How I Got So Much GHT

With Good Heart Week nearing its end and my schedule unlikely to let me write another post, I have one last one to share about how I managed to earn as much GHT as I did.

As of this writing I’m in the top 5. I’ve been as high as #3 but currently I’m sitting at #4. How did I do it?

Well, to start off, when we got the initial announcement on April 1st I tried to think of something that would be both a fun use of Good Heart Tokens and would earn me a lot of them. The best idea I came up with after a few minutes of thinking was a donation lottery.

That was all in good fun for April Fool’s Day, but then when GHT got extended to a week long I had to come up with some plan to produce real posts if I wanted to earn GHT. Since I had committed the lottery to pay out at least $1000 regardless of what happened, I now had an incentive: earn as much GHT as possible to offset the amount I was personally going to have to donate!

I wrote a shortform post asking folks for ideas on things they might like to see me write. I didn’t get any bites.

Later in the day, I thought to myself “what the heck, this is a week for writing things I wouldn’t otherwise feel were worth the effort to write, but given the changed incentives we’ve cross the threshold where, at least for a week, I’m willing to spend time writing things that are less likely to be novel or enduring in exchange for GHT”.

That said, I didn’t want to spend a lot of time writing since I still have to do my day job. “Oh, hey!” went my brain, “what if I wrote about my day job!”.

So I ended up doing just that. Most of the posts I wrote this week were on the subject, either directly or obliquely:

The last one isn’t explicitly about software engineering, but it does reflect a lot of what I’ve learned about communicating effectively in a large organization.

I tried to think of other things to write about. I came up with two other ideas of things that I’ve talked about at various times but didn’t recall having written down:

I also posted one question to try to get some additional conversations going, but so far it hasn’t gotten much traction:

In all cases I followed a few rules with these posts:

  • Write about something I already know well. No topics where I have to think up ideas I haven’t already spent a lot of time thinking about. The ideas need to be crystalized in my mind and ready to be communicated.

  • Write from personal experience. Following on from the last point, if I can’t think of an impersonal approach to a topic, just write about it from my personal perspective. If I get stuck, just say “I don’t know how it is for others, but here’s what I’ve seen” and keep going. Anecdotes are Bayesian data points and valuable to readers who know how to weigh them as long as they are clearly marked, so don’t keep them a secret.

  • Don’t spend too much time on any post. All these posts were written in 1-2 hours. Generally there was a first pass to get what I had to say out and then a second and maybe third to do some light editing. The posts are more narrative and less structured than my normal posts.

  • Avoid controversial topics. I could have written some stuff about developmental psychology, cybernetics, or epistemology. I think all of these topics would have been more controversial or run afoul of my next rule.

  • Avoid specialist topics. I didn’t want to write about things that would only appeal to a small audience since it would be harder to get a lot of GHT that way (people probably don’t upvote things they don’t understand). So I needed to write about this that would have a broad audience and be understood readily by that audience. This meant that AI safety and Zen were also off the table.

  • Focus on posts. Posts are worth 4 GHT per upvote vs. 1 on comments. My comments don’t usually perform very well, but I can usually get 10-20 upvotes per post. Posts seemed like the better strategy.

I think this basically worked, and I think I saw others attempting a similar strategy. Why did I do so well? I think a few things helped:

  • I have some brand recognition on LessWrong. Not top-tier but I’m probably a solid B-list LessWrong contributor.

  • This got amplified by the donor lottery post and the first software engineering post doing well. Both were engaging enough to get a lot of votes. This meant I was now a name to watch, at least for this week. It also helped that Ben called the donation lottery post out in the post announcing Good Heart Week as something that was now banned.

  • This is important, because attention is limited. I doubt many people read everything on LessWrong every day. Having some brand recognition gets you marginally more readers. Before you object, I think you know this is true. If Eliezer wrote a post titled “My current thinking” you’d probably read it, but if NewUser1234 wrote a post titled “My current thinking” you’d probably ignore it. I promise you there is at least one LessWrong reader out there who uses author to help decide what to read and I’m him. There’s likely others.

  • I tried to snowball my engagement a couple different ways: post every day, link back to my previous posts so I could get some more engagement on them, post a second time if the first post didn’t do well.

  • I’ve been around LessWrong a long time. I don’t normally go for applause lights or lean into what’s trendy, but I can if I want to. I don’t think of this as abusing the system so much as giving the people content they want. For example, lots of people here are interested in topics like software engineering, moral mazes, and the LessWrong community. Writing about those topics is going to draw more eyeballs, and if I wrote something good that’s going to up my chances of getting GHT. I’ll save my posts for bugging you all about good epistemology some other time.

  • I was willing to just write things and see what happened. If it didn’t go well I could always just give up and cut my losses. My life wasn’t going to be much worse if I put low effort into a post and it didn’t get many upvotes.

  • Not enough of the people who could have displaced me tried. A few did and they’re in the leaderboard alongside me. If enough top-tier LessWrong authors had engaged with Good Heart Week the way I did, I probably wouldn’t have even made the leaderboard after day 1.

There’s probably some lessons to extract from this. The biggest one is probably the same lessons others have tried to impart and I’ve failed to learn until this week: dramatically lower my bar for posting on LessWrong. The votes will sort it out. If it’s not interesting, people will just ignore it, but sometimes a post that seems not worth writing is actually something people are really excited to read, and people are also happy to read a quickly written post rather than no post at all. Additionally, there doesn’t seem to be a lot of marginal value in taking a lot of time to edit my posts carefully; I won’t get dinged much if I don’t heavily edit a post to optimize the delivery. Heck, maybe I’ve been making my posts worse by doing that!

Obvious there’s some risk I take this too far and start spamming LessWrong with crap posts. I’m not too worried about that happening (I’m a bit too busy to post like I did this week most weeks), but let me know if you think I ever err too far in the other direction.

I’ll leave that to you all to figure out any other big takeaways. Probably something something incentives somethings.

This week has been fun, but I’m also very much looking forward to LessWrong getting back to normal on Friday!