willwhitney

7 ms·

1.
Computing Inside an AI(willwhitney.com)
1 points by willwhitney 2y ago|0 comments
2.
Learning 3D physics simulators from video(willwhitney.com)
2 points by willwhitney 3y ago|0 comments
3.
by willwhitney 5y ago
Would it be bad if we made advertising... illegal? In small part because it would reduce the principal-agent disparity between what you want to do and what Google wants you to do, but more importantly because it would remove the incentive t
4.
Parallelizing neural networks on one GPU with JAX(willwhitney.com)
1 points by willwhitney 6y ago|0 comments
5.
The New York Times(robrhinehart.com)
3 points by willwhitney 6y ago|1 comments
6.
by willwhitney 6y ago
That dip is more than 4M people, so I think it's still unexplained. Perhaps due to a shift in response rates or another artifact of COVID churn in the survey methodology?
7.
There’s no better time for cities to take space away from cars(theverge.com)
2 points by willwhitney 6y ago|0 comments
8.
by willwhitney 8y ago
You can actually drag and drop multiple notes into a tag, no find and replace necessary. Minor point though.
9.
by willwhitney 9y ago
They have a grammatical error ("deliciously looking") in their (2) step.
10.
by willwhitney 10y ago
Handling delays (and the uncertainty they entail) is a huge challenge, and I think it'll be a rich area of research. The simplest part of the problem is that delays in action or perception also slow the propagation of reward signals, a
11.
by willwhitney 10y ago
Yup, it's definitely an advantage to get all the correct values from the game state. But not as much as you might think; the vision portion of a DQN or similar trains quite quickly. Plus, our bot doesn't have any clue about projec
12.
Beating the World’s Best at Super Smash Bros. with Deep Reinforcement Learning(arxiv.org)
202 points by willwhitney 10y ago|55 comments
13.
by willwhitney 10y ago
Once a startup gets bought, I look for alternatives. If they don't shut down, I still learned something new! But usually they do. In this case, I landed on Fantastical [1] — it's a little pricy but completely worth it. (Mac/i
14.
Understanding Visual Concepts with Continuation Learning(willwhitney.github.io)
1 points by willwhitney 11y ago|0 comments
15.
by willwhitney 11y ago
I independently wrote a version of this for OS X: https://github.com/willwhitney/live-earth-desktop
16.
Ask HN: Blacklist Forbes?
164 points by willwhitney 11y ago|41 comments
17.
Popcorn time revived(reddit.com)
3 points by willwhitney 11y ago|0 comments
18.
by willwhitney 11y ago
Help.
19.
by willwhitney 11y ago
I'd love to do live variable inspection (like the Chrome Dev Tools) — it's something I'm thinking about for the future. There's a (nascent) project called `jmp` (for Jupyter Messaging Protocol) that implements kernel com
20.
by willwhitney 11y ago
:D I mean, if you're offering...
21.
by willwhitney 11y ago
Right now it only works with local kernels, but that's on the to do list for sure.
22.
by willwhitney 11y ago
The development experience for statically typed languages is often _amazing_ compared to that of dynamically typed ones; we sometimes choose dynamically typed ones anyway because of the flexibility and speed of development they give. I&#x27
23.
by willwhitney 11y ago
I haven't tried the Ruby kernel - let me know how it works, and post an issue if anything breaks!
24.
by willwhitney 11y ago
Thanks Bill! Atom has been improving really fast - it's my editor of choice now.
25.
by willwhitney 11y ago
I'm looking forward to that Elixir kernel!
26.
by willwhitney 11y ago
Jupyter provides a spec for a ZeroMQ-based messaging API, which kernels in each language have implemented. Hydrogen acts as a client to those kernels the same way that the Jupyter frontend does, sending execution requests to the kernel and
27.
by willwhitney 11y ago
I like to break my code up into little functions that are easy to test, so that flow works well for me. Even more powerful, though, is autocomplete from the running environment. It means that (just like in the Chrome Dev Tools) when you sta
28.
by willwhitney 11y ago
Thanks for the support! I've been thinking a lot about what makes for a good development experience, and how to properly "factorize" an IDE to provide a consistent experience across different languages. The monolithic IDE exp
29.
by willwhitney 11y ago
Most of the credit goes to the Jupyter team and the people who have implemented the kernels for various languages. The Jupyter API spec is what has made this possible. I'm really looking forward to what can be done next (hint: https:&
30.
by willwhitney 11y ago
I'm actively using it with Julia and Torch. It should work with any Jupyter kernel (and there are lots: https://github.com/ipython/ipython/wiki/IPython-kernels-for-... ). In practice, different kernels

More ›