Science & Observation

This is how science works:
1) do experiment  & observe the results
2) develop a hypothesis which explains the results
3) done...until someone observes something that cannot be explained

What if we cannot do the experiment? Like switching from democracy to something else? Or making drugs legal? etc. What if the duration of impact of the experiment spans lifetime of an individual and makes it impossible to observe. Any experience whose patterns "size" is bigger than the amount of time we spend in that experience would remain a mystery for us. While stuck in a traffic jam we can wonder what is causing it, but we cannot really know what happened because we will only be able to see anything when the jam is over.  Does a software rewrite makes sense for a startup? If the startup runs out of funds before completing the rewrite or never does a rewrite because of fear that it will run out of funds, the experiment remains incomplete.

The only way to understand such patterns is via communication.  I am using the word communication in the extreme sense. Not restricted to just people talking, but newspaper, history, journals, blogs, tweets, etc. Don't know how theory of relativity looks when we add the extra element of communication between observers into it.

Agriculture

Agriculture is a high risk business. It is probably also low returns business. And most of rural India does agriculture for living.

1. Loan from Government (Private Banks don't give loans to small farmers)
2. Weather predictions, untimely rains
3. Lock in, can only grow one thing at one time
4. Access to electricity, water, quality seeds,  fertilizers, etc
5. No access to global demand/supply patterns. No collaboration.
6. Intermediaries involved in final sale. 
7. Transportation costs, storage costs.

So many things can go wrong in a single cultivation cycle and even if that doesn't happens other conditions  can cause losses. Every (few) years, government ends up waving off the loans to support farmers. Diesel prices are kept artificially low to help farmers. No income tax on income from agriculture. In Punjab government gives free electricity to farmers. It is magical at in spite of all these measures, farmers continue to live miserable lives. Agriculture is an important and big business and yet the farmers who takes all the risk and government pays all the loans from tax payers money and it is the intermediaries who make all the money.

I believe we can do this better. Agriculture Micro VC - AMVC

1. Like any high risk business, agriculture should be driven by investment and not loan. AMVC invests in farmers every season. The funding depends upon final share of the output. If the final produce is less, AMVC bears the loss. If produce is good, both farmer and AMVC makes money. 
2. To make it high return business, we need AMVC who could operate at a large scale to cut out the intermediaries. The difference between price to farmer and price to final consumer is anywhere between 3X to 6X. AMVC should represent its member farmers to operate as a BIG farmer who could have much more negotiating power than individual farmers. Moreover the time of harvesting could be decided subjected to deals/sales to cut out the storage requirements and unnecessary transportation. 
3. AMVC should have access to what the local/global requirements are can control what is cultivated and in how much quantity. Every body cultivating the same thing, only brings prices down and losses for farmers. By helping farmers decide on what to cultivate based on what others are cultivating, farmers can cut down their losses because of high supply/low demand problems.
4. Since it is AMVC whose money is at stake all the time, it becomes its responsibility to do everything necessary to make sure that farmers are successful, because that is the only way for AMVC to be successful.

I don't know the specifics of the agriculture industry, but I feel AMVC model can help in uplifting farmers and reducing the risk burden they carry all the time. 

Probabilistic Voting

Voting based on majority has one problem. Majority wins, always.

This is not a good thing as the whole idea of voting is representation. Majority based voting will always cut out the minority. Instead, consider probabilistic voting. Instead of making the decision based on majority, if the votes only decide the probability of decision, then the final outcome can be anything except that majority decision will be chosen with high probability. This ensures representation of all classes and over a reasonable number of decisions each class gets probabilistic representation.

e-Voting

Over the last few years we have seen use of IT in the voting. We now have e-voting machines. Some people also talk about online voting and then their are tons of debates about security of such mechanisms.

Lets assume for a while that security concerns will be addressed sometime in the future. That just leaves us with the same old democracy just happening online. I think the next level of democracy should not be limited to just "automation" of the voting process. Technology can play a vital role in defining what democracy of the future looks like. Think of why we have to choose our representatives and why only once in five years? It is a technical problem. You can't have a billion people come together and sit under one roof and decide about their future. The machinery was designed to solve this very problem. family, small villages and corporate boardrooms don't have representatives, they don't need as all stake holders can fit in one room.

In next four five years everyone in India will have a mobile phone (if they don't have it already). All reality shows are doing voting using SMS to choose their top stars and making money along the way. If people can vote for their favorite starts every week, why can't they choose their representatives every week. Why can't they suspend their vote for the representative for a week and choose to participate directly in the law making.

Today many people in India will be willing to cast their vote in favor of a candidate for may be Rs 500 - Rs 1000. Since a single winner is elected per constituency, depending upon number of candidates in election, all you need is may be 10%-20% votes to get elected. For a five year term, per day cost of vote is less than a rupee.

What If:

• Their were no elections
• You could vote using SMS, any time of the day, week, month, year
• You could change your vote anytime
• You could split your vote 
• You could loan your vote to someone who you think can better decide how you use your vote
• If their was no parliament, everything happens over Goto-Meeting and as many people who want to participate directly or through their representatives could participate. 

Basically the whole election process becomes flexible in terms of time (duration of being a representative) and in space (number of representatives). Think of it like a stock exchange for politics. A true representative of public's political sentiments.


What do we get:

• People who sell their vote, get a better price throughout their lives
• People are always in control, politicians are always on toes
• Much better participation of people. Everyone can find time to send one SMS (from anywhere)
• The whole disconnect of "people" and "government" goes away as people are government - all the time. 
• No need for nominations. Anyone can make anybody else their representative and let them choose their representative or participate themselves. 
• If you want to do good for the country, convince people first

Taking Input - Program/UI

I went to pay the electricity bill at BDA. The machine refused to read the bar code on the bill, so I choose the manual method. I had to fill about seven or eight fields. I couldn't find three of them on the bill and gave up. I am sure the bar code doesn't encodes all of inputs, but possibly some unique number which can find out all the eight fields that the machine wanted me to enter. If electricity department could print the number represented by bar code in decimal digits on the bill itself, I could have easily paid my bills and had to just enter one number.

Passport form is another example of the same problem. You need to write your address at least 4 times at various places. Simple website would go a long way.

Any website which wants to know your country will invariably show you a dropdown  with more than 200 entries. Auto complete would be so much nice to use.

Car is a great example of good user interface. Steering is large enough to control turn, only five gears, not two and not twenty, just right, clutch, brake and accelerator. Minimal and complete. It takes couple of weeks to get used to it, but after that it works well. The feedback is immediate and speed/fuel/rpm monitoring is right in front. The big horn to shout is right on the steering, along with indicators. Each of the controls are of a certain size and at a certain distance from the driver and I bet that is the metric of importance/frequency of use of the control.

I wish someone writes a UI framework which can use user feedback (not another form but by logging what user does with the UI) to make it easier for the user to use the UI.  It could hide the features which are not used, keep a cache of recently used features, increase/decrease size of the buttons (clickable, touchable area) to make it easy to click or may be allow user to create a shortcut. Microsoft Office does some of these things and I guess others can do it too and probably do a much better job of it.