This should have been posted earlier this week, but because I had midweek meetings, I could not complete it.
Anyway, this Monday and Tuesday, I attended a 2-day workshop in my campus, about the importance and new directions of research in New Media. We had talks from experts in medical image, sound compression, geometric modelling, etc. Actually, I mainly came there to
get the free meals see the scheduled talk from Ming C. Lin, whose research in motion planning in Computer Graphics coincides with me, but it was unfortunate that she could not come because of some immigration issues. However, I found the most interesting part to me is not in the technical talks themselves, but rather in issues which I have written here a few times: whether academic research in should be market-led or fundamental. Well, the workshop specifically only discussed researches in media technology, but I think this kind of debate also occurs in other fields.
For a background, Computer Graphics to me looks like an applied field, in which academicians from various disciplines and backgrounds—Computer Science, Electrical Engineering, Pure and Applied Maths, Medicine, Biology, Art, Psychology, and maybe some other fields you never imagine before; Philosophy?—have very close ties with people in the industry. Cross-business-model collaborations are not uncommon, and companies like nVidia, Autodesk, Microsoft, and Pixar have had great contributions in the scientific and artistic advancements alike in Computer Graphics, either from the end-products (movies, games), development tools (modelling software like 3DS Max and Maya), hardware (graphics processors like nVidia geForce), and programming libraries (e.g. nVidia Physx).
Before I proceed further, I would like to introduce the backgrounds of the 6 panelists.
- [Nadia Magnenat-Thalmann] is a professor who has spent most of her career in Europe, particularly in Switzerland, and just a few years ago came to Singapore to expand her research in new media.
- [Daniel Cohen-Or] is a professor who spends most of his research career in the USA and Israel.
- [Susanto Rahardja] is a professor from Indonesia, but got his degrees from Singapore and has been working here as well since the 1990s.
- The fourth panelist was [Franz-Erich Wolter], a professor from Germany, but was also a professor in some top universities in the USA.
- We also had [Kari Pulli] as the fifth panelist. I think he is quite international, as he has worked in Finland, USA, and Italy. However, he comes from different background from other panelists, as has mainly worked in the industry.
- Lastly, there was [Shi-Min Hu], a professor from China.
So the main question is, as I mentioned above, whether research in new media should be market-led or fundamental. The session was started by Prof Nadia as she mentioned that there seemed to be a contrasting difference between funding agencies in Switzerland and Singapore. I know little about research in other countries to summarise it here, but she said that in Singapore, she had always been asked about how her research would benefit the Singapore. I can confirm this one, because I had to fill the same exact question when I wrote a research proposal a few months ago. On the other hand, it is sometimes quite hard to determine how a particular research will benefit the society, because it is often not motivated by what people in the society need, but rather due to curiosity. This view was then also backed by Prof Daniel, who said that he did not like to be “directed”.
An opposing opinion came from Prof Susanto. He mentioned that it is the general preference of Singapore government to support researches which have economic impact. Therefore, if one group desires to conduct fundamental research, they can make a proportion, e.g. 90% for applied research and 10% for fundamental research. Indeed, this proportion was proved to be controversial, as soon after that Prof Franz refuted it, saying that 90-10 proportion was certainly not a good balance (and later on Prof Susanto clarified that it had not been to be rigidly followed too). However, Dr Kari had a somehow approving remark, saying that we should go with respect to the market need. He wanted us to keep in mind though, that companies usually have better resources to conduct this kind of research. Prof Shi-Min Hu ended the first part of the panel discussion with an example of how his research went into commercial product, and his remark on how applied research is important to get money or funding, in order to support future research. He also argued that mere papers cannot change people’s lives, and they have to be implemented in real products before becoming useful.
What sounded interesting to me was Prof Nadia’s immediate wonder, that one person’s preference towards the answer (if any) for this question depends on her/his background. Prof Nadia noticed that there seemed to be different perspectives between developing Asian nations, which often go towards the applied direction, and developed European countries, who already underwent this phase in the past and now prefer doing fundamental things. Prof Susanto actually further elaborated this, by saying that since Singapore does not have much natural resources, what they basically have is human resources. Brains, he said. That is why they have to use this opportunity to develop their nation. However, Prof Nadia admitted that she would be happy to compare the case of Singapore with her home country, Switzerland, which is equally not in possession of abundant natural resources, but can boast its 26 Nobel laureates. Another interesting fact is that, as Prof Franz described, in Germany, both directions are entertained by the government by having two distinct but equivalently distinguished bodies, namely Max Planck Institutes for fundamental research, and Fraunhofer Society for applied research. For example, for the field of Computer Graphics, the former has this [Computer Graphics department in Max-Planck Institut für Informatik (MPII)], affiliated to University of Saarland, whereas the latter has [Fraunhofer-Institut für Graphische Datenverarbeitung (IGD)].
I could not notice any concluding remark in the workshop, hence an open issue. Nevertheless, I think I can relate to Prof Nadia’s statement, that whereas this question is hardly a rigid “this only or that only” question, one’s background plays major role in her/his preference in doing research. Growing up in Indonesia, which I can say is a third-world country, I often read from the positive-sounding-news-deprived newspapers that there are always problems in my home country, and this always motivates me to do something, although not exclusively for my home country, but for civilisation in general, as I wrote in [my previous Indonesian post] (sorry for non-Indonesian speakers), through technology, hence an applied research. However, my stance towards fundamental research itself is divided and inconsistent. I have always expressed my cynicism towards parts of pure science, particularly Mathematics, which do not go to application level. This thought was firstly published in [this very old post], and revisited in the previously linked article. However, attending this workshop advances me one step further. Prof Franz mentioned an important difference between pure and applied researches. While applied research can have immediate impact, pure research can benefit in a somewhat longer term.
I just came across [this page], which seems to be the abstract of an upcoming talk from Eric Lengyel, a well known figure in Computer Graphics, in this year’s [WSCG’2012] conference. It mentions that in search of efficient computation, Computer Graphics researchers have repeatedly re-discover and use linear algebra and geometry theories presented centuries ago, having consistently and exclusively lying in the theoretical realm for the same period of time. A prominent example is the concept of [quaternion], which was invented quite accidentally by Sir William Rowan Hamilton when he was crossing Brougham Bridge in Dublin in 1843, although clues have been firstly given by Leonard Euler and Olinde Rodrigues long before that. It seems to me that in the course of about one century and a half, only theoreticians exclusively used it, until the emergence of Computer Graphics, which often [utilises quaternion as a mean to rotate objects], makes it possible for laymen to use it. So, if you remember that you played Tomb Raider back in mid-90s, you can celebrate yourself for [using quaternion somewhere in its lower layer]. This is a good thing, and I suppose (and hope!) this phenomenon can be found in other new fields of study as well.
To conclude this post, it seems to me it is always a never-ending question whether one should go fundamental or applied. However, both approaches do benefit people in different ways, and therefore both have to be proportionately appreciated. Although I do realise that some countries might prefer one direction than the other, it is just my opinion that both ways have to be supported.
If anybody attending (or even happened to conduct a talk) in the workshop found that I made some mistakes in re-telling how it went, I will be grateful if you can inform and correct me.