We should not judge people by their peak of excellence; but by the distance they have traveled from the point where they started.
Showing posts with label Futuristic technologies. Show all posts
Showing posts with label Futuristic technologies. Show all posts

Monday, June 8, 2015

kids, zoos, moms, Internet

Yes, the title is a bit weird.

I visited Melbourne Zoo today. I have not visited any zoo since I was a kid, so while deciding what to see in Melbourne City, I decided to pay a quick visit to its zoo.

I spent like 3 hours in zoo. The most common thing around me during this time was kids. They were every where with mostly accompanied by their moms. Moms were trying to teach kids what each animal was, where it was located in its enclosure (since mostly these animals try to hide in some remote corner away from public eye). It was fun to watch them.

Was I excited. Not really. Why? Well, I think because as adults we tend to watch a lot of Discovery, National Geography, where animals are show cased in plenty, so their attraction does not stay that strong. You see them in enclosure and you think..... whatever.

But, yes, for kids its a best experience as they are yet learning and their cognitive abilities get better by being in physical presence of these animals than just seeing them on TVs. So Zoos still have their place in the modern society. The question is, will their place be ever replaced totally by the advent of the information age, where every thing is available on the finger tips using Internet and related technologies for e.g. such as virtual reality? Zoos have to keep on re-inventing themselves to draw bigger crowds, they have to keep on making themselves attractive so that parents want to bring their kids.

Or was I being too nerd? Have I lost my ability to enjoy simple things, because I try to reason behind why this thing is happening the way it is happening? I think yes. 

Human driven and self driving carI s

I am in Melbourne, Australia to present my research paper. This is the top-most conference in data management. The paper presentation went very well. After the conference usually I spend couple of days doing local sight-seeing.

I drove from Melbourne to Sydney after the conference was over. Its a distance of around 900 km. I was seeking an adventure and decided to drive down than flying as that would give a perspective of Australian landscape, culture, driving habits, roads etc. Road trips teach a lot of about the country,
as that forces you to interact with locals and gives an impression about the overall culture.

Driving itself was a lot stressful in the beginning as I was not in my best shape due to lack of proper sleep, and on top of that constant rains that made road visibility very poor. Also I was not much confident of driving alone in a new country, in new conditions. So the initial 2 hours were a lot stressful. But as the rains vanished and I got used to the driving it started getting better. But the initial feeling was quite stressful. Suddenly you realize that you are responsible for extreme precaution as a slight mistake can cost you a severe accident and your life. Having that kind of responsibility while surrounded by zooming cars, and pouring rain could be immensely stressful.

While driving I was constantly appreciating the advent and prospect of self driving cars. A lot of daily stress in human life arrives from driving. Since the person has to take care to not to make any mistake. What if this exercise is made enjoyable and relaxing where a machine is able to take care of all the necessary precautions etc to make sure you are being driven safely. I think when the self driving cars, or cars that have abilities to correct or warn driver of possible accidents becomes common place, quality of human life will change for ever. These kind of technologies are already present in expensive cars, but yet to appear in cheaper ones.

I constantly felt how great it would have been had my car had all the safety mechanisms to warn me in case there was a possibility of accident. My stress would be much less. Also I realized, I take decisions during driving such as which routes to take, which exit to follow, how much speed-limit to follow by watching the road signs. If sensors could replace me, they should be also be able to do the same. The only problem that remains is the exceptional conditions when human intelligence can take judgement.

My journey went on fine after I got used to the driving and after some power naps by resting by road side stops. But, the thought of having a self driven car taking me from point A to point B was much pleasurable considering the amount of stress driving could involve some times. The time could be better spend in more constructive things, to make some things more better. 

Sunday, January 18, 2015

Crowd intelligence vs Expert intelligence

Out of my recent travels with my brother in a rented car arrived lot of scenic photographs. At least for last 4-5 years, I have not done so much extensive outdoor travel that I did in last 1 month. A friend suggested to put these photos on www.500px.com than flickr, so I tried it. 

(The link for photographs is here) www.flickr.com/mrunalg

500px is a market place for photographers where fellow photographers decide on the quality of photographs by rating it, and the algorithms decide whether a photograph is worthy enough to be moved up in the ranks so that after a while it could be put on the stock photography market place to sell. This model is more like the Reality TV show setup, where audience decides whether a particular singer is good and vote for him. More votes decides the champion along with a few other factors. 
So 500px essentially brings this aspect of reality TV kind of voting in the online photography world to decide the best photographs. 

I decided to test their algorithms and how good the overall setup is, so I experimented by putting in different kinds of photographs hoping to see how they get rated etc. I was quite disappointed with the overall results at how even the best of the photographs might not appeal to crowd. And how the entire logic works on how well you connected are, how many people in your network view the photographs etc. So the crux to mention is, the popularity has nothing to do with the quality of the photo. I have had similar problems with the occasional viewing of the reality show setups where many times the singers I feel are great, never win. So that made me think would the setup be good if only experts were judging the photographs?

I submitted some of the photos to high quality stock photography sites such as www.shutterclock.com. These photos were out right rejected saying they lack correct exposure, lack of focus, noise, lens artifacts etc. 

I then thought of another process to review the submitted material the way it occurs in the academic publishing world where peer reviewing is the norm to get research published. Depending on the quality of the conference / journal you are trying to publish in the reviews could get so nasty and critical that many times one feels like giving up. Writing is a very interpretation oriented skill. People look at it from their own biases, expectations etc. So chances of getting an article in, in a top publication through reviewers is even more tough. 

So now we have seen the two extreme sides of getting your work to the forefront. Crowd intelligence vs Expert intelligence. And both have its drawbacks and advantages. 

The most interesting question in this aspect then becomes how to combine the two in a kind of hybrid way so that the best of both the world could be obtained. And this is a tough question. The beauty of Internet is it has democratized the way information gets accessed. So now crowds can have their say in choosing the outstanding work unlike the earlier times, when only a few elite experts were allowed to do that. But there should also be a way to let the two modes integrate, so that we see the best possible outcome, because some times crowds might not be able to judge perfectly, and some times, the experts might be too harsh. 

Balance is the crux of everything and it needs to be present in the world of Technology too. 

Thursday, June 13, 2013

Raffaello D'Andrea: The astounding athletic power of quadcopters


How would the future in the machine world look like? 

A great demonstration of power of quadcopters, and their athletic abilities of balancing acts .... Sci-fi fiction just got real.



Monday, April 1, 2013

$30 million, 17 year old guy, and the Yahoo acquisition

17 year old whizkid's company gets acquired by Yahoo for $30 million. Wow. Nice. I read this news first on Economist.

http://www.economist.com/news/business/21574497-yahoo-buys-teenagers-start-up-youll-never-work-home
My immediate reaction was, the technology that goes behind such kind of a product is impossible to be at the hands of a 17 year kid. What is the real story? This technology needs PhD level knowledge.

A quick search about the technology revealed that this boy was backed by www.sri.com, a silicon valley non-profit r & d firm. Here is a link to the company web-site and what it does.

However, for a layman the news remains as flashy as told by the news anchor or by the reporter of the story, that, a 17 year created a technology which was acquired by Yahoo. In the articles that I read on Internet I never came across name of sri.

This is what I call news without verified facts. However, who cares, because this is how the business gets done. Steve Jobs gets all the credit for all the businesses he started, no matter, he did not invent the technologies behind many of them. (By the way sri is also the place where computer mouse was invented, they licensed it to Xerox park research center in Silicon Valley, and Steve Jobs being the business genius he was, saw the great potential it hold in revolutionizing consumer PC, and brought it to the market and the rest is history).





Sunday, November 11, 2012

Clouds in a Room ...


I am always amazed by clouds. Especially in Netherlands where the weather is always cloudy, that seeing blue clear sky with sun, is a treasure. While returning from Istanbul from a conference and watching clouds float in the sky from the plane. I imagined about clouds, the science behind clouds, and a lot of other related things. I wrote some of those ideas on my plane napkin. 

One of the prominent idea I imagined was, what if I can create decorative small clouds using some small machine (like a fog machine) in my own room.. Will it be possible? What opportunities lie in such an idea. 

For example, lets say, there is a children's birth day party and I want to    simulate a room filled with clouds instead of Baloons. What would it take to make clouds in a room? Can I make thunderstorms happen in those clouds artificially? Some wild creative imaginations.  


Today I was delighted to see the exact thing I thought counted amongst one of the best inventions by Time in 2012. Here is a picture. I do not know the authenticity of the claim below. The feeling that I get ideas which are worthy is fine to feel happy :-). 


Saturday, June 30, 2012

Food, hunger and studies

Hunger and food are a delicate issue for any graduate student.

As an American graduate student, I used to crave for free food such as Pizza, sandwitches that used to get served in seminars. Seminars also hosted free food to guarantee an audience to fill the rooms. Thus it was complimentary for both hosts and attendees.

I have always wondered how easy life would be if one could satisfy ones food needs without much efforts. Food is considered one of the basic needs for life. What if there is an amazing answer to let humans live off without having any food for days like some animals for example, beers do when they hibernate in winter.

Huge amount of defence research gets done to prepare food bars which would stuff maximum energy in minimal quantity which soldiers can carry with them when on mission in remote areas and toughest environmental and geographical conditions. Astronauts are provided with special kind of food to let them sustain in space. 

General public at the most gets the energy bars (Cliffhanger brand from US) that can help make one feel full for some time. But, by far a good solution does not exists. Is there a market for such a product? What if there are pills for different type of food items which would make you feel full and provide the body with enough energy that would have obtained by eating normal food. Worlds food problems could be solved and there would be much peace. But considering the fact that it takes 15 years for a drug pill to get introduced in market, after its first envisioned and the amount of research money spent (up-to 1 billion dollars) on these drug pills. Creating an artificial food pill looks a herculean task.

Who would eat such food because after all many people live their life so that they can eat to fullest and live to enjoy :) I would be the first client if such a pill would ever come in existence.  I would love to eat a pill which would make me feel full when I am working on some deadline, and without having enough time / will to prepare and eat food.

I wish. 

Friday, June 8, 2012

Technology, knowledge, talk and Implementation

Technology has made a lot of information available to people in a very convenient manner. The more information people have, the more knowledgeable they become, and the more topics they have to discuss and debate.

However, one strong observation that I am having is, these debates need not result into actions most of the times. People are happy to talk about things at length, without taking conscious efforts to change things. The pleasure of intellectual debates pleases most people, with an intense feeling of satisfying intellectual hunger to certain extent. We see academics debating on complex issues, research papers being written to length, committees being formed, but implementations many times lack the same rigorousness that is seen during debates.

What is it that can make people take action and implement instead of just having debates and discussions? Implementation requires discipline, persistence, hard work, and motivation. Is this where humans differ from robots, that humans can think and everybody thinks differently, where as robots are programmed and will do what one is asked to do.

Can humans be programmed to act like robots in certain circumstances? Imagine, 4 people having a strong debate on some complex issue, and after the debate is over, they actually implement the things that they strongly debated about during discussion. And this pattern is repeated, across every discussion. A kind of robotic behavior, where pattern is repeated  multiple times always. The productivity will be huge and so will be the outcome of such discussions.

Knowledge will be useful in the best possible manner, if some thing of this type could be achievable for solving certain complex problems. The problems which have eluded mankind because of lack of consensus and lack of willingness to implement the solutions. 

Thursday, April 5, 2012

Behaviorial aspects of technology

Experts from Yvette Subramanian of CITRIS.

There are a growing number of persuasive technology apps and systems that aim to make individuals (and therefore society) healthier, greener, more financially sound, smarter, happier, and so on. Analysis of such technologies often focuses on the aggregate results, for example the total percentage of residential electricity saved across households. However, an aggregate focus misses important aspects of the behavioral responses at an individual level. This talk focuses on experiments and models that examine behavior at the level of the individual, and demonstrates how such analyses could guide the effective design of behavior change technologies and better predict their impacts on different users' segments.


Sunday, March 4, 2012

Global Indian company vs local Indian consumer

While reading different articles on Economic Times, I could not stop myself thinking about the different ways conditions in India could be improved for the common man ....

For example. When I read this article by Sunil Mittal about how he wants to replicate the success of "minute" calling in "Data revolution" in India, I felt emphatic. I felt happy because some day, I could imagine myself roaming around in streets of India and using Google Maps to find out in real time, the latest happenings around that street. A pure silicon valley dream, however, if we have enough data available in India, this dream should not be a distinct dream.

However, as I turned to read the comments on the article by various readers, I understood one difference immediately. In reading the article, I was looking at the global picture. The local picture was totally different. Go to any grass root level and there are enormous level service problems. From the irregular infrastructure, to in-adequate trained staff, to third class customer service. The real problem a customer faces is huge. 

So here is the next thing that I thought. What would be the best way a virtual customer service be provided. At present we have the concept of consumer forums, customer service departments, etc. where complaints for the services could be launched. However, most of the times, due to sheer number of complaints and lots of problems in the process, a customer still feels helpless and is totally clueless about whether he would get any justice at all.

Could there arrive some business model, which would provide guarantee of quality of service by linking together, the accountable company who is supposed to provide the services, the consumer who is provided with faulty services, and the intermediate third party who guarantees this service. Could technology be of any use in this case?

For example. The article that I read, itself could speak for itself. There is the article, which talks about hopes of data growth and there are consumer negativity. Could reaction to such articles which are also read by the company people be directly propagated to masses, to generate a support about inefficiency....blah blah blah...I know some private companies trying to do this, however this model could be more fine tuned to understand how it could be made to work in a better manner. 

Some time soon, I would wish to see a happy Indian consumer.....

Saturday, January 14, 2012

Teaching machine....

Courtesy article from famous behavioral scientist "B.F.Skinner".
Teaching Machines and Programmed Instruction

Skinner's children were growing up. When the younger was in fourth grade, on November 11, 1953, Skinner attended her math class for Father's Day. The visit altered his life. As he sat at the back of that typical fourth grade math class, what he saw suddenly hit him with the force of an inspiration. As he put it, "through no fault of her own the teacher was violating almost everything we knew about the learning process." In shaping, you adapt what you ask of an animal to the animal's current performance level. But in the math class, clearly some of the students had no idea of how to solve the problems, while others whipped through the exercise sheet, learning nothing new. In shaping, each best response is immediately reinforced. Skinner had researched delay of reinforcement and knew how it hampered performance. But in the math class, the children did not find out if one problem was correct before doing the next. They had to answer a whole page before getting any feedback, and then probably not until the next day. But how could one teacher with 20 or 30 children possibly shape mathematical behavior in each one? Clearly teachers needed help. That afternoon, Skinner constructed his first teaching machine.

Skinner's first teaching machine simply presented problems in random order for students to do, with feedback after each one. But this machine did not teach new behavior. All it did was give more practice on skills already learned. Within three years, however, Skinner developed programmed instruction, where through careful sequencing, students responded to material broken into small steps. The steps were similar to what a skilled tutor would ask of a student working with one student at a time. The first responses of each sequence were prompted, but as performance improved, less and less help was given. By the end, a student was doing something he or she could not have done at the beginning. For about ten years, Skinner was caught up in the teaching machine movement, answering every one of thousands of letters from parents, schools, and business and industry. With a grant, Skinner hired James G. Holland who with Skinner's supervision, created The Analysis of Behavior for Skinner's class of Harvard students to take on a mechanical machine. (There were no microcomputers yet.) The field of education embraced this newest teaching method, but many of the materials were poorly written and no company wanted to design materials for a teaching machine that might go out of production. So most programmed instruction was put into book form. But the a book does not maintain the contingencies: Students can look at the answer before writing their own. By around 1968 education publishers stopped printing programmed instruction. That same year Skinner published The Technology of Teaching, a collection of his writings on education. Some of the better programs from the 60's are still used and with the coming of the computer and internet, the perfect machine that Skinner lacked is now available. Increasingly, instructional designers are realizing that, as Skinner insisted, tutorials must do more than present blocks of content with quizzes at the end. Effective instruction requires learners to respond to what each screen of information presents and to get feedback on their performance before advancing to the next. In addition, the sequencing of steps is critical.

Sunday, September 25, 2011

Technology and Poverty line in India

The role of technology in deciding the poverty line in India at least shows some improvement in using infrastructure such as Tablet computers to collect data. I got an opportunity to attend one such function through a new friend I made today through a mutual friend.

The discussion was held under by the socialist movement of India. Government of India is planning to have a country wide census to decide the poverty line and determine the families living before poverty line. Rs. 32 per day is the income criteria decided for poverty in Cities and Rs. 26 is decided for rural poverty. Are you kidding me?

This discussion was about different strategy formations to make concerned representatives of people from various Indian villages aware about their role in this activity and make them understand the flaws in the way definition of poverty is defined.

I do not want to comment out on the ways the definition of poverty is defined. This is a huge propaganda to make another show off by the Indian government without any sincere attempts behind it. However, some striking positives out of this meeting that I noticed were.

1. One union leader stressed on why Government is not considering educational qualification and only insisting on economic qualifications.

2. How Government is going to verify authenticity and not manipulate the data.

3. How people's involvement could be made more such that they could be able to assist better in this survey.

4. How there is a need for people to be aware of the potential harms of manipulation of such schemes and how people should prevent it, by stressing on correct corrections. An interesting suggestion was, "Instead of counting number of poor, count the number of rich, and let them defend their richness and then declare all the rest of them to be poor".

5. Make government aware that food, healthcare, education are basic necessity of anybody and not just poor and the government should try to shed off its responsibility by trying to decide number of poor based on some nonsense norms, than actual facts and then get involved in corruption.

6. Role of technology such as tabloid computers in collecting information.

Well, this discussion made me aware about how different people at low level think and discuss when government introduces new schemes, how they are totally helpless when it comes to implementing the feedback provided by them to government, How Maharashtra government is a total moron and no will at all to implement any schemes with sincerity as compared to rest of the state governments in India, what are the different ways in which people could be made aware.

However, I strongly felt, most of the points which were stated were well known to any person who is well aware of these problems and why they exist. The discussion lacked any effective mechanism by which these things could have got implemented, the suggestions could have reached the government and their would have been a successful channel to propagate these discussions to grassroot people.

Technology has a huge role to play in such discussions. Community radio is a great idea that could help in having such discussions across different sections. However, a better scheme to have is make people aware the incentive behind attending such functions. People are selfish and if you are able to tune up your motivations into correct incentives which hit well with the people's moderate ambitions to make their lives a little better either in terms of economic improvements or materialistic comforts, people would turn in hoard for such must discussions. Well, that is what politicians in India do. They provide economic incentives for people to vote and people react. I wish our activists have a vision to make some logical tacticts like this to attract more crowd and make them dedicated to a cause, by making them realize its importance for them in the long run.

An interesting idea came to my mind in deciding about poverty of a person. Can a person by identified as poor by looking at his picture? For example. A poor person would be thin, he will be underweight, he will have dark skin because of constant labor etc. So there are some points in a person physical characteristics that could decide whether a person is poor or not. Calorie indicator is one such method of measurement for poverty.

So, here is an interesting question. Lets say I provide you with 100 digital photographs and ask you tell me who is poor amongst these? Could you do that solely on the basis of picture? I believe yes to some extent. There are sophisticated image processing algorithms, which could decide different parameters by taking a look at the person, for example, height of the person, ratio of various stats etc,. This is a hugely complicated problem, however I believe with more and more fine grained parameter control, such an automated process could be done to some extent.

Its a huge irony that in a country like America women activist take protests to get a right to bare their body in public citing reasons that male are allowed to go bare chested, then why not us? While in a totally different part of world like India some women do not have proper clothes to cover and could not find people to support this cause. Huge contrast across different sections of society across world.

Tuesday, September 20, 2011

Mobile based jobs

Could mobiles be used to generate employment in developing countries? This is an interesting question.

Mobiles by far are used as the means for communication in the present context. Right from an top end executive to a low end construction worker owns a mobile now. And the penetration in the low end section of the society in India is going to increase in coming time.

A crucial question to ask is. Could this vast technological device be used to generate legal employment. Could low end micro-tasks be generated, distributed and completed on mobile phones? If yes, what type of tasks could they be? Who would generate them? What kind of payment method be employed for such type of jobs.

This is a hugely interesting question and as mobiles become more powerful mobile based employment would start picking up slowly. A key question could be, if such jobs are available, would the low end people be willing to work on such jobs in temporary time? For example. imagine an auto driver instead of chitchatting with his friends in his idle period of waiting completing a temporary job on mobile and earning some income.

Interesting proposition and I feel slowly this type of trend will pick up. There are already some moves in that direction.

Tuesday, August 16, 2011

Jobs in Internet Cafes - Amazon Mechanical Turk - Low skilled workers

Will people be willing to work on low skilled jobs on Computers / Mobile phones if enough payment is provided to them? I recently finished off with this project and results were very good.

Amazon Mechanical Turk. www.mturk.com is a marketplace where low skilled jobs are put for people to work. These jobs are categorised mainly into three types
1. Transcription type jobs.
2. Article writing type of jobs
3. other menial jobs.

The payment varies from $20 to $.01 depending on the complexity of job. It was found that these jobs are not much popular and not picking up fast. Why is it so? After a good research the answer was found to be in the User Interface of such jobs. Low skilled people are too un-techsavy to understand fluent use of the interface provided by mturk.

The next question that could be raised was. If enough payment is provided, will people be willing to work on similar jobs if a good user interface is provided to them? How much is enough? We designed jobs which will pay users around Rs. 100 per hour.

We tried answering this question by running experiments in Internet cafes in India and Africa for around 2 months. We designed our own setup, strategy, user interaction, data collection and analysis. The results were encouraging. Different places displayed different trends, and users had different expectations. It was good to mingle with low end businessmen, low skilled people, and observe their behaviour.

This research will be published in the research track for "Technology for emerging countries".