AMA: We are the Google Brain team. We'd love to answer your questions about machine learning.

https://www.reddit.com/r/MachineLearning/comments/4w6tsv/ama_we_are_the_google_brain_team_wed_love_to/

i missed this, even while reading askscience then...

Taken in isolation, part 1 left me confused. Part 2 greatly improves the value of the transcript.

Zebra is dealing with a set of problems, which may or not stem from a single Problem. I saw your questions as an effort, not only to clarify the issues, but also determine the structure of the problem set. Here, you describe the problem spiral, where solving one issue raises further issues because you keep thinking of the whole set instead of taking each issue in isolation.

Note there are two seemingly conflicting strategies here. One is to solve part of the problem, focusing on it for a given time, trying to jumpstart a success spiral. But how would you differentiate this from bikeshedding? How can you be sure you're not focusing on irrelevant things?

On the other hand, carefully thinking about the Problem, and how to solve it all at once, or by the correct sequence of actions, may lead to an efficient strategy. However, taking on the Problem as a whole shatters all motivation, warps the view of the problem set and makes small tasks seems undoable.

You linked Zebra, in the transcript, to Nate Soares's Replacing Guilt series. I would pinpoint my advice with Moving towards the goal. Solving the Problem (the cluster of problems, taken as a single eldritch entity) has to be set aside. You can picture the goal state, but the planning fallacy goes both ways: you can overestimate the difficulty of the solution. What matters is to make progress.

(I may have paraprashed your post here - I wanted to write down my own understanding of it)

What are rationalist presumptions?

I am new to this rationality and Bayesian ways of thinking. I am reading the sequence, but I have few questions along the way. These questions is from the first article (http://lesswrong.com/lw/31/what_do_we_mean_by_rationality/)

Epistemic rationality

I suppose we do presume things, like we are not dreaming/under global and permanent illusion by a demon/a brain in a vat/in a Truman show/in a matrix. And, sufficiently frequently, you mean what I think you meant. I am wondering, if there is a list of things that rationalist presume and take for granted without further proof. Are there anything that is self evident?

Instrumental rationality

Sometimes a value could derive from other value. (e.g. I do not value monarchy because I hold the value that all men are created equal). But either we have circular values or we take some value to be evident (We hold these truths to be self-evident, that all men are created equal). I think circular values make no sense. So my question is, what are the values that most rationalists agree to be intrinsically valuable, or self evident, or could be presumed to be valuable in and of itself?

(meta: I'm not sure if I should make a Discussion post for this, so I'm posting here. Feedback most welcome)

I'm exploring the following hypothesis : sometimes, you have to give up constructive actions for the sake of focus.

Most productivity methods suggest the obvious, to keep wasteful activities in check. It could be gaming, chatting, checking news websites. They all share a common trait: you don't gain any significant utility (nor money, nor fun, nor rest) for spending more time on it. You achieve the same result by spending a little time on it, rather than a full day.

With productive activites, time spent and value created aren't proportional. Sometimes you're lacking energy, inspiration, and it's okay: you don't have to work yourself ragged.

If you have multiple tasks to be achieved in parallel, you should treat them as sequential anyway. Focusing on one task at a time yields better results than task switching all the time.

Problems arise when you find inspiration, or a sudden peak of interest for a certain task which is, useful in isolation, but which doesn't fit in your schedule. Maybe a discussion with a friend sparked the idea of a story to write. Maybe you're considering moving some furniture because you're well-rested and full of energy.

Even if you could be maximally productive for a given useful task, you should treat it as a wasteful activity if you have something else you planned to do. If the idea sounds good, write it down. If it's really good, hype will come back another time. If you're energetic, do the most physical thing you had planned to do. Energy will come back another time.

The goal is not to add another task on your current schedule and mess with the plan you've set for the day, like you'd do with "classical" wasteful activities. You can convince yourself easily that news websites can wait another day. The unintuitive part is this also holds with most productive activities, even though you're training yourself to not defer work!

When you update, you're not simply imagining what you would believe in a world where E was true, you're changing your actual beliefs about this world. The point of updates is to change your behavior in response to evidence. I'm not going to change my behavior in this world simply because I'm imagining what I would believe in a hypothetical world where E is definitely true. I'm going to change my behavior because observation has led me to change the credence I attach to E being true in this world.

Consider P(E) = 1/3. We can consider three worlds, W1, W2 and W3, all with the same probability, with E being true in W3 only. Placing yourself in W3, you can evaluate the probability of H while updating P(E) = 1 (because you're placing yourself in the world where E is true with certainty.

In the same way, by placing yourself in W1 and W2, you evaluate H with P(E) = 0.

The thing is, you're "updating" on an hypothetical fact. You're not certain of being in W1, W2, or W3. So you're not actually updating, you're artificially considering a world where the probabilities are shifted to 0 or 1, and weighting the outcomes by the probabilities of that world happening.

[Survey Taken Thread]

By ancient tradition, if you take the survey you may comment saying you have done so here, and people will upvote you and you will get karma.

Let's make these comments a reply to this post. That way we continue the tradition, but keep the discussion a bit cleaner.

I really enjoyed this post. The process of making things, and the motivation behind interests me. Also, congrats on making this whiteboard!

Did you, at any point, thought "is this really worth my time"? From your description, I suppose the fun justifies the whole thing, and the fact that you made a usable thing adds to the value. I'm often overthinking fun and how I spend my time, so I wonder how to mitigate these feelings of "am I doing the right choice doing that at all?"

Post in this thread if you live in France.

The burning is the unsatisfied desire for sex, and lifting the branch is offering sex. At the end of the story, the boy goes to prison for attempted rape. I presume you were joking in saying that you did not recognize this, or that you simply intended to say that you consider it a bad analogy.

In any case, I agree that such an analogy is pointless, and that is why I downvoted the post.