All of dybypy's Comments + Replies

dybypy20

In deep learning, Goodfellow writes in response to NFL

The philosophy of deep learning in general ... is that a wide range of tasks (such as all the intellectual tasks people can do) may all be solved effectively using very general-purpose forms of regularization

Where regularization is changes made to an algorithm to reduce generalization error but not training error, and are basically one type of assumption made about the task to be learned.

3Adrià Garriga-alonso
Good find! Yeah, this is a good explanation for learning, and the NFL razor does not discard it. I think that almost no deep learning professor believes the bad explanation that “deep learning works because NNs are universal approximators”. But it’s more common with students and non-experts (I believed it for a while!)
dybypy60

I've done something similar to this

I found my overconsumption of reddit and YouTube distasteful as I was spending a majority of my day on these sites mindless scrolling or refreshing.

I ended up cutting them out entirely for about two weeks and felt much better, working through a backlog of books instead. but have since relapsed

I like the idea of writing summaries I think I will commit myself to that.

1Willa
My book reading habit has definitely felt the benefits of this experiment :) I'm reading more books and am quite happy about that. Writing summaries for YouTube videos helped me watch less videos in total, but also improved my filtering (intuitions or memory to help predict what videos are more or less worthwhile to watch) so I now watch less total videos of higher quality than I used to. If you commit to writing summaries, let me know how that goes! I'm curious to see if the summary writing is as helpful for you as it has been for me (everyone's different so seeing if that intervention works for > n = 1 would be nice).
dybypy110

Isn't this horses?

It sounds like you want a way to set up a system which autonomously creates a product that fills some certain parameters that you set.  In this case you want self building/designing cars.  

I feel like the alien perspective already answers your question of how.  Car manufacturers are already these self modifying car building systems, with humans as an integral part.

I don't think we can get to "growing cars" without humans as the substrate until something else is able to fill that role such as AI, or in the case of horses, natural selection.

dybypy40

Could the time worked be the number you gamify? Though that might incentivize taking no breaks.

2bfinn
Indeed. For years I’ve set myself a target for time worked per day, which I track in a spreadsheet. I don’t always hit it each day, but try to ensure my monthly average does.
dybypy90

I'm not sure what points can actually be made by this line of reasoning.  What I mean is taking the naive view throws out any useful information and leaves us with the status quo.

And you conclude with boycott-itarianism + the status quo essentially with each side cancelling each other out in all other subgroups.  The boycott-itarianism addition comes mainly from the assumption of utilitarianism (since we're talking about EA and all).

Let's use this reasoning in a different situation e.g. human population.

Whether a human life is net negative or net... (read more)

dybypy10

So you got better at management by dropping most of the programming work you used to do and instead delegating it to others?