laarc.io. Looking for AI work. DMs open. ML discord: discordapp.com/invite/x52Xz3… projects: status.shawwn.com

Lake St Louis, MO
Joined January 2009
Last night, someone asked me what I've been up to since 2010. My reply turned into a short autobiography. I considered deleting it, but people encouraged me to post it instead: gist.github.com/shawwn/3110a… If you're unhappy with your life, it's important to believe you can fix it.
13
6
1
163
oops, but
1
2
1
15
395
Show this thread
Shawn Presser retweeted
I'm regularly floored by all the amazing things people are doing with JAX, and I'm so glad we can do our little bit to help researchers push the boundaries of humanity's knowledge and capabilities. Nice work @shoyer et al.! nextplatform.com/2021/10/18/…
0
8
2
79
Shawn Presser retweeted
gfycat.com/disgustingunfinis… rotate_3d: [cos(radians(6)),0,-sin(radians(6))/sqrt(2),sin(radians(6))/sqrt(2)] translate_x: -1700*sin(radians(6))
8
25
4
213
3,688
Shawn Presser retweeted
ClipArtify :p
1
3
0
32
633
I'm genuinely surprised at this result. It's hard to imagine ever leaving twitter. I don't really have a social circle. I just treat twitter like the Vietnam war, where friends randomly disappear. Yet the "75-100%" answer was just 3%. You wouldn't stay?
how many people in your social circle would have to use a less corrupt twitter competitor before you switch? pls elaborate in comments
7
1
1
10
Shawn Presser retweeted
Here's a Business Insider story about how a load more anonymous Google employees are censoring AI research, allegedly crossing out refs to things like fairness. businessinsider.com/google-e… This recent trend has directly impacted me. A thread...
4
80
17
239
Show this thread
One of my mentors wrote over 500 scientific papers. She taught me one remarkable tactic: She wouldn't write her own papers. She only SENT her ideas TO COLLEAGUES. I wrote over 1,000 papers using the same technique. Step by step🧵👇🏻👇🏻👇🏻
One of my mentors raised over $500 billion and went public. He taught me one remarkable fundraising tactic: He wouldn't accept money wired from the VC firm. He only SENT wire transfers TO VCs. I raised over $100 billion using the same technique. Step by step🧵👇🏻👇🏻👇🏻
5
3
1
34
It’s always impressive seeing someone commit to doing something every single day. She’s logged 57 straight days of ML hacking. Apparently the goal is 66 days. I can barely commit to something for 66 minutes, let alone 66 days.
Day 32/66 of #66daysofdata: Went down the JAX rabbit hole.
0
0
0
9
Shawn Presser retweeted
All polynomials with coefficients ±1, orders from 3 to 6. The values are Tanh-scaled.
0
6
0
30
Dear dev twitter: Is there an open source license like BSD-2-Clause, which also covers program *outputs*? BSD2C requires that people credit the original repo. But with NFTs, they're selling program outputs, not "using the code in a bigger code project." What are the options?
Replying to @theshawwn
that would be awesome! it would benefit a number of generative artists who are going through the same problem
9
5
2
29
Unfortunately, this idea can't work. It's an open legal question whether a developer can copyright program outputs: nitter.net/alexjc/status/14… The case quoted in the article mentions computer art programs.
Applying a copyright license that covers the output of software is an open legal question, so far that failed in the U.S. on the Ninth Circuit. internetandtechnologylaw.com…
3
0
0
1
Adobe doesn't get copyright over all Photoshop outputs. Why should the developer of a generative art program be able to wield copyright over any generated outputs? I guess they shouldn't. How unsatisfying. There's no way to prohibit NFTs. Sorry @marceloprates_.
3
1
1
4
I'm not sure your closed-source idea would stop NFTs, either. If they can use your program at all, they can sell the output. When people are taking advantage of you, and there's nothing you can do about it, it's usually best to ignore them.
5
1
0
2
This is an insightful thread. Anyone who's interested in ML should consider following them (@KimboChen). They have 14 followers, and just joined this month. I can tell they're taking twitter seriously, and I remember how hard it was for me when I had very few followers.
Inspired by Aleksa's informative Tweet, I want to do something similar. Here's a TL;DR on the design choices of @Tesla 's Dojo Computer: 1 / N
Show this thread
1
0
0
7
Here's another thread by them:
I attended ICCV Undergraduates in Computer Vision held by @_sam_sinha_ , Stephen Tian, Alex Yu, and @cchoi314 . Here are the things that I learned from Prof. @hugo_larochelle 's AMA, the student panel's talk, and Dr. @rosemary_ke 's talk. A thread: 1 / N
Show this thread
0
0
0
0
This isn’t the way open source works. You can’t release something with a permissive license and then say you can’t sell it for profit. All that matters is the terms of the license. I don’t personally condone selling open source work that you didn’t put effort into, though.
2
0
0
18
Now, I understand he’s not saying “they’re not allowed to do this.” However, he is saying “It was not my intention…” That’s the part about open source. You can’t have “intent” when making something open. You can only have a license. This is their confusion. Licenses matter.
4
0
0
12