.

The RW-led Yahapalanaya government performed abysmally in this regard, in comparison with the previous MR one. To put it plainly, they won on a deceitful ticket of dubious merit – the cry of “rogues, rogues”, a well-know political gambit that appeals to the downtrodden, who observe the rich and powerful strutting their stuff with envy, and pity themselves. They did little except to weaken government and bestow undue power on their otherwise apathetic leader, whilst allowing their cronies to embezzle over 10 billion rupees from the state coffers on the side.

The only serious charge that was substantiated through the Yahapalanaya government’s infamous “campaign against corruption”, was that of the reallocation of some state funds outside of financial regulations (FR) by the previous MR government, as a grant for prayer cloths for Buddhist devotees! This is after operating a special police taskforce for over three years to bring to justice those involved in supposed major financial crimes. Mind you, this “special” taskforce operated with brazen political bias under the direction of the Prime Minister, arresting or questioning all and sundry from the previous government on a daily basis, alas to no avail.

Let us come to the crux of the dilemma facing us today. Let’s be generous, and steel man the case for a so-called “constitutional crisis”. President MS, after working closely with, or rather attempting to work closely with the RW team for years, found himself to be increasingly irrelevant, and witnessing a rogue political agenda that was derailing Lanka’s economic progress. Worse, he found himself to be the target of a plausible assassination plot with high government connections, and made a quick decision to use his political clout to kick RW and his team out and restore some semblance of controlled, purposeful government. He consulted his legal advisors, and finding a loophole in the constitution that would serve him well in explanations later on, sent RW his dismissal note, and appointed his more capable former ally MR as the new Prime Minister.

Here is my key point. The same pundits who touted the “all are rogues” theory (like the JVP for instance) are screaming that due process is sacrosanct, and if process breaks down, all hell breaks loose. Who says? Why, if there weren’t revolutions in human society, we’d be stuck in a tribal, Neolithic world. Constitutions are drawn up (and amended) to uphold values and good practices as best understood at a given time in history. They however are ultimately just a means to an end, which is the overall wellbeing of the people at large. Means do not always supersede ends (just as ends do not always supersede means), especially if the means are preventing us from stopping a calamity like bloodshed or economic regression in this case.

We create due process to help us preserve human wellbeing based on existing knowledge, and when we discover a novel situation that needs urgent action outside of previous precedent, we first break the coded rules in the interest of time, and then amend them for future benefit. That’s why there have been hundreds of unconstitutional Executive Orders and Acts of Congress with sweeping consequences in American history, why the Australian Prime Minister was sacked unconstitutionally by the Governor General in 1975, and how Abraham Lincoln emancipated slaves.

President MS’s little constitutional coup is not such a remarkable action. So he exploited a loophole in the 19th Amendment to sack a grossly underperforming Prime Minister. The Supreme Court is the final authority to decide on the constitutionality of this action. Perhaps RW knows in his legal mind that MS was technically correct, since there doesn’t appear to be any move so far to clarify the matter with the Supreme Court. In any case MS did it to right a pretty bad situation. The rupee was in freefall, the Prime Minister was covering up the bond scam against a mountain of evidence, there appeared to be no purposeful moves to defend and strengthen the economy, taxation was rising with no corresponding increase in available public utilities or benefits for the disadvantaged.

In fact, benefits to the disadvantaged were being taxed, agriculture was neglected, infrastructure development was neglected, and there was evidence of a plot to murder The President. So all in all a good political move! Strongman-ish perhaps in nature but bloodless and easily democratized through parliament within the next few days. The President struck when the iron was hot, to the chagrin of his incompetent opponents who were trusting precedent and loyalty – two worthless values in the face of real problems.

A couple of other points for us to ponder on the present political situation. We now see yet another red herring being tossed up in the air, to distract us ordinary folks from the core issue of the failure of RW to perform sensibly. It is once again a version of the pitiful cry of “rogues, rogues”, this time taking the form of financial inducement for taking up ministerial posts. Listening to the first three audio recordings of MP Ranga Bandara’s phone conversations and his subsequent analysis of them, it is plainly apparent that the said Ranga Bandara is the one who is stitching three different conversations with three different people together, with his own unsubstantiated explanation of what is going on.

The first conversation sounds like a credible one, between himself and S.B. Dissanayake, a Government minister. In summary, the minister was urging him to cross over and join the the new government, before the 30 available cabinet positions are taken up by others. A perfectly reasonable and ethical conversation, that a minister from the new government would have with a UNP MP, to canvass support against RW, whose leadership they (the new government) consider as an active obstacle to the nation’s progress. The second and third conversations, which are suggestive of inducement, are between Ranga Bandara and two perfectly unknown persons, one of whom claims to be an agent of SB Dissanayake, and the other whom Ranga Bandara claims to be an agent of Yoshitha Rajapakse. Where is the evidence that SB Dissanayake offered money to Ranga Bandara, or that these two unknown people are in fact agents of the new government? Why, any pickpocket can be hired from the street to discuss a bribe over the phone, claiming to represent someone else.

I wouldn’t fret over this red herring, unless we can find evidence that clearly shows these two people acted on SB Dissanayake’s instructions. Transparency international has submitted this “evidence” to a court, lets see what the legal experts have to say.

The other more general point was that, for the umpteenth time, the RW camp is trying its level best to turn away our attention from administrative performance towards abstract morality. Getting the speaker to voice his personal displeasure over the “immorality” of RW’s sacking and the prorogation of parliament, the talk of bribes, the talk of dictatorships and unconstitutional government, prostration in front of foreign emissaries etc., are all part of a clever yet (unfortunately) regressive political campaign to gain sympathy and rekindle the nonperforming Yahapalanaya government. I urge all well meaning representatives and citizens to not get lost in these dubious details, but to stay focused on the big picture and act accordingly.

Was there not a gross failure in the administration of our country over the past three years, and didn’t The President make the right move to change the administrative leadership? Sift through the evidence and come to your own conclusions, ladies and gents.

Developing Principled Algorithms and the increasing necessary for “Ethics Testing” in Software Quality Assurance

If you were a software developer at Waymo, what would you program a car to do if a child sprang across its path from the pavement at the last second, while chasing a ball that rolled onto the street? Would you have the car turn sharply into the adjacent lane to definitively avoid the child, and risk killing the backseat passenger through a mid-body collision, or would you have it apply brakes and risk running over the child?

In the past, this sort of ethical dilemma used to be the topic of thought experiments designed for the amusement of moral philosophers. The “trolley problem” is the generalized thought experiment that highlights this type of moral conundrum. There was no compulsion for anyone to come up with a final preferred solution to the trolley problem, since the context was entirely theoretical. It used to be that there always was in practice a human being making an instant decision, and the judgment fell upon this human to exercise her so-called “free will” at the last second. In the case of the car and the child, the driver would do whatever that reflexively came to his mind at that moment. Not anymore.

With the advent of AI systems and their underlying decision-making algorithms, we have reached an era where we need to codify ethical judgment upfront into our software. And so we must begin to develop formal solutions to some tricky and sometimes bizarre ethical conundrums that were previously left to instant human judgment.

One nifty solution might be to trigger a random selection of one of the two options available, so as to superficially mimic what happens with human drivers, and yield an unpredictable result. You may sometimes favor the car, and sometimes the child. It is certainly an option that a software developer grounded in ethics might consider. As opposed to implementing the more myopic choice, of always prioritizing the life of the passenger.

The above scenario may be somewhat of a caricature, but it demonstrates why futurists and intellectual leaders have focused our attention towards developing principled algorithms, which incorporate broader ethical considerations when processing information and making important decisions that affect human wellbeing. With the advancement of AI, ethical conundrums raise their head in the most unexpected places, and unless we are trained to keep an explicit lookout for them as software engineers, we may unwittingly codify an ethically deficient or suboptimal solution.

Let us consider another example, of what appears at first glance to be a completely innocuous type of algorithm that we often encounter – personalized product recommendations – that can yet potentially cause seriously bad consequences if codified without care. Many of us have gotten used to considering what online stores like Amazon prompt us to buy next, based on their observations of our past curiosity and purchases. We are shown a variety of interesting choices by these stores, and unless we are searching for a specific item, we are happy to first browse these recommendations, and often to make a quick purchase. While this might be a great way to buy electronic gadgets, it can potentially be a serious handicap when it comes to broadening our knowledge through books.

Let us suppose that we had heard of a famous liberal intellectual like Noam Chomsky for the first time, and decided to search for one of his books in our favorite online store. Let us say we reviewed the book’s synopsis, and purchased it immediately. When we visit the store again, we are tempted with 5 new books, all having great looking titles and interesting outlines. Let us say we were to buy two more of them. More books would be recommended the next time, and we’d again buy some of them, and so on.

When we initially purchased the Chomsky book, the store would peg our reader interest under some kind of classification hierarchy. To keep things simple, let us say we were initially identified as liking books by liberal thinkers. So the first round of recommendations would be all books by authors who are broadly classified as being of a liberal mindset. As we bought more books through store recommendations, let us say our store’s algorithm identified deeper interests, such as liberal economics or suchlike. Here’s the catch. If your algorithm is weighted too much on past interest, you may never be recommended a book that has a brilliant argument in favor of an opposing conservative opinion.

Worse, your store is likely to tune their recommendation algorithm to make you buy more, rather than broaden your mind. So it would probably recognize and categorize you as having a weakness for a certain broad sociopolitical identity, and then pump you with books that are self gratifying and hence tempting to purchase. The algorithm would create a personalized intellectual echo chamber, that you would love to be a part of.

Now I’m not saying that the recommendation algorithms at Amazon actually do this. But it’s a real danger facing the patrons of such a store, unless the store’s algorithms are purposefully designed to be principled. This could be achieved in many ways, such as by purposefully mixing in books written by authors with opposing sociopolitical identities, or by including random new publications in every recommendation set, or both. But the important point is that the software engineers who design the algorithms need to have these broader ethical motives in mind when doing so.

The above examples are just two of a great many other instances where algorithms can veer away from serving the customer’s best interest, towards the dark world of mind hacking for profit. Media bots that mistake popularity for newsworthiness and favor the snowball effect, big data algorithms that ignore multivariate analysis and make false predictions or proclamations, enrollment algorithms that favor set outcomes for identities such as gender or race, over equal opportunity on a case by case basis, are some other examples of where algorithms can become (perhaps unwittingly) immoral.

Some might argue that its not software engineers who decide the broad motivations behind algorithms, but rather that it is company owners and big businesses who do so. This might indeed be true, but if software engineering as a profession was to be firmly grounded in secular ethics, and IT courses were to have compulsory learning modules that demonstrate how algorithms can go awry and produce unethical outcomes for its human consumers, then this would be an excellent start. In fact, it would be important for universities to introduce an entire new category of software testing into the traditional bandwagon of Functionality, Usability, Security, Performance and Data testing. One could call this category “Ethics Testing” or “Fairness Testing”, and it would signify test cases and test scenarios to identify threats to the physical and mental wellbeing of the human consumers of the software.

NB: Original article on LinkedIn Pulse: https://www.linkedin.com/pulse/developing-principled-algorithms-increasing-necessary-ruwan-rajapakse/

From neurons to knowledge; the inspiration from our brains that sparked the quest for Artificial Intelligence

The buzz on AI is bigger than ever in digital innovation land these days. We all tend to get excited when we hear of new devices and systems that are being developed, which are said to be “intelligent” and which mimic human behavior. Ranging from IBM’s Watson API to virtual agents to self driving trucks to smart anti-personnel drones, we see an explosion of technology that was originally inspired by the workings of our own Natural Intelligence; which is a creation of our brains and the neural networks that underpin them.

An Artificial Intelligence is a goal-driven machine or system deriving semantics – i.e. meaning or relevance from the environment it navigates – and thereby taking actions that increase its chances of achieving its goals.

So how does a goal-driven system becoming intelligent? Let us consider how intelligence arises in our brains through neural computation, which most AI systems also mimic in one form or another.

There are a few basic facts we must first know about our brains. Our brains consist of vast networks of interconnected neurons (a type of specialized cell). Here is a typical neuron.

A neuron is a vastly complex entity, and is almost a separate living organism in itself, but for the purposes of understanding the basics of how intelligence dawns on us, we can think of it as a sort of fancy Logic Gate (recall to mind your electronics lesson from GCE O-Level days). As you can see, there are several types of connectors protruding from the cell body. The Dendrites serve as Inputs to the cell, and the Axon serves as the one single Output. The Inputs and Output consist of electrical impulses of a variable frequency.

Apart from the Inputs and Output, a typical neuron also has two other important variables that determine its function. One is what is known as the Synaptic Strength, and the other is the Threshold. The Synaptic Strength (aka Weight) is a chemically controllable barrier in the connection between the Axon (Output) of one neuron and the Dendrite (Input) of another. Like a sort of chemically controlled Variable Resister connecting two neurons. The Threshold is an internal barrier potential that must be exceeded for an Output to be triggered. Consider the following simplified functional model of a neuron.

How a neuron works is that if (Input1 X Weight1 + Input2 X Weight2 + Input3 X Weight3) is greater than the Threshold, then the Output is 1. Else the Output is 0.

Given this functional behavior, we can easily show how a neuron can be trained to behave as an AND Gate. Let us say that the Weights are 0.3, 0.3 and 0.4 respectively, and the Threshold is set at 1 Hz. If each input is triggered with a frequency of 1 Hz, then 1 X 0.3 + 1 X 0.3 + 1 X 0.4 = 1, and therefore the Output becomes 1 Hz (i.e. True). If any one or more of the Inputs are 0, then the Output becomes 0 Hz (False). This neuron is now behaving like a multi-input AND Gate.

Let us now consider a neuron that is trained to behave as an OR Gate. We can set all the Weights at 1, and leave the Threshold at 1 Hz. In such a circumstance, the Output will fire (become True) if any one (or more) of the Inputs are triggered at 1 Hz. You have your OR Gate.

So far so good. But how does a network of conditional switches grasp truths about the world? The beauty is that once we know that neurons can function as multi-input AND and OR Logic Gates, we don’t need to care anymore about the above details like Thresholds and Synaptic Weights. We can begin to see how a network of such multi-input Logic Gates could be made to recognize real world objects and concepts with the aid of input sensors.

Let us consider a very simple real-world concept like a line. Let us say its important to recognize and act upon a line-like object, given a set of five linear point-like sensors. Lets us also assume that in order for an object to be a line, it must at least be of three sensors in length. Now take a close look at the below network of Logic Gates.

As you can see from the above diagram, a line-like object (in red) is pressing against three of the five sensors. Using common logic, we can see that the output of the gate network will be True. The beauty is that if we slide the line anywhere along the sensors, it will return True. But any object lesser then three contiguous sensors in length will return False, no matter where you slide it. This key insight is the origin of generating meaning through a network of neuron-mimicking logic gates.

We can extend the above line-detector concept in our minds and see that we can wire a network in three dimensions, for a 5 X 5 square sensor panel, which will be able to detect a line greater than three sensors in length, touching the sensor panel in any orientation.

Extending this concept much further, and with the aid of inhibition, feedback and other complex factors that mimic the actual function of neural networks in our brains, scientists are able to show that such networks can focus on edges, detect movement, and recognize shadows, shapes, words, and other important characteristics in our environment. A single “downwind” neuron (or equivalent logic gate) can be made the holder such a complex characteristic as the face of one’s mother. A Jennifer Aniston neuron was actually discovered in the brain of a human subject, in a ground breaking experiment conducted by a team from the University of Leicester, some years ago.

The science of Artificial Intelligence based on networks that perform neural computation – the most exciting and hopeful frontier of human advancement that one can foresee – has developed over the years to become an extensive and exceedingly complex one. Yet it is a truly exhilarating feeling one gets when one begins to understand the rudimentary concept that inspired this revolution in the first place.

Republished from my LinkedIn article: https://www.linkedin.com/pulse/from-neurons-knowledge-inspiration-our-brains-sparked-ruwan-rajapakse/?published=t

Is there even an alternative to agility?

I keep bumping into business owners, CIOs, and, astonishingly, sometimes even tech entrepreneurs who like to begin the conversation by saying “we don’t believe in agile development”. What’s more, I’ve seen a good many others who, whilst paying lip service to agility and the good practices of Scrum, Kanban and similar frameworks, still fail to adopt the fundamental frame shift from rigidity to agility. This has been particularly evident in how these folks managed their requirements, costs and relationships on the ground. I’d like a quick word with these agile holdouts, and anyone else who needs motivation to adopt the agile development mindset.

The question to ask here is fairly simple, and best exemplified with an analogy from Physics. Is Scrum a historical step forward from Waterfall, like the step General Relativity took, when it outperformed Newtonian Gravitation, yielding more accurate results under extraordinary circumstances? Or, is Scrum much more like the leap forward that Newtonian Mechanics made, when it entirely replaced Aristotelian Mechanics – which in hindsight turned out to be plain batty? I’m inclined to think the latter.

Why? Building a software solution is not quite like building a bridge. The reason for this lies, surprisingly, not so much in the details of the engineering process, but rather in our big-picture objective of building that software solution in the first place. Unlike in the case of bridge-building, software solution development involves extensive, ongoing learning from the environment. In this case, the “environment” is the collection of intentions and fancies that the human users who would interact with the solution, have. Rather like in Darwin’s view of nature, a rudimentary piece of software evolves through hindsight into a highly useful – and consequently complex – one. We couldn’t possibly fathom its complete purpose and usage at the onset.

So, the process for understanding user expectations and providing a digital solution for a given domain (such as dating or crypto currency) must necessarily be gradual, involving the production of an initial artefact that collides with end-users, and then changing it continually according to a steady stream of end-user feedback. Therefore, a MVP release aught never to be seen as a mere step or two away from the final deal.

Also, development teams should not be evaluated only for their peak performance, when they are smashing the ball around the park in the first month or two. The whole delivery framework must be geared towards gradual learning, and the development team evaluated for domain interest, stamina and openness to change.

As such, nurturing a motivated, collaborative development team, whose contribution can be scaled up or down based on the velocity of end-user demands, is absolutely essential in today’s ultra-competitive business environment. Business owners, CIOs and tech entrepreneurs should gravitate towards appropriate contractual and/or compensation models that foster such collaborative teams. Scrum and Kanban methodology and their associated compensation models for extended offsite teams, such as monthly retainer, have matured sufficiently to deliver software product increments that can chain-react with end-users and evolve very successfully.

Its time to bury “waterfall” and similar low-collaboration delivery frameworks, and their associated adversarial mindsets, into the annals of history.

What’s in a Business Analyst these days?

A few weeks ago, a graduate from a reputed university, who had studied software engineering, stopped by for an interview. He had applied for a role in Business Analysis, and naturally the first thing I asked him was, “what do you think a Business Analyst (BA) does, in todays agile development environment?” He answered that a BA’s role is to help developers understand product requirements. So I asked him if he could imagine himself working on a project, and describe in some detail what he’d do to elicit product requirements from a business owner.

He struggled with a clear answer. At first I thought it was just his nerves acting up at the interview, but the more I got him comfortable, by asking him smaller questions like “what diagrams or artifacts would you use?”, the more he seemed to grope in the dark about the overall job role. He managed to name some of the diagrams and techniques used in business analysis, like sequence diagrams, use cases and flowcharts. And he was even able to describe what purposes they served. But he was unable to string together the different activities and tools, and describe the job role in a cogent manner.

After some discussion, I realized that he actually was not a poor candidate, at least in terms of his intelligence. Rather, the big picture of how one would set about serving the role of a BA seemed not to have been a discussed, as a part of his academic training.

I think he is not alone. I recall to mind the relatively small number of BA interviews I’ve taken part in (I’ve participated in over a thousand developer interviews, and yet only a half-dozen or so BA ones), and it seems to me that all of them ran the same course. So I thought it might be helpful to software engineering students, if I could lay out a plausible big picture approach to Business Analysis, which is aligned with today’s agile app development zeitgeist.

A BA would ideally commence his job brainstorming with the business owner or entrepreneur seeking to develop the technology solution, and defining the solution holistically, by penning down the overall idea in a couple of English Language paragraphs. Let us consider a Company “Leave Management” Solution as a hypothetical example.

I’d like to build a Company Leave Management Solution, where company Employees can see their leave balances and apply for leave through a mobile app, and where their Line Managers can approve leave via this same app, and where Executives can see reports on leave entitlement, usage and trends. The HR Team would be able to update employee leave balances through a Web portal, for speed and convenience. I’d like to have employees and companies enroll for the solution as a readily available service, by downloading the app on smartphones and signing up. We’d like to deploy a minimal solution within 3 months at the onset, learn from our experiences with our client companies, and thereafter extend our solution to suit their deeper needs.

This type of big picture vision, stated in simple language, is so important for the success of any project in app development or digital transformation. It sets people in the right direction, in terms of the broad vision, functions and actors that the solution encompasses, as can be seen at the solution’s inception. I would even suggest producing a marketing brochure or Website describing the concept to its customers, prior to actually building the solution itself. Understanding the big picture is so important. Once this big picture is understood, we can then move on towards defining the actual software solution in an agile, evolving manner.

We can then identify the actors in the solution, as can be seen at this moment. Like for example, a HR Manager, an Executive, an Employee and a Line Manager. We’d then setup a brainstorming session, and list down the User Stories that describe the expectations that each actor would have of the solution. A User Story from the previous example of the Leave Management solution might be, “As an Employee, I want to open a Leave Application Form in my mobile app. I’d like to see my Leave Balances for each Leave Type, in this form. I’d want to pick the Leave Type I’m applying for, and schedule the dates through a Calendar Control. I’d then like to Submit my Leave Application, which would be routed automatically to my designated Line Manager.

Once we’ve understood the solution’s User Stories, we can begin to prototype the User Experience (UX) in earnest. This would typically take several brainstorming sessions to conclude, depending on the size of the solution. We’d first white board rough sketches of the user screens and their input/output controls, so as to deliver the input/output expectations we have in our User Stories. It might be helpful to initially list down the approximate screen names, that correspond to the overall mental picture we have, of how users would intuitively engage with the system. This UX development activity should always be done in discussion with the product owner(s), a UX/UI developer and other helpful strategists, in a requirements elaboration workshop.

Once a few screens are conceptualized and white boarded, the workshop should break, allowing for offline thinking time for the participating team, as well as time for the UX/UI developer to produce a low-fidelity mockup of the discussed user experiences. There would always be refinements that would get voiced at the next workshop session, about the previous session’s screens, whose mockups would be shared at the beginning of each new session.

After several sessions of UX white boarding, we’d end up with a complete first cut of the solution’s User Experience, and the mockups would now be fit for high-fidelity creative enhancement. It would be important at this stage, to begin creating a Software Requirements Specification (SRS) document, in which the User Stories and mockup screens would be embedded. Business rules that apply to the user actions in each screen would be described, along with any input validation rules that apply for the screen.

Let us consider the example of the Leave Management solution. We would perhaps only allow integers values and multiples of 0.5 for the Leave Quantity field, and it should not exceed the Balance Leave for the Leave Type chosen to be applied for. This would be logged as an input validation rule. The error messaging for input validation violationswould be jotted down, such as “You have applied beyond your available Casual Leave balance. Please try again”. Most importantly, the business rule that (say) the submitted leave will be immediately deducted from the Leave Balance for the chosen Leave Type, on pressing the Submit button, would be noted. We’d ideally also record any special data types for storing the information gathered from the user, in this SRS narrative.

The above process is the most rudimentary one for eliciting and defining requirements for a piece of software to be developed. There may be several other important sections in a SRS document, for communicating non functional requirements such as security and performance expectations for the solution. There may even be some software solutions that are so sophisticated in scope, that they require complex mathematical models or other tools to help define them completely. But it is highly likely that the overwhelming majority of software app solutions would follow a BA process like the one we’ve described above.

In conclusion, its important to understand that this BA process is not a waterfall-like, irreversible process. The essential activities may be repeated, from Release Cycle to Release Cycle, where previously defined requirements may be changed, and the SRS updated as a living document.

–Republished from my LinkedIn Article: https://www.linkedin.com/pulse/whats-business-analyst-days-ruwan-rajapakse/

Measures of enlightenment: Do all prominent factors for a happier existence show signs of steady historical progress?

Steven Pinker, a Harvard College Professor and one of the world’s leading authorities on Language and The Mind, is due to publish a new book titled “Enlightenment Now”, demonstrating historical progress in the human condition. Or at the very least, he expects to show us that humanity’s approach towards progress has been a learning experience, and that we’ve steadily incorporated measures that foster better living conditions for ourselves. This piece of research is a sequel to his equally illuminating earlier thesis titled “The Better Angeles Of Our Nature”, in which he elegantly demonstrated how human violence has declined over the decades, centuries and millennia.

I keenly look forward to his new book. I would also like to raise a question that arises as a natural consequence of the book’s title, and the preview of its core thesis that Prof. Pinker has shared here. I agree that the steady decline of physical violence, the evidence of economic growth not being at cross purposes with environmental conservation, the improvement in living conditions through increasing adoption of technology etc, are all contributory factors to increasing wellbeing. Yet I still wonder if human attitudes towards each other have shown an equally steady improvement, leading to higher levels of inner satisfaction and external social equilibrium.

In simpler words, have we learned to live happier with each other and with ourselves? I think this is a justifiable, empirical question that can be posed objectively. The factors hinted by professor Pinker unquestionably impinge directly or indirectly on our happiness, such as iPhones bringing in easy conveniences and opportunities for relaxation or the reduction in oil spills leading to a safer environment. However, these kinds factors are the more easily measured, “material” outcomes of human progress, and whilst important to human happiness, are nonetheless incomplete measures of our progress towards enlightenment and the lessening of human suffering.

Daniel Kahneman tells us that we communicate our subjective wellbeing to others in two ways: through our experiencing self and our remembering self. In brief, our experiencing self is our moment-to-moment feelings, expressed through our speech, smiles, laughter, tears and such. Our remembering self is our memory of past experiences. Kahneman tells us that these two kinds of reports can sometimes contradict each other, for one and the same experience, such as whether we enjoyed a particular vacation. The important point though, is that our happiness in both cases can be measured at least to some degree of approximation.

Equally important is the concept of “social technology”. By “social technology” I speak of the innovations throughout history in human attitudes, that reshape (or claim to reshape) our character for the better, and lead to enhanced inner happiness as reported by their adopters and as experienced by those whom they come into contact with. There has been a vast range of such social technologies invented throughout history, some arguably obsolete, and some potentially of great value to this day.

Gods one can appeal to and seek solace when in difficulty, The Golden Rule of the ancients, the dissolution of the self and the practice of self-detachment in the face of adversity (as advocated by the ancient Buddhists and some modern secular thinkers), the art of forgiveness of the early Christians, the stance of non-punitive, restorative justice, the precedence of evidence over speculation or hearsay, the abandonment of superstition and the application of pragmatic solutions to life’s problems, the scientific method, psychology and the science of counseling, the art of reasoning through differences of opinion, the limits of income towards increasing happiness, or even such mentifacts as Sam Harris’s base moral stance of not harming the “wellbeing of conscious creatures” are some diverse examples of social technology that emerged in our turbulent history.

These types of social technologies have been claimed to help us live together in greater harmony, and/or yield greater inner fulfillment. It may be possible to research and chart our progressive adoption of such “social technologies”, over history. Have we adopted the ones that yield results, to the same extent that we brush our teeth, and abandoned the ones that don’t, to the same extent that we don’t use charcoal anymore to brush our teeth?

Yet another angle to my question is to inquire into the progress of our spiritual development. I used the word “spiritual” advisedly here, in the same context as someone like Thomas Metzinger would, where spirituality is understood to be a deep sense of intellectual honesty. Are we more intellectually honest today than we were in the past?

There is of course the tricky problem of determining measurable outcomes that demonstrate the mental wellbeing and spiritual development of humanity. What tangible outcomes can we survey to assess the qualitative progress in human wellbeing, beyond the material substratum shared in Prof. Pinker’s preview? We are not talking about oil spills here. Also, we are unlikely to get a very meaningful result by simply asking people “are you happy” over a period of several decades. Instead, I propose a survey of the following factors.

  1. Do young adults communicate a greater sense of optimism now than in the past?
  2. Are we more prolific today in our hobbies and pursuits per capita, adjusted for age and other relevant demographics, than in 1950?
  3. Do we claim to feel less anxious or threatened on a moment-to-moment basis?
  4. Do we report feeling progressively lesser mental duress – at work for example – over history, and exhibit lesser numbers of stress-related maladies and social consequences?
  5. On a mundane basis, are we increasingly free to be ourselves, express our opinions uninhibited (within the limits of non-violent conduct), without experiencing harmful consequences?
  6. Have we become less judgmental and more objective over time, focusing on resolving ethical dilemmas rather than assuming stances and polarizing society?
  7. When transgressions happen, are we more forgiving, understanding the frailty of the human condition?
  8. Have we progressively given less credence to speculation and hearsay, and focused more and more on evidence, in the media for example?
  9. Are we less superstitious, and increasingly rational?
  10. Are we kinder to each other, and do we help each other oftener, volunteering to provide advice and guidance?
  11. Do we have more fulfilling social lives, and have more friends and soul mates than (say) 50 years ago (the kind whom we discuss issues with, not our friends list on Facebook)?
  12. Are we more satisfied with what we’ve done with our lives, and have fewer regrets?
  13. Is jealousy on the decline?
  14. Are parents being more like friends and educators to their children, and less like authoritarian figures?
  15. Is politics becoming increasingly about team performance, and less about clan identity?
  16. Do we smile more, and laugh more?
  17. Are sociopathic personalities increasingly filtered out for positions of authority and for public office?
  18. Is public intellectualism becoming increasingly and uniformly popular in both the East and West?
  19. Do our education systems distill the progressive social technologies in an objective, secular manner, and instill them in our children?
  20. Do we brainstorm more often as families and social units?

Obviously one would have to structure these rather loose and diverse factors into a more cogent scientific proposition. It would certainly make an interesting research project, and perhaps Prof. Pinker’s new book addresses the above. Research into some of the individual factors mentioned may already be available, and it would perhaps be a case of mining the data and looking for an overall trend in the context of happiness and wellbeing.

My own hypothesis is that human beings have become less authentic and more hemmed in by retrograde social technologies like extreme political correctness, or by downright fear of unforeseen negative reactions. A few decades ago, a person would at worst have been politely chided for an inarticulate, yet well-meaning remark, such as “Tamils are good people”, for which one would be ostracized today. I suspect we have become so finicky that we hardly voice our thoughts out aloud in our own peculiar way.

Furthermore, I conjecture that in the past, there have been at least two peaks in human mental wellbeing, with troughs in between. One peak was at the height of the flourishing of the hunter-gatherer (or “noble savage” as Yuval Harari would say), just prior to civilization and the establishment of extensive social hierarchies. Wellbeing took a nosedive after civilization, because one’s actions were now largely controlled by a system beyond ones power to change, and because vast numbers were literally enslaved for the purposes of economic progress.

A more recent peak was reached shortly before the rapid acceleration of the information technology revolution, somewhere in the 1950s. This was the time when Western economies were growing, scientific optimism was high, population pressure was lower in the East, most occupations were not so specialized, and lucrative opportunities were widespread for those of average cognitive ability.

I theorize that at the present moment we are approaching some sought of a local trough in human wellbeing, where the rat race is so cranked up, where population pressure in the East is still very high, where lucrative opportunities are shifting to the right tail of the ability bell curve, and where earlier socialist or welfare state ideals are frowned upon to such an extent that the baby it encapsulated (aspirations of fair play and a more even distribution of comforts) has been thrown out with the bathwater (the hegemony of the former communist dictatorships).

One hopes get more answers from Enlightenment Now, and to research this topic further, about whether we are progressively happier to our own estimation, and whether our attitudes towards our fellow beings have become steadily more enlightened.