Interests: China+ML, natsec+tech, brains+words+absurdity | Current: @CSETGeorgetown (opinions my own) | Former: @open_phil

Joined November 2014
Pinned Tweet
New paper out today! This one was written with my brilliant-but-twitterless colleague Zach Arnold, and (due to his influence) is especially fun. 🧵:
👏New Policy Brief👏 #AI accidents are already happening, but if AI continues its rapid rollout, they are sure to get more frequent and severe. CSET's Zachary Arnold + @hlntnr paint a picture of what that might look like, and what policy can do to help: cset.georgetown.edu/publicat…
Show this thread
2
9
1
39
Show this thread
Watching the livestream of @jackclarkSF at the @SeedAIOrg launch event, good stuff: seedai.org/event
1
2
0
20
When you say "AI" in DC, most people think "big data." It's not totally wrong, but this idea lags behind progress in the field. In @sciam today, @HsjChahal and I introduce one up and coming "small data" paradigm, transfer learning: 1/4 scientificamerican.com/artic…
2
25
3
80
To be clear, I wouldn't say the era of big data is over - we're still pushing the bounds of what giant models trained on giant compute clusters using giant datasets can do. But thinking of AI *only* in terms of big data is clearly not right either. 3/4
Very few people realize that there’s a massive tech paradigm switch happening between old-school ML (dare I say “big data”?) and new generation ML (transfer learning). All the tools/platforms/processes will be different from one to the other.
1
0
0
4
Shout-out to @maosbot, @mattsheehan88, @timrudner, @itschrispainter, @stuhlmueller and many others for helping us think through this project over the course of a year or more. And of course the bulk of the credit goes to lead author @HsjChahal! 🙌🏻 4/4
0
0
0
9
If you ever read CSET's stuff and would be down to do us a favor... please take this survey to help us understand our audience! georgetown.az1.qualtrics.com…
As CSET nears its 3rd birthday, we want to learn more about our audience and how we can better serve your interests. If you’re interested in helping, you can do so by filling out a short, anonymous survey: georgetown.az1.qualtrics.com… Thanks for your support!
0
0
0
0
Every non-engineer: So this is a nation-state attack right? Every SRE: Sometimes turning things back on is VERY HARD, OKAY
10
461
32
3,418
Show this thread
Helen Toner retweeted
Beautiful thread about emergent AI behaviour that nobody deliberately designed yet seems meaningful.
Replying to @JoelBurgess
Among Skyrim players, you'll occasionally see this tip: if you see a wild fox, follow it and you'll be led to treasure. Sometime shortly after shipping, we saw this going around online, and an informal investigation started. Who made foxes do this?!
Show this thread
3
66
1
238
Rainbow lorikeets, Australia's friendliest bird ambassadors!
Merkel visited a bird zoo today and the photos are everything. This isn't photoshop, it actually happened
Show this thread
0
0
0
11
Update 1: this was a false alarm 😑 unclear why I got that email, but I still had to do a medical exam so in retrospect it couldn't have been an approval Update 2: Submitted medical a few weeks ago, then today I received the exact same email... so maybe this time for real?! 🤞🤞
No big deal, just sitting here alone in my apartment sobbing because I got an email from USCIS that I *think* means my green card has been approved 😭
7
0
0
45
I'm thrilled and incredibly honored to share this news! I can't wait to work with @sama, @gdb, @ilyasut and the rest of the team to contribute to the mission of building AI that benefits all of humanity.
We're excited to announce @hlntnr of @CSETGeorgetown is joining our board of directors. Her deep understanding of AI policy will help us achieve our mission to deploy safe and responsible general-purpose AI. openai.com/blog/helen-toner-…
18
16
2
270
Saw a Silicon Valley dude tweet the other day about how China is doing all this crazy tech regulation stuff "and no one is talking about it" and just about lost it
every single time you see a “and no one is talking about it!” tweet, people are just telling you the shape of their bubble
Show this thread
1
0
0
36
Great thread on China's new draft regs for recommender systems. Had fun working with @DigiChn folks today on translation & commentary to add to Kendra's: digichina.stanford.edu/news/… All credit to @China_Digital for translating, and kudos to @gwbstr for making DigiChina so excellent.
My goodness. China's cyberspace watchdog, the CAC, just published a long (and unprecedented) set of draft regulations for recommendation algorithms. The short version: they will be tightly controlled. Key points below. 1/ cac.gov.cn/2021-08/27/c_1631…
Show this thread
1
6
0
19
Exactly 2.5 years ago we huddled nervously around a table, 3 days out from launching the CSET website and announcing ourselves to the world. I can't express how lovely it is to see our amazing team—*and* smart external folks who read our work—now celebrating this milestone. 🥰
📣 News Release! 📣 CSET is happy to announce an increase in grant funding, bringing our total funds to more than $100 million, to continue our work at the intersection of emerging tech and security. cset.georgetown.edu/article/…
Show this thread
1
0
1
84
Helen Toner retweeted
You guys. 🚨(SUPER FUN) JOB ALERT🚨We're hiring and it's a spectacular opportunity to work in #biosecurity with an amazing #NTIbio team. Spread the word and/or apply! @NTI_WMD 👉tinyurl.com/rkyn2s9d
1
26
2
30
No big deal, just sitting here alone in my apartment sobbing because I got an email from USCIS that I *think* means my green card has been approved 😭
22
1
1
340
Helen Toner retweeted
An impt Chinese gov think tank (CAICT) recently worked w/ e-commerce giant JD to put out China's first white paper on Trustworthy AI (可信AI). I translated the infographic w/ their framework for Trustworthy AI. Suggestions welcome. Full paper: caict.ac.cn/kxyj/qwfb/bps/20…
3
30
3
89
Show this thread