Tuesday, September 28, 2010

The increasing enticement for fraud with multiple insurance covers for a car

I recently got our Honda City insured from Honda Assure (I chose National Insurance as the insurer) for INR ~12,000. I asked the agent if it is possible to give multiple insurance covers to a car. He nodded in agreement, assuring me however that I don't really need multiple covers. Then I asked him if one has the obligation to let each insurance company know that the vehicle has been insured from other insurance companies. He answered in negative, but appeared a bit surprised at my questions.

Anyway, his answers had me thinking. Five simultaneous comprehensive insurance covers (say from Bajaj Allianz, ICICI Lombard, National Insurance, Reliance General Insurance, and Tata AIG) would cumulatively value our Honda City at INR ~2,150,000, and I would've to shell out only INR ~60,000 to get these covers. Not to say that I'm thinking on these lines, but a theft of this car would - in case all five of the companies approve my claim - get me much more easy money than is the cost of even a new City.

This looks odd. If it were this easy, people would've been abusing this scheme for eons. Getting their car stolen and then making insurance claims to the multiple insurers to get undeserved money shouldn't be that easy, after all. I don't know if this is really possible, but if it is, people might already be misusing this scheme to make easy money.

Monday, September 27, 2010

Adaptive keyboards and my years-old idea

A few days back there was an article on Microsoft Adaptive Keyboard in Punjab Kesari (okay, I usually stay away from this newspaper but occasionally have a quick glance through it). Intrigued, I had a read and realized that it was similar to something I had sketched when I was in class 11.

Class 11 was the first time I had my own computer. That was the time when every deficiency that I would feel in my computing experience would give me a new idea. And I would write each of those ideas.

One such idea came to my mind when I played Recoil (by Electronic Arts). You have multiple weapons in Recoil, and you switch between them using numeric keys on the computer's keyboard. Initially, I found it annoying to remember which key stood for which weapon, so I scribbled something like this on my notebook:

There should be a keyboard, each of whose keys is composed of a tiny LCD that can display colors. These LCDs will allow the keyboard to be "dynamic", meaning that the functions of keys can be changed depending on the "context", or the "environment" being operated using the keyboard. For example, when playing Recoil, the numeric keys shall stop displaying numeric digits, and each of the keys shall start displaying the icon of the weapon assigned to that key, making it both convenient and quick for a human to switch between weapons, without having to remember anything (except, of course, the icons).

Although an abstract idea is no match for an actual implementation (even if the initial implementation is only a prototype), it's flattering to see that folks at Microsoft have developed something similar to what I thought of over 8 years ago.

When an American company with no real understanding of Indians tries to market...

...its wares in India using an advertisement, this is what happens:


Whoever designed this advert, did he not know that:
  1. Indians don't know what a QR code is.
  2. Most Indians use phones that are neither smart, nor feature.
  3. Your car is in the low-price segment. Do you really think that potential buyers in this segment own phones capable of processing QR codes (and of displaying rich content)?
  4. Most Indians haven't used the Internet on a full-sized computer. A negligible fraction uses it on their phones (both because the costs are high and the speeds are low).
Whoever approved the idea of adding a QR code in this advertisement doesn't have ground knowledge about India.

An unbeatable benchmark for Web browsers

"I'm pleased to announce the timeless version of HomoSapienTest, an eons-old browser benchmark. More than Kraken, Sunspider, V8, and Dromaeo, HomoSapienTest focuses on only the real workloads. I believe that the benchmarks used in HomoSapienTest are the best in terms of reflecting realistic workloads, and give every other artificial benchmark a kick in the groin." - A parody of text from Release The Kraken, Mozilla Blog

There are so many performance tests available for Web browsers. Unfortunately, none of these appears to honestly indicate the real-world performance of contemporary Web browsers. Different vendors like to report the performance of their respective browsers based on the results of only select benchmarks - ones that make a vendor's browser appear snappier (for example, Apple uses a relatively unknown test called iBench, because it projects Safari 5 as the fastest). Unfortunately, all of these tests merely try to be representative of reality.

But who decides what this reality is? Humans, of course.

No matter what these tests say, a human is almost-instantly able to decide which browser is the faster one, and whether or not the level of performance is acceptable. Let Mozilla beat the Firefox-is-fast trumpet a million times. Let Microsoft beat the IE9-the-fastest trumpet a billion times. Fact remains that - and this can be felt by a human in mere seconds - that IE and even Firefox can be unacceptably sluggish on many machines that are just a few years old, whereas Opera and especially Chrome can provide acceptable performance on such systems.

I'm not sure if there have been research studies that compare the numbers produced by these artificial benchmarks to thumbs-up and thumbs-down ratings given by human users to different browsers. If a benchmark's results can match the ratings given by humans, then it can probably be called a good benchmark.

Tuesday, September 21, 2010

Punjab's corrupt and frustrated traffic police... and how their excesses would be avoided in future

I was fined today on the following two counts:
  1. Red light jumped
  2. Driver not wearing seat-belt
I plead not guilty to both the allegations - I didn't commit either of the offenses. I am 100% sure that I crossed while the light was green. I don't drive without the belt on. But can I prove that I did not jump red light, and that I was wearing the seat-belt? No. Not yet, at least.

It's fairly safe for me to predict that in the not-so-distant future, cars (whatever form they take by then) will include some sort of instrument that mimics the functionality of a flight data recorder (FDR) installed aboard every modern commercial airliner. With real-time tracking of multiple parameters of every vehicle, situations such as the one I faced today will be a thing of past. Using Galileo/GLONASS/GPS/Etc., the device will record a vehicle's precise location at all times during the flight, and a driver will be able to review his past positions before claiming that he did not jump the red light (actually, even this won't be required, as fines will probably be charged automatically, and there will be minimal human involvement in enforcement). Similarly, this device will record whether or not a seat-belt was worn.

The broad idea is that as we move into the future, human decisions - many of which can be plain incorrect - will increasingly be removed from worldwide systems and supplanted by error-free decisions made by machines. Tracking will ensure that the data required for appeals is available (although I don't expect machines to make mistakes fining rogue drivers).

Think and it's already done :(

It has happened many times that I got a thought, one which I deemed non-trivial, and I checked on the Internet, only to find that someone, somewhere had already either penned that thought, or had executed it.

Disappointment. But also an assurance, that my train of thought moves in the right direction.

This happened most recently on 19-Sep-10. I woke up in the morning, and had a thought that Microsoft should tie-up with Facebook to gain access to the valuable Like data building-up in Facebook's databases. Such a move would not only provide Bing with a few days worth of free marketing and elevate the perception of Bing among users' minds, but would also potentially make material improvements - both perceptual and real - to Bing's search results.

I queried on Google, and to my disappointment (and assurance, as said before), there were very recent news stories talking exactly this. Despite implied claims by Ballmer and Microsoft's executives that the relevance of Bing's results is at least as good and perhaps better than Google's, my first-hand experience is that Bing has a long, long way to go before its results start feeling as relevant as Google's results are today (and Google would've become even better by then).

Sunday, September 19, 2010

Comparison of insurance with online advertising

The following thought struck my mind today.

As a post on Chitika Research points out, a very small proportion of Internet users pay for the remaining users (and for themselves). This fact is the bedrock of online advertising, and of all the free applications and content that have become available on the Web.

In contrast, in insurance, most users - who make few or no claims - pay for the very small proportion of users who make significant claims (and this small proportion contributes too). This shows the interesting difference between insurance and online advertising in terms of the proportion of feeder customers.

Friday, September 17, 2010

Microsoft is back to playing lock-in games with Pinned Sites feature in Internet Explorer 9

This post is based on an unverified assumption that the Pinned Sites feature in IE9 is "closed", that is, it's a proprietary Microsoft feature, and cannot be implemented in other browsers.

Sometimes I wonder whether I admire or despise Microsoft Corporation. The company houses brilliant-minded individuals and churns out high-quality products. And the same company also uses evil practices to choke competitors and lock customers into its products. Windows API is perhaps the best example of the latter.

Pinned Sites is the latest in a long series of intelligent, but customer-locking features launched by Microsoft.

Official description of Pinned Sites (Source: Microsoft)

This clever feature - a wolf hiding in a sheep's clothes - is marketed in the familiar customer-friendly tone typically used in Microsoft documentation. But most won't see the predator hidden in this feature. The feature essentially locks the icon of a Pinned Site to Internet Explorer 9, instead of working in a browser-neutral fashion. Windows 7 users using this feature will see icons of their favorite Web destinations in the taskbar of Windows 7, and will click these icons to reach the desired destinations - in Internet Explorer 9.

The genius part of this feature is that it exposes that the Web browser is merely a means to an end. From the user's perspective, the icons in Windows 7's taskbar apparently remove the browser from the equation by taking a user directly to the desired destinations. Also, this feature - I don't know if others realize its importance - allows tight integration of Web based applications with the OS, helping to bridge the gap between Web applications and traditional applications. The evil part of this feature is that it can unknowingly force users to use IE 9. On their parents' machines, kiddies will create one-click icons to the Web destinations of choice of their parents, and the parents will thus forcibly use IE 9. Websites might strike deals with Dell, HP, Sony, etc., that mandate placement of one or more Pinned Site icons to these websites, in exchange for a few bucks per machine, forcing most users who use these websites to use IE 9 (most users try the easiest/quickest methods they can think of, and they can think of only little, and Pinned Sites clearly is for such users).

I don't know if an antitrust complaint will be filed by a competitor (Google, Mozilla, Opera, etc.), but I believe it should be - this genuinely helpful feature should be browser-neutral.

Update (19-Sep-10): I totally hate the fact that this feature allows traditionally-platform-agnostic websites to create platform-specific extensions (currently only for Windows 7). I'm probably not able to imagine Microsoft's long-term goals with Pinned Sites (and other features in the pipeline), but I'm worried that such features might start locking websites to Windows OS.

Tuesday, September 14, 2010

GMAC should convert GMAT Paper Tests into a new book, complete with detailed explanations

After having tried 5 of the 9 ETS-created GMAT Paper Tests sold by GMAC (I possess all the 9), I'm convinced that:
  1. The level of most Quantitative questions in Paper Tests is quite low
  2. A lack of explanations is felt each time you get a question incorrect, but can't find out why (especially in Verbal)
  3. Paper Tests can't be used as a representative of the real GMAT, since Paper Tests aren't CAT-based, and because the level and number of questions differs significantly from the GMAT of today
  4. Despite all of the above, Paper Tests are helpful because the questions are high-quality, the answers are battle-tested, and the problems are real GMAT problems (from the past)
The four observations above lead me to conclude that GMAC should create a fourth book called The Official Collection of GMAT Paper Tests. A lack of explanations make the USD 75 Paper Tests look like a lot, lot less value compared to the USD ~73 bundle containing all three Official Guide books. So, The Official Collection of GMAT Paper Tests should include detailed solutions for every problem in the Paper Tests (including AWA problems).


I'm convinced that this fourth book, if released, has the potential to register significant sales, because addition of detailed explanations for all of the 900+ past GMAT questions will be sufficient to convince potential buyers into purchasing the book. Besides, an orange-yellow book will look great sitting beside the current trio!

Sunday, September 12, 2010

No software version of Google Mini or GSA!

I like both Google Mini and Google Search Appliance. However, what I don't like is that they're available only as standalone hardware devices. That means, no software version to deploy on one's own servers! And that also means possibly-avoidable increase in costs.


Why hasn't Google provided a software version? Possible reasons:
  1. To prevent reverse-engineering of its search algorithms
  2. To provide consistent/predictable and reliable performance
  3. To avoid the problem of piracy

Thursday, September 09, 2010

Are we living with a sub-optimal mathematics? Is it possible for us to design a better mathematics?

I'm no mathematician, but I believe I have a fair understanding of basic mathematics and I can reason decently. Here's a rant about some things I don't understand.

I don't like the following facts:
  1. We have complex numbers, imaginary numbers, and negative numbers
  2. Division by zero is not defined
  3. Tangent of Ï€/2 is not defined
  4. Square root of a negative number is an imaginary number
  5. A finite value for π hasn't been reached
  6. 0! has arbitrarily been equated to 1
  7. ...
Is it possible that fundamental flaws in the basic elements that constitute our mathematics and in the principles that relate these elements to each other are reasons why we have to resort to self-created filler solutions such as complex numbers, imaginary numbers, "not defined", etc.?

I sometimes feel so. I sometimes doubt the need of 'zero'. I sometimes feel that negative numbers should not exist at all (i.e., numbers should only be positive). Not sure, but I sometimes feel that it might be possible to construct a better and error-free mathematics from scratch, in which every operation is defined, and everything is reasonable.

Is the (unadulterated) composition of Earth's atmosphere ideal for life? Is "better" air possible?

The following figure represents the normally-accepted composition of the Earth's atmosphere.


Assuming that the above figure represents unadulterated and unpolluted composition of the air we breathe, I wonder, is this composition the best that can be (for humans)? Can an artificially-made gaseous mixture (with constituents and their proportions different from the constituents and proportions of natural air) beat natural air, in terms of health benefits, longevity, etc.? If yes, how can we conveniently and economically expose humans to "air 2.0", so they can experience its health benefits?

Monday, September 06, 2010

BA's alliance with Kingfisher looks like it'll help BA more than it'll help Kingfisher

A quick look at fares from India to Europe and vice versa reveals why the recent codesharing alliance between British Airways and Kingfisher Airlines will probably help the former more than the latter.

In summary, the alliance allows:
  1. BA flyers to fly to various Indian cities on Kingfisher's Indian network
  2. Kingfisher flyers to fly to various European cities on BA's European network
When BA flyers arrive at BOM, DEL, etc., Kingfisher will fly them to the smaller Indian cities that BA doesn't serve directly. This will provide these flyers with increased convenience in terms of baggage-transfer, booking, check-in, etc., and thus help both BA and its customers.

When Kingfisher flyers arrive at LHR, BA will fly them to other European cities that Kingfisher doesn't serve directly. At this point, it might appear that Kingfisher (and its customers) will get a benefit similar to that obtained by BA (and its customers). However, a quick look at the fares from BOM, DEL, etc., to LHR and to other European cities reveals that carriers of the Middle East usually offer significantly lower ticket prices for flights from India to various European destinations, compared to prices offered by BA and Kingfisher. BA's and Kingfisher's ticket prices usually tend to be higher than those offered by airlines of the Middle East (this is based on the hundreds of times I've checked ticket prices between different cities of the world).

What, if anything, does Kingfisher stand to gain from this alliance? Why will a price-sensitive Indian flyer - and most Indians are very price-sensitive - fly to a non-LHR European destination on a (probably) costlier Kingfisher+BA plan (with LHR as the hub), when a Middle East carrier is willing to carry him to the same destination for lesser money (with a Middle East destination as the hub)?

Saturday, September 04, 2010

One possible future of computing

One possible future of computing will be composed of the following:
  1. ARM-based chips
  2. A lean OS, with itself is silent and stays in the background
  3. An optimized Web browser (or an evolved version of what we know as a Web browser) as the platform for applications
  4. All applications are Web applications, in the sense that they were all initially served from the Web, update off the Web, execute at least part of their code in the Web, and are functionally integrated inseparably with the Web
  5. An application need not be served off the Web each time it is launched. If it can be proved that the latest versions of all parts of the application already exist locally, it'll be executed from the local cache
  6. Most code executes on the Web, and only some code executes on the local device
  7. Applications do not require porting from one OS to the other. Only the Web browser needs to be ported from one OS to the other (thus mimicking JRE)

Wednesday, September 01, 2010

Why is CSeries being compared to 737 and A320?

When Bombardier was developing the CSeries, it had a choice:
  1. Go against 737/A320
  2. Stay away from 737/A320
Now we know what the CSeries is, and Bombardier appears to be in a weird situation - its aircraft doesn't actually overlap with 737/A320 - these aircraft differ considerably in terms of range and revenue seat-capacity - but people all over the world (especially the guys at GLG News) are busy criticizing the CSeries as being inferior to 737/A320. Why are these people even comparing CSeries to 737/A320?

One thing can be said with a good amount of certainty - that with CSeries, Bombardier did not intend to directly compete with 737/A320, because an aircraft with lower range and seat-capacity simply cannot compete with 737/A320, even if it provides a double-digit fuel-burn improvement. Any aircraft that wants to directly compete with 737/A320 must - in general - match or exceed the established players' range and seat-capacity (else it would be in a different segment, invalidating the comparison).

In summary, it appears that Bombardier is in the worst position it can be in. It developed a relatively less capable aircraft to, apparently, not compete directly with 737/A320, and people are busy complaining that the CSeries has less range and seat-capacity than 737/A320. What a pity!