The LessWrong 2021 Review: Intellectual Circle Expansion

Ruby and Raemon
1 Dec 2022 21:17 UTC
95 points
55 commentsLW link

If you’re already familiar with the LessWrong annual review, feel free to skip this opening and jump to What’s new in 2021? or Unveiling the 2020 Annual Review books

The tradition continues! We will do for 2021 the same as we did for 2018, 2019, and 2020: approximately one year after the year is complete, we look back with the benefit of some distance and ask how good were the posts written then? After the hype has died down, do their claims still seem valid? Were they useful or interesting in a lasting way? Etc.

On LessWrong we’re trying to make intellectual progress on problems that matter, but it can be hard to tell whether we’re making progress. The Annual Review is the once-yearly Schelling time to reflect, debate, discuss and vote. It’s a poke to say “remember that idea you and everyone else were obsessed with in 20xx? Did it continue to matter or was it just a fad?” It’s also the time when we can determine which contributions most affected people in the long-term, and reward those authors and inspire others to follow suit. I think it’s pretty great that when an author who invested hours into a post and greatly benefits the community, the community eventually recognizes and rewards them in some fashion.

The full technical details of how the Annual Review works is described in the final section of this post, but the core of it is the same as the past few years. There are three phases:

  1. Preliminary Voting Phase (Dec 1- 14): Site members start identifying posts especially worthy of consideration in the review. Posts with 2 preliminary votes move into the Discussion Phase.

  2. Discussion Phase (Dec 14 - Jan 11): Site members begin reviewing and debating posts in earnest. This discussion will inform members on how to cast their final votes. Posts that receive at least one written review move to the Final Voting phase.

  3. Final Voting (Jan 11 to Jan 25): Site members allocate their vote budget (quadratic voting) amount nominated posts. The outcome of the final determines the Annual Review results.

For more of the philosophy of the Annual Review, see the previous announcement posts here and here and here.

Getting Started

If you are super eager to get started participating in the review, you can start preliminary voting on posts from the All 2021 Posts page, or the View Your Past Upvotes page. Note: only users with accounts registered before January 2021 are eligible to vote.

New in 2021: Offsite Content

Largely, the system is the same. With one exception–and this might be a terrible idea because it was a last-minute decision and it isn’t fully thought out yet:

This year, we are allowing people to submit off-site, non-LessWrong content to the Review. If you want something posted on another website, e.g. ACX or arXiv, to be considered for the review, make a linkpost for it and then let the LessWrong mod team know, we will backdate and it can be included. (Let us know by writing a comment on this post or contacting us).

What kind of external resources are worth including?

There are two kinds of content that seem particularly worth including in my mind:

  1. Content that was broadly part of the community’s intellectual diet even if it wasn’t posted on LessWrong. For example, suppose we were doing the Annual Review of 2014 and Meditations on Moloch hadn’t been crossposted to LessWrong. Since it was a major contribution to the community’s intellectual commons, I think it would have been good to include it in that year’s review.

  2. Highly impactful Alignment content that’s not on LessWrong/​Alignment Forum. AI Alignment is the most notable specific intellectual problem our community is trying to tackle, and I’d like to see our work evaluated alongside major contributions on the problem, no matter where they happen.

Why include external content?

I don’t feel entirely philosophically convinced that this is correct, but some arguments in favor:

If our goal is to determine which content from 2021 was the most significant intellectual progress of the community, I want to know not just how good a given post was among content on LessWrong, but also among the content our community was generating and consuming generally. If the best LessWrong post is only the #10 best post of 2021, I’d like to know that, and vice versa. I don’t think we’re going to get a comprehensive review of everything on the Internet and rank ourselves among that, but I do want to offer community members the opportunity to nominate something they think is especially worthy.

If there are significant and valuable ideas being generated that haven’t been circulated on LessWrong, I’d like to know what they are (and figure out why they’re missing).

If there are some ideas or concepts and posts that are influential to our intellectual community, I want them to get subjected to the same scrutiny and epistemic checks that we conduct for ideas posted on LessWrong.

What if this is a terrible idea?

It might be and that really depends on how many people nominate offsite content and what that content is, but I feel the EV of experimentation here is positive. I/​LessWrong team reserve the right to change my mind about external content and decide to exclude it from the Annual Review process at any point.

What about prizes and books?

Historically authors of posts voted highly in the review get prize money and have their posts included in published print books. I do not yet know how external content that gets highly voted will be treated for these purposes. It seems like offsite content shouldn’t obviously be included in a “Best of LessWrong” volume but it’s not clear and we haven’t figured it out. Tbd. I think that as content gets nominated, we’ll be able to form a coherent position on this.

2020 Books: The Carving of Reality

After the Annual Reviews of 2018 and 2019, we made beautiful print books containing the top-rated posts of those years’ reviews. It turns out that making gorgeous books is a lot of work and it’s unclear if it’s worth the opportunity cost, so we weren’t sure we’d do it for the 2020 review. In the end, we decided to!

Because we got started late, the 2020 books will not be ready in time to be delivered for the 2022 holidays, but hopefully, you can have them by January. They look pretty dope though! (Covers below). You can already see the top posts from the books on the Best of LessWrong page. A proper announcement post for the 2020 Books will be coming shortly.

We’re still unsure about the long-term future of the Annual Review books given the opportunity cost. For this year’s review, we intend to make at least one book (with fewer posts perhaps) in order to test some ideas for how to make them faster/​more efficiently. We’ll see how that goes to determine what we do in future years.

How does the review work?

Phase 1: Preliminary Voting

To nominate a post, cast a preliminary vote for it. Eligible voters will see this UI:

If you think a post was an important intellectual contribution, you can cast a vote indicating roughly how important it was. A vote of 1 means “it was good.” A vote of 4 means “it was quite important”, and is weighted 4x a vote of 1. A vote of 9x means it was “a crucial piece of intellectual progress.”

You can vote at the top of a post page, or anywhere the post appears in a list (like the All Posts page, or the new View Your Past Upvotes page).

Posts that get at least one positive vote go to the Voting Dashboard, where other users can vote on it. You’re encouraged to give at least a rough vote based on what you remember from last year. It’s okay (encouraged!) to change your mind later.

Writing a short review

If you feel a post was important, you’re also encouraged to write up at least a short review of it saying what stands out about the post and why it matters. (You’re allowed to write multiple reviews of a post, if you want to start by jotting down your quick impressions, and later review it in more detail)

Posts with at least one review get sorted to the top of the list of posts to vote on, so if you’d like a post to get more attention it’s helpful to review it.

Why preliminary voting? Why two voting phases?

Each year, more posts get written on LessWrong. The first Review of 2018 considered 1,500 posts. In 2020, there were 3,000. Processing that many posts is a lot of work.

Preliminary voting is designed to help handle the increased number of posts. Instead of simply nominating posts, we start directly with a vote. At the end of the Preliminary Voting phase, the results of the vote will be published. This will help the LessWrong community prioritize reviews. Posts that are highly ranked can invite more investigation of how they stand the tests of time. If you think a post was (unfairly) ranked low, you are welcome to write a positive review arguing it should be considered more strongly.

Posts which everyone agrees are “meh” can get deprioritized, making more time for more interesting posts.

How is preliminary voting calculated?

You can cast an unlimited number of votes. However, the more votes you cast, and the higher your total “score” (where a “9” vote counts for 9x the score of a “1” vote) the less influential each of your votes will be. We normalize voting strength so that all users who are past a certain “score” threshold exert roughly the same amount of total influence.

On the back end, we use a modified quadratic voting system, which allocates a fixed number of “points” across your votes based on how strong they are.

Fine details: A vote of 1 costs 1 point. A vote of 4 costs 10 points. A vote of 9 costs 45 points. If you spend more than 500 points, your votes start to become proportionally weaker.

Phase 2: Reviews

The second phase is a month long, and focuses entirely on writing reviews. Reviews are special comments that evaluate a post. Good questions to answer in a review include:

  • What does this post add to the conversation?

  • How did this post affect you, your thinking, and your actions?

  • Does it make accurate claims? Does it carve reality at the joints? How do you know?

  • Is there a subclaim of this post that you can test?

  • What followup work would you like to see building on this post?

During the Review Phase, we’ll be giving out prizes for people who write especially good reviews. We’ll announce details later. For reference, last year we gave out $3600 in prizes for reviews.

Phase 3: Final Voting

Posts that receive at least one review move on the Final Voting Phase.

The UI will require voters to at least briefly skim reviews before finalizing their vote for each post, so arguments about each post can be considered.

As with last year, we’ll publish the voting results for users with 1000+ karma, as well as all users. The LessWrong moderation team will take the voting results as a strong indicator of which posts to include in the Best of 2020 sequence.

(Note: I am currently uncertain whether Final Voting will use the fine-tuned quadratic system from last year. I plan to take last year’s voting data, round each vote to the nearest “1, 4, or 9”, and see if the results are significantly different from the original vote. If they aren’t very different, I suspect it may not make sense to encourage everyone to spend a bunch of time fine-tuning their quadratic points. I’m open to arguments in either direction)

Voting on Important Intellectual Progress

In past years, the vote was officially for creating a published book. This made it easier to reason about what exactly you were voting for, but also meant that some types of posts were harder conceptually to reward. Some important progress isn’t very fun to read. Some important posts are massively long, and couldn’t possibly fit in a book.

So this year, I’d like to formally ask that you vote based on how important an intellectual contribution a post made, rather than whether you think it makes sense to publish.

The LessWrong moderation team will take stock of the top-rated posts, and make judgment calls on how to best reward them. Some may fit best into anthology style books. Some may be more appropriate for (eventual) textbooks. Some might be important-but-tedious empirical work that makes more sense to give an honorable mention to in the books, while primarily rewarding them with prize money.

In practice, this is not that different from how we’ve been assembling the books in previous years. But it had been a bit ambiguous, and I thought it best to make it official.

You are welcome to use your own taste in what you consider important intellectual progress. But some questions that might inform your vote include:

  • Does this post introduce a concept that helps you understand the world?

  • Does the post provide useful and accurate empirical data?

  • Does this post teach a skill that has helped you?

  • Does this post summarize or distill information that makes it easier to grasp?

  • Do the central arguments of the post make sense?

  • Does this post promote an important and interesting hypothesis?

While writing reviews, it’s also worth exploring questions like:

  • How does this fit into the broader intellectual landscape?

  • What further work would you like to see?

Go Forth and Review!

Head over to your past upvotes page and start voting!

Ruby and Raemon
1 Dec 2022 21:17 UTC
95 points
55 commentsLW link