Re: high school graduation speech

Walking down memory lane with my gmail archive, I found an interesting email I sent in response to Paul Graham's graduation speech. I had emailed him on Fri, Jan 21, 2005 at 3:57 PM, the following:

---

Hello Paul,

I hope you're not being inundated with slashdot readers writing their feedback. I'm certainly not lessening the effect by writing, but I feel I should write since I enjoyed reading your graduation speech and am as dismayed as you that it couldn't be delivered. I'm not in highschool anymore, but mentoring highschool students in my free time I know your advices and anecdotes to unveil reality are good ones. I've even forwarded your webpage to my younger cousins who themselves are in college but still very curious about life, the division between childhood and adulthood, and the "point of life" - to stay upwind as you put it.

Apart from my compliments, I also have a suggestion, which I'll get to after a short anecdote of my own. I was very mathematically inclined as a kid and delved into software more than most I knew. All along, I had only one friend whom I could learn from and he equally learned from me. Even my father who was a software developer was uninterested in the murkier topics of computer science theory, such as lambda calculus, frequency analysis, and feistel networks, as the practicing world cared more about programming libraries. Needless to say, it was a difficult journey to learn and a lonely one at that. I knew college would be better, but that's eons away when you're in eighth grade. I stumbled upon linux, open source, and a community working on things without monetary purpose nearing the end of my tenth grade. My first email communique outside of my highschool was to Andrew Tridgell, then the sole samba developer. I had my vague notions of tcp and udp and OS datagram frames but in one email response he clarified questions I would've spent the next year analyzing. Instantly, solitary learning where I learn from my mistakes like an ape became human learning where I stood on civilization learning the past mistakes of all. It was incredible, and I wish I hadn't had to have stumbled on that revelation.

I'm not saying open source is the answer to everyone's grade school intellectual doldrums, even if it was my answer; I'm saying that a useful community will inevitably exist outside college, and highschool students impatient for that college life can tap into it earlier. The key element is people. Most students, including both the academic ones trying to learn and the non-academic ones trying to be popular, will benefit from the idea that there are more people to know and learn from than those in their own school. I know of far too many students who never communicate in email or chat beyond their school classmates, and parents unfortunately find that comforting. The concept I couldn't grasp was how many 6 billion people are and yet how only a few dozen people would be interested in samba in 1995. As a ratio, it's astounding. However, these non-popular -- distinctly separate from the unpopular -- projects are bastions of clever lonely people, the perfect type for a student with little to offer besides attention and a lot to gain such as knowledge.

Ergo, my suggestion. I believe it would be highly useful to impress upon students how they should look beyond their region - that while their home town may be the only place in the world to know certain inside jokes and terms, there are things far grander in the world and fractious enough as to make the individual teams small, closeknit, and meaningful. While being trendy and knowing the latest, local, and popular things can make one feel good about oneself, the eclectic, esoteric, and historical things are longer lasting benefits which compound with themselves in value over time. To be eclectic, however, one cannot be content with what is provided. I was very distrusting of supposed quality, and knew the world contained a spectrum of quality far greater than I could contemplate. Far too many people cling to the first anchorman, reporter, or developer they meet if they're interested in that subject. I suggest exploring and discerning whom to emulate.

Being mindful of history is necessary in order to be eclectic. Old news tends to be overlooked, or be seen as unprofitable, making it somewhat immune from the noise of limelight seekers' premature ideas and from marketing propaganda. There's money in making software, not in writing a taxonomy of comparisons between, or meticulously documenting the concepts wielded by, different software. There are even taxonomies on taxonomies, each level becoming less popular and less profitable and taking longer to complete, thereby missing the slim window of public interest. However, by sitting in 1995 and reading tcp/ip lessons in 1994 about lessons from 1991 etc., dating back to the release of the cornell worm in 1970, one could learn a lot more than if one had only read some arbitrary book circa 1980 focusing purely on unproven and trendy '80s paradigms. I've given the following advice to many: we should look at the thirty year topics that are still alive today, even if barely, and trace the discussion back to their origins. By learning from the collective discussion, any contemporary incident can be seen through the lens of a learned person, yielding an amalgam of concepts surviving a darwinian massacre of preceding years' ideas. Once the lens has been shaped, it serves as a crude tool to craft finer tools. Apply that lens on another, more recent, incident, and repeat the process until you're looking at present day situations with an extremely well-pruned, eclectic corpus of knowledge with which to interpret anything you choose. This strategy works tremendously well at solving the problem of "noise" -- too much nonsense posing as quality work; and, interestingly, this same approach is taken by bayesian spam filtering.

Sorry for writing as much as I have - it was originally meant to be two or three short paragraphs. Pardon me for any typographical errors.

Cheers,
(My name)