Predictably Irrational

Posted by Brian Wed, 03 Mar 2010 23:13:00 GMT

Predictably Irrational: The Hidden Forces That Shape Our Decisions
By Dan Ariely

5/5

Predictably Irrational was the right book at the right time for me, making it one of the best books I have ever read. In it Dan Ariely discusses the findings of behavioral science and how they contradict the assumptions of classical economics.

Classical economics assumes that the actors in a market are completely rational and will always act to maximize their personal gain. Behavioral economics breaks from these assumptions by putting them to the test and, as any honest person can attest, finds that they simply are not true. We are not rational beings. What is being discovered is that we are irrational in consistent and predictable ways though.

Ariely primarily highlights our irrationality through experiments. Each chapter discusses an experiment he or his colleagues have ran to test some aspect of our supposed rational actions. For example, he discusses the influence of sexual arousal on the choices we make. Participants in the study answered a series of questions about their sexual activities and preferences. They were asked to estimate what their preferences would be when aroused. Then the questions were repeated when they were highly aroused to see if they correctly predicted. The results were clear: people can’t estimate how their behavior will change when aroused. The most interesting question referred to condom usage. When unaroused the vast majority of participants said they would take the time during a sexual encounter to put on a condom. However, when they answered the same question aroused, the number plummeted. He then extrapolates that perhaps these types of experiments should be taken into account when advocating abstinence only education.

Another experiment analyzed the effect of a placebo pain killer. Two groups of patients were told they were being given a new pain killer, but one group was told that it cost $2.50/dose and the other group was told it was only $0.10/dose. Those told it was more expensive were more likely by a significant margin to feel that it worked. Experiments such as these can have a profound effect on health policy.

In a way this is a return to the economics of Adam Smith, who classical economists often trumpet as a free market champion. What they either don’t realize or simply ignore is that Smith wrote The Theory of Moral Sentiments in 1759, which, as Ariely reminds us, “notes that emotions, feelings, and morality are aspects of human behavior which the economist should not ignore (or, worse, deny).” Classical economics gives us many useful tools, but they must prove their mettle in a reality that accepts the fact that we are irrational beings.

Posted in  | Tags  | no comments

The Innovator's Dilemma

Posted by Brian Thu, 25 Feb 2010 01:36:00 GMT

The Innovator’s Dilemma: When New Technologies Cause Great Firms to Fail
By Clayton Christensen

5/5

How do successful companies fail? Often the answer is obvious: poor management or an economic downturn are two common culprits. More interesting are the companies that seem to do everything right, but a few years later are in a steep decline. How is it possible that management which only a few short years ago was being lauded as a model for the industry can come to be regarded a blathering idiots with no clue as to what their customers want? In The Innovator’s Dilemma Clayton Christensen puts forth that this is caused by the introduction of disruptive technology into the market.

What is disruptive technology though? Christensen defines it as a technology which has worse performance, at least in the near term, in what has been considered the key market measurement but which is still considered acceptable. It trades off max performance in this key measurement for features that customers outside of what has been to now the core market care about more. He uses the disk drive industry as his major example throughout the book. When a new size disk drive was developed the established players repeatedly ignored it because its storage capacity wasn’t interesting to their current customers. However, its smaller size was interesting to a new market and because technological innovation often moves faster than market demand, eventually the disruptive technology is able to displace the existing technology, along with the companies pushing it. He explains a similar process with hydraulics in excavation, minimills in steel production, and discount retailers.

I was pleased to learn that The Innovator’s Dilemma is often used in MBA programs now, although I do wonder how it is received by both faculty and students. The idea that good management can be a direct cause of failure is probably a non-intuitive and disquieting thought to many of them. Christensen is an engaging writer with the data to back up his theory. Highly recommended.

Posted in  | Tags , , ,  | no comments

Batch Modify 0.4.1 Released

Posted by Brian Sun, 21 Feb 2010 21:13:00 GMT

On Friday I released a bug fix to the Trac batch modify plugin that addresses a problem users of Trac 0.11 and 0.11.1 were seeing. 0.4 changed the batch modify form to respect the restrict_owner setting, but it does this through a method in the Trac ticket API that did not exist until 0.11.2. This bug fix simply checks to make sure your version of Trac supports this call before attempting to use it.

Posted in  | Tags , ,  | no comments

Batch Modify 0.4 Released

Posted by Brian Wed, 17 Feb 2010 01:32:00 GMT

I released the next version (0.4) of the Trac batch modify plugin today. This release fixes a bug that caused the form checkboxes not to enable the corresponding field when checked. This bug was introduced in 0.3 when I removed some deprecated JQuery syntax. A couple of enhancements were included as well. The restrict_owner config option will now be respected on the form and new keywords will now be merged with the old keywords on each ticket. Previously the old keywords were simply thrown away.

Posted in  | Tags , ,  | no comments

Misapplying the Golden Rule

Posted by Brian Sun, 14 Feb 2010 05:10:00 GMT

Do unto others as you would have them do unto you.

I have learned that the Golden Rule as stated above is worthless to me. The reason is simply that what I would have others do unto me is very different than what others want done unto them. For example, in debates I prefer people to simply tell me that my idea is stupid and logically list out all of the reasons why it is stupid. Once both sides have listed out their opening reasons a healthy debate can ensue much more quickly. In a debate I feel that my feelings are irrelevant to the goal of reaching the correct conclusion. Unfortunately, others do not feel this way. If I apply the Golden Rule in a such a scenario the other side will probably view it as a combative personal attack. While it may be combative, it most certainly is not a personal attack as long as the debate stays focused on the ideas. While the reasons may imply that the person presenting the idea is stupid, that is immaterial to the debate itself.

Of course, my application of the Golden Rule is naive. A extension of this is the so-called Platinum Rule advanced by Karl Popper, “The golden rule is a good standard which is further improved by doing unto others, wherever reasonable, as they want to be done by”. This seems closer to to societal expectation of how the Golden Rule is to be applied. Given this it seems that my role in life is to be subservient to the social norms of others, even though I would rather not be treated that way.

no comments

The Mythical Man-Month

Posted by Brian Sat, 13 Feb 2010 02:04:00 GMT

The Mythical Man-Month By Fred Brooks

4/5

The Mythical Man-Month is the most influential software project management book ever and I am not conceited enough to think that I will have any original thoughts on its contents, but here goes.

The most widely known item to come out of this work is Brooks’ Law, which states that adding manpower to a late software project makes it later. This is a very counter-intuitive statement to the non-programmer, but makes perfect sense to those who practice the craft. Every project is different and the time it takes new team members to get familiar with the project pushes the rest of the project behind schedule even more. This of course causes management to add even more manpower and the project enters a death spiral.

He also gave a citation of the oft-mentioned 10:1 programmer productivity ratio, which states that an excellent programmer can be ten times as productive as a mediocre one. It comes from a study by Sackman, Erikson, and Grant.

While the core is still strong, around the edges he book is showing its age. Whenever Brooks talks about specific technologies, programming languages, or methodologies everything feels dated. At this time it was still a big debate as to the usefulness of languages like FORTRAN over assembler. He also pushed what I felt was a heavy-handed organizational concept that he called a “surgical” team. The team involved a large number of support people around a a central architect who made all the final decisions. A number of these people wouldn’t be needed today because of technological advances, but I still wasn’t sold on the idea.

Overall, this is of course a must read for any programmer or manager of. It is important that Brooks is talking about management in the context of very large projects though. My work comes no where near to the scope he is used to and the ideas can seem heavy-handed because of this. With that being said, any book that has this much staying power has to have something to it. While everything here is not gold, a large chunk still applies.

Posted in  | Tags , , ,  | no comments

Open Sources 2.0

Posted by Brian Thu, 04 Feb 2010 00:21:00 GMT

Open Sources 2.0: The Continuing Evolution

4.5/5

Much has changed in the world of open source since the publication of the original Open Sources in 1999. Open Sources 2.0 is another collection of essays that discuss many of these changes, touching on a wide array of topics in open source software and in the application of open source principles in other areas. As with any collection of essays there is no coherent narrative, so I will discuss several essays that I enjoyed. This review would be very long if I discussed all of them as only a few were uninteresting to me.

Jeremy Allison, a lead developer of Samba, contributed an article comparing the POSIX and Win32 standards. POSIX is a series of related standards that defines the interface for a UNIX operating system, while Win32 is Microsoft’s de facto standard API for Windows programming. Allison’s Samba experience makes him uniquely qualified to compare the two. He gives a balanced take, looking at one bad and one good feature of each. For POSIX he looks at the horribly broken locking calls (AT&T submitted a terrible proposal in the standardization process and nobody else cared to complain) and its impressive record of future proofing. For Win32 he gives kudos for excellent Unicode support before any other vendor, and a big thumbs down for a powerful, but overly complicated security model. It’s actually possible to write very secure code with Win32, but it is so difficult to use the security model that nobody really wants to. He concludes by advising you to stay away from de facto standards.

Andrew Hessel contributed an essay focused on applying open source principles to biotechnology. He describes how current IP practices have contributed to rising drug development costs and waves of mergers and acquisitions as firms try to consolidate fragmented IP. He makes an interesting observation that DNA is binary code as only two base pair combinations are possible and extrapolates upon that to wonder if an author of synthetic DNA could copyright OSS style. Very few biotechnology companies now can produce and market and product profitably so the value of proprietary IP may be overstated in this industry. I don’t hold out much hope for this approach being adopted by any large players.

Overall, this collection of essays is extremely thought provoking. The focus here is not code, but the issues surrounding open source software. By extension several essays also look at how thee principles can be applied in other areas. Highly recommended.

Posted in  | Tags , ,  | no comments

New Camelot Center Website

Posted by Brian Sat, 30 Jan 2010 00:30:00 GMT

The Camelot Center is a charity that offers hippotherapy to those with disabilities. I have been working on a new website for them in my spare time. Last night the initial version went live.

This is my first live website written with Ruby on Rails and I must say it was an enjoyable experience. In addition it was the first website I have done attempting to use proper semantic markup and CSS design. There are no tables in the layout. I’m not very good at graphic design, but given my lack of talent I am fairly happy with the visuals I came up with. There is still a lot of work that needs to be done, but right now it gives them what they need.

Tags ,  | no comments

Batch Modify 0.3 Released

Posted by Brian Tue, 26 Jan 2010 23:30:00 GMT

A few weeks ago I started maintaining the batch modify plugin for Trac. This plugin does pretty much what its name says. It allows you to modify Trac tickets in a batch. Prior to my taking it over it had not been maintained for over a year and didn’t even work on Trac 0.11. This version fixes a couple of bugs so the plugin actually works and gives a couple of UI enhancements. Details are available here.

There are a lot more bugs and feature requests to go through. Hopefully I will have another release in about a month.

On a personal note this is my first open source project that people actually use and it is quite enjoyable. pyFish is a lot more complicated of a project, but nobody besides me actually uses it.

Posted in  | Tags ,  | no comments

My Apache-Fu is Weak

Posted by Brian Sun, 24 Jan 2010 00:49:00 GMT

Update: My Apache-Fu is strong again. Old links should now be working again.

So the upgrade changed my URL’s yet again. Now article URL’s are of the form /%year%/%month%/%day%/%article-name%. This breaks any existing links to the site and any existing links within the site. I thought I had Apache setup to redirect these correctly, but I apparently don’t. Right now most links to this site are broken. Just drop the leading /blog/article and it should work until I figure out why my .htaccess isn’t redirecting properly.

Tags  | no comments