Sunday, October 09, 2016

On the mansplaining hysteria

A large computer program which may appear as complex as a city, is essentially a web of decisions made by its programmers, aka developers. In my opinion, there are few human endeavors which involve decision making to the extent that programing does.

Given the long-lasting nature of many of our decisions, we, developers, try to chose wisely among alternatives. We do this by learning from past experience and discussing the choices among peers. The most important decision made during the lifetime of any program is the initial choice of the programming language. Indeed, the choice of the language has profound implications on the structure of the program being developed. Moreover, the choice of the language in a given program is for all practical irreversible.

Although it is relatively easy to learn a new language, it takes several years to become proficient in it. Given the investment involved, we, developers, tend to be very protective of the programming languages we already know well. Thus, the statement "language X sucks" is met by "that's because you do not understand the beauty of language X." This pattern has millions of instances on the Internet, repeating itself on a daily basis.

As a personal example, let me mention the following anecdote. Around 2010, I came across Scala, a programming language developed at EPFL, my Alma Mater. Appreciating the expressive power of the language, I started the "Scala Enthusiast Group Lausanne (Swizerland)" and chaired the group for a few years. At one point during this period, I complained about a particular feature of Scala with a tweet. A Scala enthusiasts quickly tweeted back that my complaint had no merit and that it should be viewed in light of my apparent anti-Scala bias. This is like accusing the editor of an LGBT journal of homophobia. Have I mentioned that we programmers can be over-protective of their favorite language?

Very recently, I stumbled upon a tweet by Jen Golbeck where she complained about Java, a widely-used programming language. Ludovic Reenaers tweeted back with "learn it well and you will never want to use anything else". To this, Golbeck replied that she gave Java classes and "shit" like this was very frustrating. I intervened by remarking that given the protectiveness of programmers regarding their preferred programing language, Reenaers' original comment could not and should not be construed as sexist.

Golbeck replied that she did not "want another dude to explain how I [she] should feel about things!" Then a bunch of ladies chimed in with how "manspanners were flooding in" and "this is mansplanning elite". My attempts at providing another perspective got no traction and were angrily dismissed.

Given the CS culture with which Jen Golbeck is presumably familiar with, I still fail to understand how she, a college professor teaching CS, can be offended by a common discussion pattern and detect a gender bias where there was none.

The exchange with Golbeck et al. is interesting insofar as it illustrates unwillingness to hear a different perspective. Not only that, any attempt at a conversation is met with accusations of sexism. Those flinging frivolous accusations of bigotry based solely on the gender of their interlocutor are being bigoted themselves. Such behavior is an insult to reason as well as human decency.

Monday, July 11, 2016

Milking the Tesla S auto-pilot accident

Fortune Magazine recently published an article accusing Tesla Inc. of hiding material information in its $2 Billion stock offering dated May 18th, 2016. The article suggests that by failing to disclose the adverse affects of the fatal accident dated May 7th, Tesla Inc. may have misguided buyers about the future of the company.

 Elon Musk' reacted angrily to by calling the article BS.

This less than amicable exchange was mentioned in an article published by my local (Swiss) newspaper. This article, like many others, seemed to focus on Tesla's follow up on the accident and the future of auto-pilot, while ignoring salient facts about the circumstances of the accident. More specifically, the fact that Mr. Frank Baressi, the driver of the tractor trailer, cut the path of the Tesla S driven by the late Joshua Brown on a separated high-way, is mentioned only fleetingly. Cutting the path of traffic in a separated highway is a high-risk maneuver. While auto-pilot may share some of the blame, automatically exonerating both Mr. Frank Baressi and the late Joshua Brown from any responsibility in the accident seems disingenuous.

As for the $2 Billion stock offering, for a company bleeding cash at the rate of $500 million per trimester, nothing seems more natural than raising cash for future investments. Fortune magazine must know this.

The press seems to think that articles on reactions to articles published in some other paper is somehow newsworthy. The signal to noise ratio of the numerous articles published by Fortune magazine on Tesla seems very low and on par with click-bait internet pages we all love to hate. It looks like Fortune magazine, like many other papers and magazines, are merely milking the Tesla phenomenon.

Elon Musk took a risk in integrating auto-pilot before other car manufactures. He paved the way for auto-pilot technology to enter mainstream. Some may call this being brave.

Saturday, July 02, 2016

Impatiently awaiting the unconditional surrender of the human driver

In recent days, much was written about the fatal traffic accident involving a Tesla S in auto-pilot mode. In an ironic twist, the deceased driver, Joshua Brown, 40, of Ohio, published several fascinating videos about the auto-pilot mode of his beloved car.

It appears that the car's camera did not detect the trailer crossing the road due direct glare from the sun. The long range radar of the vehicle apparently also failed to detect the crane due to its hollow shape. There are some indications that Mr. Brown's Tesla was moving very fast. This would explain why Frank Baressi, age 62, the driver of the tractor trailer, did not see the Tesla and cut its path.

With full details of the accident still missing, one can reasonably conjecture that, cut off by a large vehicle with no warning, the accident would have occurred even in the presence of a fully alert human driver.

Notwithstanding the dozens of articles about the accident, the responsibility of Frank Baressi, the driver of the tractor trailer, is rarely, if ever, mentioned. Had the accident involved two human drivers without auto-pilot, we would have instinctively assigned some of the blame on Mr. Barressi. With auto-pilot in the picture, we tend to focus on the technology. Thus, we seem to set a higher bar of safety for auto-pilot, a technology in its infancy, than we do for human drivers.

In my mind, by focusing on the technology, we implicitly admit that humans can be (are?) bad drivers. We get impatient; we get tired; we get old; we drive under the influence of substances. The machine will never get tired, old, impatient or drunk. It will never overtake before a turn, succumb to road rage or cut the path of a bicycle. There is little doubt that after initial kinks solved, auto-pilot will significantly reduce road fatalities throughout the world, avoiding injuries and saving millions of lives. As such, I am impatiently awaiting the unconditional surrender of the human driver.