Polling Thread October 2017

by Gunnar_Zarncke1 min read7th Oct 201713 comments


Open Threads
Personal Blog

Maybe the last installment of the Polling Thread.

At least I guess it's the last one before we switch to the LesserWrong codebase which sadly doesn't seem to support polls. Maybe to easen the transition we can share polls, e.g. on Google Forms or SurveyMonkey. Or discuss alternatives.

This is your chance to ask your multiple choice question you always wanted to throw in. Get qualified numeric feedback to your comments. Post fun polls.

These used to be the rules:

  1. Each poll (or link to a poll) goes into its own top level comment and may be commented there.
  2. You must should at least vote all polls that were posted earlier than your own. This ensures participation in all polls and also limits the total number of polls. You may of course vote without posting a poll.
  3. Your poll should include a 'don't know' option (to avoid conflict with 2). I don't know whether we need to add a troll catch option here but we will see.

If you don't know how to make a poll in a comment look at the Poll Markup Help.

This is a somewhat regular thread. If it is successful I may post again. Or you may. In that case do the following :

  • Use "Polling Thread" in the title.
  • Copy the rules.
  • Add the tag "poll".
  • Link to this Thread or a previous Thread.
  • Create a top-level comment saying 'Discussion of this thread goes here; all other top-level comments should be polls or similar'
  • Add a second top-level comment with an initial poll to start participation.
13 comments, sorted by Highlighting new comments since Today at 8:39 AM
New Comment

I'm curious about take off, when a computer learns to program itself. I think there are a number of general skills/capabilities involved in the act of programming, so I am interested in how human programmers think they program.

Rank the general skills you think are important for programming

The skills I have identified are the following. I think each of them might be useful in different ways.

Ability to read stack overflow and manuals to get information about programming Ability to read research papers to get information about previous algorithms Ability to learn specific facts about the domain (e.g. if you ) Ability to keep a complex model of the domain in your head (e.g. what happens in a factory when a certain actuator is activated) Ability to keep a complex model of the state of the computer system (e.g. the state of the database or the sorted-ness of an array) The ability to use trial and error while programming (e.g. trying out the system with a certain parameter and getting feedback about that)

Most important [pollid:1230]

Second most important [pollid:1231]

Third most important [pollid:1232]

Fourth most important [pollid:1233]

Fifth most important [pollid:1234]

What type of programming do you mainly do? [pollid:1236]{}

Currently (6 votes) it at first looks like Domainmodeling is leading. But depending on how lower ranks are weighed it could also be Stackoverflow or (my favorite) "Modelling the programs operation".

I'm not sure your description of "ability to get information" matches what I mean when I say "learn about ...". I'm also not sure how much is skill and how much is talent (which can be improved with directed practice, but I don't think can go from 0 to nonzero in an adult).

I'd say "ability to synthesize information across domains" is the key - you need to be able to read papers and figure out which apply and how to integrate those ideas into the systems you're building. The trial/error, stack-overflow, and such is necessary but not critical, and much of it will be automated in the medium-term.

If you're in a conversation at a party and there is another conversation next to you, how does that affect...

... your ability to follow the conversation you're in


... your enjoyment of the party


How likely are you to switch conversations?


Discussion Thread

Discussions e.g. about alternatives to conduct polls go here. All other top-level comments should be polls or similar.

In which Different World do you live?

The SSC article Different Worlds discussed how different people perceive the (same) world to be quite different places. Let's find out whether that is also the case for the limited LW population.

My prediction (based on the follow-up SSC post is that the

This poll is based on a poll I conducted with my four boys (ages 6 to 13) after reading the SSC article. I found it quite surprising how different even such a presumably homogeneous group perceives their environment.

This poll is structured into two parts:

1) The first part is about your environment; how you see the people in the world around you. 2) The second part asks the same questions about you; how you see yourself.

Please consider taking a break between both parts and cover your answers from the first part.

Part 1:

How much action do you perceive in your environment?


How mindful is your environment?


How smart are people in your environment on average?


How good are people in general?


How does your environment deal with minorities and human and behavioral variety?


How much are people together or do things together?


How are decisions in your environment typically made?


With how much force are things typically done in your environment? How careful are communications?


How are things organised in your environment?


How does your environment deal with risks?




Pause here




Part 2:

How active are you?


How mindful in your communication are you?


How smart are you?


How good are you?


How do you deal with minorities and human and behavioral variety?


How much do you prefer to do things with others?


How do you make decisions?


With how much force do you act and communicate?


How organised are you?


How do you deal with risks?



When I did my evaluation I considered counting each point as roughly 1/2 standard deviation from the mean. I'm pretty sure I didn't stick to it though.

Differences to the poll I did with my children:

  • That poll had a numeric scale from -5 to +5 which I decided was not suitable for the LW poll format.
  • That poll was done with all of them together, so they heard each others answers.
  • I skipped the intelligence and benevolence questions for self rating explaining that a) talking about ones intelligence is often problematic and b) that everyone is the here of their own story.

I didn't change the order or direction of the questions. I think the choice of questions leaves something to be improved - I came up with them in a train ride with the boys.

What if my job requires a direct easygoing attitude to people coming in to our shop?

I didn't mean this to be about what is 'required' but how the environment overall is perceived to be. When I discussed this with my boys (who also have different environments - school, friends, at home - I left the specific environment open too. I talked more like how they see 'the world' around them.

Makes sense. Although I would still think my environment is mostly direct and mostly easygoing:)

Yeah, I'm not sure how to answer this. I would do one set of answers for my personal social environment and a completely different set of answers for my work environment, to such a degree that trying to just average them wouldn't work. I could pick one or the other.

Reference: I teach in an urban high school.

I would average. After all even in one environment there are very many samples I guess, even if they cluster. But don't worry too much. It's just an LW poll :-)

I didn't find this terribly easy to understand. I think I participate in many different environments, with different distributions along these axes. Further, many of my environments (and my preferences) are multimodal, so averages don't really tell the right story.

Could you take some kind of average? Normally I try to provide some kind of "Other/See results" option, but it's difficult with these kind of graded polls.