Previous month:
October 2014
Next month:
December 2014

November 2014

Google's Creative Lab

Last year, Google ran an interesting experiment in collaboration with the Royal Shakespeare Company (RSC): it used Google+ to attempt to redefine the theater-going experience using an online stage. It was called Midsummer Night's Dreaming. You can read about it here and there. It sounds a bit expensive - especially the part that involves new characters - to be easily replicated, but (I think) at the very least opens thought-provoking avenues to complement projects such as "theater/opera/dance at the movies". With 110,000 unique visitors the weekend of the project and an increase by over 700% of the number of followers of the RSC Google+ page, the project was deemed a success. 

Google hit a home run with Gmail many years ago. I also liked Google Reader, but that was discontinued. Google Analytics is also a great service that receives little attention in the mainstream media and has been around for a while. So what sort of services has Google put forward in the recent past? Google Drive has been overshadowed by Dropbox. Google+ never turned into the counterweight to Facebook that some people (including me) hoped it would become. Google Glass will remain for nerds until it is combined with statement-making glasses or sunglasses, if that ever happens. While Apple has been able to repeatedly bring innovative products to market, Google has been closer to coasting on its reputation while securing singles or doubles after the home runs of the past, although some of its marketing campaigns have been very favorably received (see here and there, although you have to wonder if part of the hype isn't due to some of those writers' desire to be hired by Google.)

It's true, though, that Google won the 2012 Print Advertising Competition. Maybe they've become better at marketing than at putting out new, groundbreaking services. Or perhaps Google+ is carving out a niche for itself in the interface between art and technology: another interesting project is the partnership between Google and Tate Modern, This Exquisite Forest. From the Tate website: "Taking as the starting point a series of short animation sequences created by artists represented in Tate’s collection, users of the website and visitors to the installation are invited to draw and animate new sequences and thus continue the ‘seeds’ begun by the artists.  As more sequences are added, the videos dynamically branch out and evolve, forming multiple new visual narratives." 

Below are some excellent tips from Google's Chief Creative Officer Robert Wong to further one's creativity, courtesy of Fast Company.

I also loved his talk for the Future of Storytelling summit. This video was posted by "Future of Storytelling". Please upvote the video on YouTube if you like it!

Google needs to do a lot more about Google+ (or any other service it is currently working on) for it to stop the perception its best innovations are behind it. (If people liked wearing glasses, contact lens manufacturers wouldn't be making so much money...)  One way might be to reposition Google+ as an online venue for ephemeral discussions in response to an event or talk or video. I would imagine, for instance, that it would make a great tool for discussions after NTLive theater broadcasts (broadcasts of live videotapings of certain London plays over the UK and the US) or GlobeOnScreen (a similar idea for Shakespeare's Globe company). But it might need far more sophisticated tools - especially comment display tools - than what is currently available. As an example, you could have a situation - in the "chat room after broadcast of theater play in movie theaters" scenario - where people type comments into comment boxes and the comments that are up-voted appear in bigger font, gathering more attention (similarly to having someone at a party who a lot of people gravitate toward). The comments would show up on the whole screen so that they would be arranged two-dimensionally instead of being stacked on top of each other. My feel is that the comments would need to be written, because Internet connections might not always support video chat involving so many different people and tech glitches would be more frustrating if the whole system relies on video. But I might be wrong.

Then you could arrange replies to the upvoted comments as spokes to the "hubs" made by the main comments. Some layers might need to be hidden if the comment chain reaches a certain length. The user could click on another comment hub to participate in another discussion. (I guess you'd need a system to sift through the comments quickly to see if someone else already made your comments, although repeating other people's point doesn't seem to stop most blog commenters today. But if you want to connect with other people who share your opinion on a given show, you need to be able to find them quickly.) Providing a forum for Google users to interact as a one-time-thing based on a shared experience like theater or a music event, where users would presumably remain civil to each other, using cutting-edge visualization tools for those comments, might not be transformational on a large scale, but could be an idea worth pursuing.

Congrats to Dr. Yang Dong, who won Honorable Mention... the 2014 INFORMS Financial Services Section student paper competition, for work I supervised as Yang's dissertation advisor. Other finalists were from Columbia University (two students), the University of Illinois at Urbana-Champaign and KAIST. I'm very proud of Yang for her great accomplishment. You can read more about her work in this previous post of mine.

Below is a picture of Yang with her certificate. Wishing her best of luck in the next phase of her career as a Senior Analyst at JP Morgan in New York City!


Operations research papers

Here are two research papers I wrote with my then-PhD-student Dr. Elcin Cetinkaya that might be of interest to my operations research readers. One is on a data-driven approximation algorithm to portfolio management with quantile constraints. While the algorithm is very simple, it performs very well in practice. (Calafiore considers a similar problem from the perspective of finding the minimum sample size that will provide certain probabilistic guarantees to the manager, in a setting slightly more general than ours.) Our paper is the latest version of the report based on our ISMP presentation back in 2012. The other is on new product launch. The key contribution of that paper is to incorporate robust optimization to the parameters of the Bass diffusion process driving innovation adoption, in a context where the decision maker seeks to maximize his Net Present Value. For tractability purposes we enforce that the uncertain parameters must take either their nominal, best-case or worst-case values. While this requires the use of integer variables in the uncertainty set, which precludes an immediate use of strong duality as is traditionally done in classical robust optimization, we show that a total unimodularity property of a parametrized version of the problem allows us to obtain tractable reformulations of the master problem.

More from DDI2014: MIT Media Lab, Hyper-Efficiency Units and Hyper-Compact Cars

Here is another post about DDI 2014 (check out my "Design vs Innovation" post for the first one). Before I begin, a disclaimer: I did think DDI 2014 was thought-provoking. I really did! It just happens that the topics that most caught my attention did so for bad reasons. In the present case, I'm referring to a presentation that was both on hyper-efficiency units and hyper-compact cars. Let's begin with hyper-efficiency units: tiny, transformable spaces that are made more livable thanks to robotic architecture. The motivation, as presented in the talk, is that cities want to make sure creative types and young graduates can find an affordable place to live within city boundaries instead of being pushed out by gentrification and similar forces. So far so good. But what the team tried to do is to fit as much as possible in the legally minimum livable space in New York City (although Kendall Square was also presented as a possible application area), which the slide said is 29 m^2 or 312 sq feet. Through tech devices that make tables and beds come out or fold in, the space can be made to seem larger than it really is.

Here is what I didn't like about this. The real issue in NYC is that the inaffordability stems from the ever-growing salary gap between Wall Street finance folks and the rest of NYC residents. Trying to mitigate the fact that their ever-increasing salaries allow landlords to extract ever-increasing rents from their tenants by making the tiniest possible space more livable (assuming the tech sensors and robotic arms and all that don't break down, although if it is the landlord's responsibility to fit the space he will surely add a hefty premium for that, and if it is up the tenant he/she probably won't have the money to splurge) seems like putting a Band-Aid on a very serious problem that will not be fixed by such measures. People move out of the cities because, once they get out of school, their idea of affordable housing involves more than staying in an apartment the size of two parking spots. (Plenty of NYers and NJers moved to the part of Pennsylvania right behind the state line so that their money could buy a moderately-sized house with a yard and fairly good schools.) You can put lipstick on a shoebox, but it's still a shoebox. 

The other thing that annoyed me was the description of self-driving mini-cars (that basically have neither a backseat nor a trunk) as the solution of the future to help fit more cars on parking lots. I don't remember what they were called so I'll call them hyper-compact. The speaker explained they have to self-drive to make sure they fit like sardines (from the French expression, "etre entasses comme des sardines", not sure if it carries over in English... imagine the feeling of being piled up in economy class in a long-distance flight and you get the idea) on the parking lot. The mostly young or youngish attendees in the audience were captivated by the demo video of the hyper-compact cars as the future method of choice to travel in cities.

But the thing is, people come to cities from elsewhere too. (Even Cambridge. You just have to sit for 25 minutes after the toll at the Cambridge/Somerville once you leave the Mass Turnpike because everybody is trying to merge on River Street to convince yourself a lot of people come to Cambridge from elsewhere too.) This might be difficult for hipster-creative-types-let's-bike-everywhere-gas-is-bad-do-we-have-a-medal-for-awesomeness-yet-we-are-so-much-better-than-everyone-else, but some people really do need to be driving their cars or pickup trucks for their work. (I drive a small car with excellent gas mileage. Blame someone else for ruining the environment.)

As much as, coming from Europe, I don't particularly care for SUVs or pickup trucks, if your work requires you to go to work sites or travel a lot in parts of the country where the roads can be really bad - and yes that includes New England - you really don't want to be driving in something that looks like the front bench of your truck wrapped in a glass cartridge. So what are you supposed to do? Leave your truck outside the city and pack all your gear and tools and material in "the cartridge"? Yeah right.

(The speaker did say the emphasis was in city building because apparently cities are where the cool things happen, or more accurately most of the population is supposed to be in cities. When I go to Paris I am amazed by the crazy illegal spots certain residents leave their Smart cars. But you can live and work in Paris without ever leaving the city or needing a car. That is not a situation you encounter too often in the US. Maybe it wouldn't be a bad thing to investigate new frameworks that would be relevant to a broader subset of the US population, although the speaker, to his credit, did emphasize that this was a new model of high-density urban living designed with Kendall Square in mind, and did not pretend this would be applicable everywhere. But it seems like a lot of effort expanded for something just for Kendall Square.)

While the speaker mentioned Hamburg in Germany plans to be car-free by 2024 and Europe is so much more forward on those matters than the US, Europe also has a very different population density than the US. When I first came to America (which meant in my case coming to Cambridge, MA) I thought it was ridiculous to let 16-year-olds drive cars instead of having them wait until they turn 18 and are presumably a bit more mature. But an American friend of mine pointed out that Americans need cars to get around in ways that Europeans do not. It's more complicated than not having public transportation. In some parts of the country it's really not efficient to have a good network of buses.

You can read more about this project here, here and here.

I wonder why having more (energy-efficient) shared rides wouldn't work, similar to Google buses in San Francisco, but not for a single company. For instance, all the companies in Kendall Square could get together and finance a system of buses for their workers. With that level of scale (tens of thousands of people work in the area, according to the speaker), you should be able to finance the scheme effectively. It will look less futuristic than glass cartridges but it might actually work.

Oil and light on canvas: the renovation of Rothko's Harvard Murals

26HARVARD1-articleLarge-v2(Photo credit: New York Times) The Harvard Art Museums are re-opening this weekend after a multi-year renovation by Renzo Piano, which has combined three museums into one, opened the rooftop and added, in typical Piano's style, glass bay windows flooding surrounding spaces with welcome light. Their re-opening coincides with the public display, for the first time in about half a century, of Rothko's Harvard Murals. Those murals were donated by Rothko to Harvard after he withdrew from the Seagram Building Commission, and installed in the university's Holyoke Center in the 1960s. Rothko made his own paints, and it wasn't known at the time that the specific mix he was using, Red Lithol, would fade under light exposure (the red pigment he was using was stable in powder form but not when mixed with the binder to make it liquid).

Thus, the murals today are extremely faded, and because of Rothko's specific technique, impossible to restore to their original condition without making irreversible changes to the work. A team of Harvard and M.I.T. research, according to the Harvard Gazette, investigated an experimental technique where colored light is projected from digital projectors hanging from the ceilings onto the panels to give the eye the impression of seeing the original colors. (The article states, "Light projection as a tool in art conservation was first described and demonstrated by Canadian conservator Raymond Lafontaine using slide projectors in the 1980s." Apparently he managed to remove the appearance of a yellowed varnish without removing the varnish itself, simply by projecting some colored light onto the painting to "absorb" the impression of yellow.)

I was dubious when I first read that the renovation of the Harvard Murals involved light projection, but the effect is truly stunning. If you look at the paintings from, say, two feet away, there is no way to guess this is not the real paint. From close enough, of course, your shadow comes onto the panel blocking the light from the projector and yes you can then see the faded paint in the same precise outline as the renovated image.

The Harvard murals differ from the paintings in three horizontal rectangular shapes stacked on top of each other, for which Rothko is best known; thus, it is particularly important for scholarship that they be put on view. They will undoubtedly broaden the knowledge by the general public of one of the leading Abstract Expressionists of his time. In the words of the Associate Curator of Modern and Contemporary Art: "One of the tragedies is the Harvard murals have not been recognized in art history in the same way they should be because they were taken down and because of the fading. One of the several things this exhibition will do is to bring them back into the discourse on Rothko’s history and the importance of them within his trajectory as an artist."

At the museum you can read a bit more information on the technological process used, which explains the approach in quite technical terms, but I haven't been able to find this online yet. In the meantime, this article in The New Scientist discusses the renovation in more detail. 

The murals will be on view until July 26, 2015. Read more here and here

Recent most-read posts

Here are my 5 most read recent posts for some reason. Quite an eclectic group!

Personal recent favorite: Links worth reading.

In the pipeline:

  1. More from DDI2014: MIT Media Lab, Hyper-Efficiency Units and Hyper-Compact Cars (going live Nov 18)
  2. Google's Creative Lab (going live Nov 29)
  3. Recent articles in Health Affairs (not yet scheduled).

The thief next to me #informs2014

So I went to get breakfast at the conference hotel this morning during INFORMS2014. A number of us, including me, had breakfast vouchers of some kind (either the pay nothing type or the pay nothing at the Starbucks or pay $10 for the complete breakfast buffet at the restaurant type). When he was done with his breakfast, the person next to me, a youngish Indian male who was an INFORMS2014 attendee (he was wearing his badge with the easily recognizable red laniard, although I don't remember if his name was easily readable) told the waiter that he had forgotten his breakfast voucher. The Hispanic waiter, working at an upscale hotel, was very courteous and said no problem, give me your name and room number and just get me your voucher within half an hour. So far so good: young conference attendee who forgot something in his room, kind waiter who wants to help him out. And then the young Indian male got up and when he had his back turned to the waiter, smiled so smugly that I just knew he had been lying through his teeth and wasn't coming back.

Now, technically speaking, if I was psychic I would have figured out the lottery numbers by now, so I did eat my breakfast very slowly to see if he was coming back - maybe I just don't trust people enough, right? Someone tried to be kind to him so he's not going to ruin it for the next person after him who really did forget his voucher in his room, right? Well, you can believe in Santa Claus if you want and hope he showed up back at the restaurant very late having ransacked his hotel room in a desperate search for his breakfast voucher, but in the half hour I waited, cutting my cantaloupe in the tiniest possible slices I could muster, the person just didn't come back.

Then what further aggravated me, when I posted a comment on Twitter, was a reply I got back about the need not to "underestimate grad students' resourcefulness in how to get free food", with a smiley face to punctuate it all. So me being me, I reply something along the lines that it's theft and disgusting. (My family was very poor. They got themselves out of poverty without stealing.) At which point it was replied to me something along the lines that such buffet crashing was not uncommon at conferences (although strongly disapproved of by my interlocutor). I thought the choice of words was telling. So this is what this generation has come to, and not any member of this generation but educated members in a professional group the vast majority of whose members hold or pursue PhDs related to operations research and data analytics and thus have or will soon have skills that will make them in high demand and well-paid in the workforce: seeing a restaurant as an opportunity for "free food" and "buffet crashing" rather than calling the behavior what it is, service theft. When you're old enough and educated enough to attend an INFORMS conference, you should know better.

Do I really have to explain the difference between getting free food because as a starving graduate student you're "crashing" a party (when the other guests aren't paying either and the food has already been ordered) and service theft at a restaurant? Really? If you have enough money to buy yourself a nice-looking blue business shirt and just as nice-looking business pants, surely you have money to buy yourself breakfast. And if you don't have money to buy yourself breakfast, I would suggest you do not steal breakfast. What are you going to do 15 years from now, overcharge your customers at your consulting company because, oh well, they have the money? record non-existing profits to convince more and more investors to buy your stock and raise its price out of thin air so you can build lavish headquarters because, oh well, you can get away with it? tell trusting investors you can achieve extraordinary returns so that they should invest their life savings with you, while you live in high style and run a Ponzi scheme because, oh well, they should've known better? You've got to start somewhere and service theft at a conference restaurant for *breakfast* when you have had enough money to pay for the airfare, hotel and business suit does have the right degree of gratuitous smugness that one would expect to precede blatant large-scale professional fraud. I guess it's no wonder with that sort of attitude that the white-collar executive world is rocked every few years by scandals of epic proportions.

Interesting projects from the MIT Mobile Experience Lab

Lots of interesting things are going on at the MIT Mobile Experience Lab at the moment. Here are a few projects that I found particularly interesting. You might notice a common thread!

  • Marriott Six Degrees: this is about redesigning the experience of the business traveller in the hotel lobby using LinkedIn information, to show how various Six Degrees users are connected and help them make new connections. The webpage has very insightful videos of an effort that positions Marriott as one of the most innovative thinkers in the hospitality sphere. You may also want to read this article in the Boston Globe or watch the video below.
  • Rethinking the MBTA Ridership Experience.
  • Mobile Applications for Tradeshows.

If you've attended TEDx conferences, you might have noticed how difficult it is to meaningfully connect with other attendees, although everybody is eager to listen to the speakers (of variable quality) and presumably shares tastes in technology, education or design. You might chat with your neighbor during the break, but you never know who else you might have connected with if you had sat somewhere else in the auditorium. TED or TEDx conferences have high ambitions about changing the world, but remain one-directional exchanges of information where (for some TEDx conferences I've attended) the organizers seem more eager to leverage the TEDx name for their own interests than to deliver a first-rate experience for attendees. The conference of the future, whether trade show or TEDx type (or...), absolutely should foster more multi-directional connections, where speakers share information with attendees but attendees also get to connect with each other. Now, you only need to have been once on a mailing list that some users furiously decided they wanted to be removed from (users who, in spite of seeing that replying to all "unsubscribe" only led to mass spamming, kept doing the same thing that annoyed them so much in others), to realize such novel mobile tools might need to focus only on the more sophisticated part of the user base at the moment. Nonetheless, a different set of tools are sorely needed so that people can make more meaningful professional connections when they attend workshops and conferences. I'll follow with interest what happens.

You can read about more projects here. Also check out the Mobile Experience Lab's portfolio book.