Chasing performance.
When good enough, isn’t enough.
Performance. No one gets there by being one dimensional.
There are three dimensions that should be of interest if you want to extend your performance envelope.
The first dimension is commitment and want. It doesn’t matter if it is 39 degrees c, 80% humidity, or escalating cyclone warnings, outside. A runner’s got to do what a runner’s got to do. We all ran this morning, despite the lashing rain. Some of us drove 18 km, one way, to run by the creek in Karachi.
The second dimension is training, process, discipline. Need to have a plan other than being stark raving mad. We mix time trials and hill training with long slow runs. Hard effort with light effort, weight training with rest days. For rest and sleep are also training.
The third is assessment, benchmarks, standards. You have to measure yourself against a scale. Who do you want to be? How far are you from that goal? Is the training working? Are you any closer this week? Compared to last week?
Computer Science education is very much like track or distance running. You can watch all the videos you want but that sub-20 5k will happen if you show up every weekend and clock the assigned mileage and training. It is easier to tackle larger bandwidth problems after you have knocked out lower intensity challenges. Think getting to the “Boss level” without all the powers, armor, spells, tools and strength.
The only person standing between you and that goal is you. Because only your legs and your mind can carry you across the finish line. You need both because when one gives up, it is the other that shoulders the burden. When that happens, you can’t follow a guide, a desire, a methodology blindly. Understand why this benchmark. The tougher the benchmark, the more important it is to answer this simple question, why? Why do we all need to suffer so you can do this?
Of the three, the third dimension is the hardest. It is easy to cheat. Pick an easy benchmark. Fib on the test. Be comfortable. Get help from unsanctioned sources. Move the needle when no one is looking.
No one would know. But you were. You knew. You do.
It is also the most important. Feedback loops are the key to stepping up. The faster and sharper the loop, the faster you improve. But there is a catch. The right benchmarks are hard. Because they change. If you are serious about chasing performance, you change them yourself. Getting comfortable with running a mile in 9 minutes, move it up to 8. Understand how to train a GAN on Tic Tac Toe. Step up to Scrabble. Figured out local hosting. Now make it run on a NIM. As you move up the ladder, lower bandwidth challenges become part of your muscle memory. By the time you parse the problem query, a solution has already been designed and proposed by your brain.
Who does that? The same people out for their long slow run in the midst of a cyclone warning. In Zen and the art of motorcycle maintenance, Pirsig calls them participants. Most people live life as spectators. Some live it as participants. It is hard being a participant. We don’t do things because someone asks us to do them. We do them because we have to do them. We don’t have a choice, because nothing else would be acceptable to us. Something’s broken, fix it. Don’t know how, figure it out.
Our path is our burden.
Like performance. We don’t chase performance for sake of performance. We chase it because performance defines who we are. A 9:34 mile is 6 minutes slower than the world record for the distance. But it is the best I could do this month. I don’t have to run it in 3:43. I am not going to run it in 3:43. But I am going to run it a bit faster than 9:34 next month. Just like I ran 24 seconds faster in August, versus July. 2 minutes faster in July, compared to June.
For now that is enough. Till the day it won’t be.
