poppingtonic

Street Fighting Mathematics! Loved taking this Sanjoy Mahajan's course on OCW in high school

Some zig zag motions are required to maintain velocity and the combat is fun. I like it!

Learning how to implement Prolog in pg's On Lisp was a fun way to spend multiple weeks programming. Doing this again this year should be a lot of fun.

Lambda, The Ultimate Container

This essay nerd sniped me hard into information theory. Absolutely love it.

Thank you for the links - I can't stop replaying Love

Reading textbooks.

Using a REPL-based language like Lisp (which I learned from reading advanced Lisp textbooks) for several years before I worked professionally as a programmer enabled me to dive into complex, underexplored parts of the (Python, C) projects at my first job, which meant I basically owned some very important parts of the stack.

Reading and applying ideas from Unix sysadmin books enabled me to be comfortable jumping around the OS very quickly, giving me a speed advantage even when the computer I used was pretty underpowered, relative to my colleagues, early in my first job.

Hello HN, co-founder (Brian Muhia), here to answer any questions!

1. You get a state-of-the art training loop that includes the latest research findings such as 1-cycle.

2. Yes. You can make everything project-specific if you want to. e.g. Image segmentation datasets are just slightly modified Dataset classes.

3. fastai extends PyTorch in a very Pythonic OO sense, so I think the only speed issues could come from that, and maybe maintaining a few extra dicts in memory. If it's about i/o, probably not, if it's about parallelization (for nlp), definitely not. In fact, I can't think of a way in which there's a significant speed penalty.

Try a schedule that keeps you on ~ two hours a day, focusing on at least one of the videos in the 14-video series. So: 14 weeks, paying attention to one video a week and whatever peripheral studies you choose to follow up on based on the ideas Jeremy pushes in his lectures.

Edit: also, find datasets online that fit in your interests. There's no better motivation than to solve problems in a domain that you're already familiar with. I grew up on a farm and built a classifier to detect diseases as part of my fast.ai course-work: https://t.me/shambadoctorbot

Wow, when I hear about perverse incentives, I don't usually think about people spending their lives showing their clients a taste of a possible future until a high enough reward gets them to stop caring. Reward hacking for humans.

Maybe if they randomized to n<=20 n-gons. I'd love to see Dactyl tackle a dodecahedron.

The conclusion of this paper is, paraphrasing, "We should stop using point estimates for what should be distributions. When we do that, we get much lower bounds and have a better chance at being more precise, whether or not we are alone." It opens up avenues for more questions, which is the correct stance given how little we know.

One year of Python, with high school and introductory college courses. Some basic linear algebra, familiarity with logarithms etc.

Givewell's founders know this, which is why they spun off the Open Philanthropy Project: https://www.openphilanthropy.org/focus/global-catastrophic-r...

I've read most of it, and I think it's a great book for experienced developers looking to dive deep into a pretty specific range of topics. As a beginner's second book after SICP, I'd still recommend it, because it assumes almost nothing. Having read it, I find it hard to see it as just a collection of intros, because you get a lot of mental algorithmic mileage out of e.g. hand-rolling an implementation of automatic differentiation, which is done in the book. You get a lot of clarity on things that are still quite useful.

Anaconda can probably help

Is this going to expand the funding that goes to the SENS program? There's a lot of basic Science that is yet to be done, and it appears (from Aubrey de Grey's talks) that a lot of it is still rather underfunded. See this talk: https://www.youtube.com/watch?v=S6ARUQ5LoUo

Which funding priority supersedes this, then? Funding the prevention of nuclear war, for example? Disaster preparedness for large cities?

No, probably evolution strategies.

https://blog.openai.com/evolution-strategies/

These two blog posts by K. Eric Drexler explain why this is important. It's related to learning subjects you have no familiarity with.

[1] http://metamodern.com/2009/05/27/how-to-learn-about-everythi...

[2] http://metamodern.com/2009/05/17/how-to-understand-everythin...

ArXiv itself does this. See the 'cite as' section of this one: https://arxiv.org/abs/1707.08475, which has the note

arXiv:1707.08475 [stat.ML] (or arXiv:1707.08475v1 [stat.ML] for this version)

This seems more like a versioning issue. Can't papers that pass review get a new DOI? I'm thinking something like semver for scientific papers or scientific data, more generally. DOIs even look like semver versions, so why not add more space for the different stages of publishing?

I just realized where the 'V' logo in "Viewpoints Research Institute" came from (you can see it here: [0]):

The diagram comes from the book "The Act of Creation"[1], where Arthur Koestler discusses the common themes in creative acts, including the idea of juxtaposition.

In research, you start with a baseline level of knowledge and do a random walk until you encounter a (not necessarily) totally orthogonal set of ideas. Once you combine the two, boom! Something new is created. A lovely book, that I'd recommend to anyone.

[0] http://www.vpri.org [1] https://en.wikipedia.org/wiki/The_Act_of_Creation

You could use SuperTuxCart. I'm setting this up, actually.

I'm really loving the choice of articles, especially since you're just getting started.

Edit: I'm referring to the journal, not the author.

I love this quote from the book "Origami Design Secrets":

  If you are a beginning designer, you should realize that no design is sacred. To learn to design, you must disregard reverence for another’s model, and be willing to pull it apart, fold it differently, change it and see the effects of your changes.

  Small ideas lead to big ideas; the concepts of design build upon one another.

You can also use Complice in a group setting, using rooms, if you think you might need accountability buddies.