Monthly Archive for April, 2004

It probably makes sense for me to have a vague plan of what use I might get out of my Ph.D., when I have my Ph.D. And honestly, I haven’t given that much thought. I just imagine I’ll have some sort of job at an established company doing research, or possibly at a university teaching and doing research. And then I toss out a “or maybe start a company of my own, if my research leads in that direction.”

So here’s the outlook. I want to lay the groundwork for a new kind of relationship between people and computers, where instead of programming computers, people teach computers, and in return expect them to have a clue about what they are talking about. Basically to establish a medium of communication where there’s actually some sort of transfer of ideas going on, and the reachitect how information is stored and manipulated around that medium.

In graduate school I want to figure out what the psychological and computational constraints are on such a medium and garner some theoretical and empirical support for the idea. In my spare time, I’ll keep implementing things as I have been.

But even if I am successful over the next seven years, I’ll have a platform without users and a product without a market. The information industry has gathered around Microsoft’s development platform like moths to a bug light. And given my belief in software libre, there wouldn’t be a product to buy even if people wanted it. Instead, the company I would start would *buzz* *buzz* provide services.

We’d have a two-prong strategy: get into schools, and get into niche markets. The first prong involves making the platform the best possible teaching tool it can be. The platform should make programming more fun and enable nearly limitless creativity, so I think this is a viable goal. The second prong is the way of the disruptive technology. Disruptive technologies are initially unsuitable for the mass market because of market inertia and a lack of refinement, but provide strong incentives to niche markets. I think this new platform fits that description quite well. So a major goal will be to identify environments where the platform would be useful, which is basically all of the places where Microsoft Office scripts and basic VBScript and whatnot are heavily used.

With children coming up the ranks knowing the platform and these niche markets leaking into the larger market, I think the platform could take over the entire industry. The benefits of providing programming capabilities to normal users are just enormous, and will make application development technologies like XAML look like electronic typewriters: fancy implementations within an obsolete paradigm.

So this yet-to-exist company would help schools set up low cost thin client linux bases computer labs, provide lesson plans and support. It would help businesses deploy the platform, provide training courses and run conferences. And it would evangelize. I’m not sure if it would ever make a ton of money, but it would be a target for venture-capitalists, and a critical spoke in the wheel of acceptance for the platform, as well as a way for me to earn a paycheck.

The news of the hour: I have decided to go to Indiana University in august, to study Human-Computer Interaction in the new School of Informatics. I have to say immediately after I reported my decision to the interested parties, I was filled with dread that I had made the wrong decision. But I’m slowly feeling better and better about it, and I think by the time I have gotten out of Storrs, the “UConn haze” will be gone, and I will feel fine.

I have been poking at some ideas about what I want to do research on. It sounds like I will be teaching for the first year, which hopefully means I won’t be “researched out” and I’ll be motivated to pursue some of this stuff independently. And I’ve been reading bios of some people doing work in this area at Indiana, but not in the School of Informatics. It’s hard to say if I’d every be able to work with any of these people, or pick their brains, but it’s worth checking out their research over the next few months and contacting people who I think would be interested in what I’m doing.

I’ve also checked out a bunch of books from the library and photocopied a bunch of papers on the philosophy of language and in particular semantics. None of it has been very fruitful, but I haven’t really gotten very far into it. I kind of wish I had someone to say “read this and this and this, they will give you a good overview of our current understanding of semantics and their computational application.” But I’m not really sure who that person is.

Regardless, these philosophy books have made me think… As a researcher, how aware should I be of my inherent biases? I wonder if ecological psychologists, for example, aren’t biased by their experience of perception and action in their account of it. In other words, perhaps the ecological explanation is only true when verified through human perceptural machinery, but evaporates when it’s not perceived but simply cataloged.

And philosophers, I think, are highly biased to precise, analytically useful explanations of things. Not just philosophy, but all of the research world is biased towards explanations which can be put on paper in great detail, simply because of the fact that research is useless unless it is such. But what of things which can’t be explained? Surely there are things in this world which cannot be put into words, and even more things which could never be encapsulated on a thousand bound pages. But perhaps more importantly, even where phenomena can be accounted for and verbalized, we as researchers are biased towards ready accounts and easy explanations. We are biased towards comprehensible and overtly consistent theories.

And in my case, I am biased towards the computable and the implementable. The possibility that natural language processing can’t be done by a Von Neumann computer is relegated to the very back of my head. I want language to be computationally tractable and every theory I work towards is biased by that. But like the researcher trying to publish work or the philsosopher trying to develop analytical tools, at some point I just need to accept my biases as a result of my goals. The question is: where is that point, and at 22 should I be avoiding it?

So, UIC isn’t offering my any financial aid. That means it’s currently down to Indiana and UConn. I should be hearing from UConn about financial aid in the next day or so, and I should be hearing from Cornell about admission within a week. Regardless, my only financially viable choice at the moment is Indiana University.

I think that’s how I’m going to look at it. I’m going to Indiana unless Cornell or UConn convince me otherwise.

And I feel pretty good about IU. It seems a lot like UConn in many ways, except bigger I guess, and in the midwest. The School of Informatics seems enthusiastic about me personall, which is nice. And I think it would work out.

Not sure about being a hoosier though.