rw

8 ms·

31.
by rw 11y ago
Is your thesis online where we can look at it?
32.
by rw 11y ago
GPUs are best under SIMD conditions: single instruction, multiple data. You're talking about running `eval` thousands of times. Each unit of execution is going to have different data, because each process is executing different code (e
33.
by rw 11y ago
Fair comparison, but I didn't intend any monopoly on fear. All I'm saying is that the risk of GAI should be taken seriously, calibrated along with the other problems you mention. There's a lot of room between 'irrelevant
34.
by rw 11y ago
I agree that the rate of progress in AI is unpredictable, which means we are probably not right on the cusp of superhuman AI. But what if we actually are on the precipice? How could you tell? You seem to be taking a bet on the following s
35.
BlindTool: Phone App That Audibly Identifies Objects (with Neural Networks)(laughingsquid.com)
4 points by rw 11y ago|1 comments
36.
Facebook uses FlatBuffers on one billion Android devices(rwinslow.com)
3 points by rw 11y ago|0 comments
37.
by rw 11y ago
Please read the comments on the post. We addressed the misunderstanding.
38.
Facebook uses FlatBuffers on one billion Android devices(rwinslow.com)
7 points by rw 11y ago|3 comments
39.
Show HN: Use FlatBuffers in Go(rwinslow.com)
10 points by rw 11y ago|0 comments
40.
Use FlatBuffers in Go(rwinslow.com)
1 points by rw 11y ago|0 comments
41.
by rw 11y ago
How does it do under SMHasher (a library to test hash quality)?
42.
by rw 11y ago
I liked this writeup. I'm curious how much this improved battery life for a typical user. FWIW, I help maintain FlatBuffers for Go and Python. I'm happy to answer any questions I can.
43.
by rw 11y ago
What data structures does it use?
44.
by rw 11y ago
Hyperinflation sequence for banknotes: http://oeis.org/A051109
45.
by rw 11y ago
Badly titled. This is not an official library.
46.
Kaggle: Restaurant Revenue Prediction has a submission with zero error(kaggle.com)
3 points by rw 11y ago|2 comments
47.
by rw 12y ago
The biggest difference is that self-teaching implies more variability in skill outcomes. Lower lows, higher highs. Look for people who can learn the structure of a problem when there's no guidebook (or textbook) available. If they can
48.
by rw 12y ago
Previously: https://news.ycombinator.com/item?id=6427525 Pardon the essay, but I've written a tool in this space before. Back in 2011, I wrote a textual steganography library and command-line application, called Plains
49.
by rw 12y ago
I used to dream in code, but now I dream in graphs that represent data structures. This happened when I began focusing on memory layout over syntax and API signatures.
50.
FlatBuffers: a memory efficient serialization library(google-opensource.blogspot.com)
126 points by rw 12y ago|62 comments
51.
Semantic Compression [in Programming](mollyrocket.com)
11 points by rw 12y ago|0 comments
52.
The Impact of Childhood Lead Exposure on Crime(nber.org)
1 points by rw 12y ago|0 comments
53.
Big, Small, Hot or Cold - Data Pipeline Examples at Stripe, Tapad, Etsy, Square(hakkalabs.co)
12 points by rw 13y ago|0 comments
54.
Tonight at Square SF: Metamarkets presents Druid(meetup.com)
3 points by rw 13y ago|0 comments
55.
by rw 13y ago
One serious use case is to seed the generator with a spam email corpus. This lets you generate messages that look like spam. Example: wget https://spamassassin.apache.org/publiccorpus/20030228_spam.tar.bz2 tar -j
56.
by rw 13y ago
TweetFS uses the SeqTweet library, which takes care of sequencing the tweets for you. Specifically, see the _list_to_twitter method: https://github.com/workmajj/seqtweet/blob/master/seqtweet/se... O
57.
by rw 13y ago
I wrote a textual steganography library and CLI in 2011, called Plainsight: https://github.com/rw/plainsight Additionally, @workmajj and I wrote TweetFS using Plainsight. It lets you recursively pack up directories and
58.
Google interprets 'Linux' as 'Lenovo' in this query(encrypted.google.com)
1 points by rw 13y ago|2 comments
59.
by rw 14y ago
QLabs, New York, NY Front-end developer About us: QLabs is a tech incubator based in Noho, NY. We’re a small group of hackers who rapidly prototype new products every 10 weeks, then test them in the market for long-term potential. Our prima
60.
U.S. Sen. Schumer Asks Google And Apple About Their "Spy Planes"(techcrunch.com)
1 points by rw 14y ago|0 comments

More ›