Hacker News Reader

Comment by jboggan | Hacker News Reader

This reminds me of the time in my doctoral program when I turned in a draft journal article to my advisor. He returned it a week later with no marks or written commentary whatsoever, other than telling me "do it again." I asked what was wrong with it and what I needed to do differently on the next draft. He replied "this isn't excellent. Don't give it back to me until it's excellent." I asked what excellence entailed and he said "you just know."

People ask me if I learned anything in grad school.

View on HN

This sounds similar to feedback I just got on my dissertation's conclusion section. They told me to rewrite it, noting that it should be different, but they weren't sure how. Asked for more detail, they said, "Make it punchier."

So did you? I'm not endorsing that minimal feedback technique, but it would certainly make me think pretty hard about how to improve the paper. Bang-bang control can work.

If you're the boss of any enterprise, and you want to make things good ('excellent') the first time, you only have your instincts and experience to go on.

But in the poster's context, "excellent" is a non-objective undefined term. In an enterprise, the incentive is for you to define how your subordinates do their jobs.

So, it's a different subject altogether.

This actually gets at what the paper describes as the problem with defining what "excellence" means. (I read about half the paper.) Everyone defines it differently, so excellence basically just means that it's judged well by whoever is doing the judging.

This is an old chestnut of a philosophical problem that many youngsters of my generation first experienced in the form of 'Zen and the Art of Motorcycle Maintenance'.

There's no solving it. And yet we all know that differences in quality are real. This is not an argument I want to engage in.

My question was, did the GP learn anything from being told 'not good enough, do it again'? They imply not, but I bet they did.

I've learned lots of things that way. Grant proposals, job applications, asking someone out on a date: lots of things you try out and don't get feedback beyond 'nope'.

I think part of the problem was that the person who had been writing all of his research papers for him the last several years had left the lab unexpectedly and the advisor had no feedback to give. I don't know, I became pretty demoralized when each attempt to write up my research got that kind of response. The advisor was super enthusiastic about me when I started and told me that my research was so good that my first paper was going in Science or Nature. He kind of built me up and let me crush myself with high expectations. It was an interesting education.

You went in with an optimization problem and came out with 3-satisfiability.

Reply on news.ycombinator.com