Russian networks flood the Internet with propaganda, aiming to corrupt AI chatbots
A pro-Russia network is internally corrupting large-language models to reproduce disinformation and propaganda.
I know we're all focused on other terrible things right now but if you want a break, check out my latest coauthored publication for the Bulletin of Atomic Scientists.
It looks at our @americansunlight.org report on Russian LLM grooming and offers some solutions...
thebulletin.org/2025/03/russ...
26.03.2025 14:08 โ ๐ 63 ๐ 27 ๐ฌ 3 ๐ 3
I caught it now, at about the 2 minute mark she switches to Hebrew. Thanks for the share.
10.03.2025 02:32 โ ๐ 1 ๐ 0 ๐ฌ 1 ๐ 0
I heard it change to what I thought was French
09.03.2025 23:36 โ ๐ 0 ๐ 0 ๐ฌ 1 ๐ 0
More of this! We all know this wonโt pass but this is a line in the sand and a great talking point.
/cc @shasti.bsky.social
08.03.2025 01:28 โ ๐ 12874 ๐ 2217 ๐ฌ 86 ๐ 90
08.03.2025 06:08 โ ๐ 29764 ๐ 7148 ๐ฌ 360 ๐ 199
Great take!
07.03.2025 22:07 โ ๐ 2 ๐ 0 ๐ฌ 0 ๐ 0
JD is probably secretly jealous of Musk and he took it out on Zelinsky
01.03.2025 23:28 โ ๐ 0 ๐ 0 ๐ฌ 0 ๐ 0
I think I recall movies of Hitler acting like that and berating his oppents
28.02.2025 22:18 โ ๐ 0 ๐ 0 ๐ฌ 0 ๐ 0
But you don't have to exit the kaisatu do? I used to do it all the time just cause I liked riding the great trains. Maybe doesn't work anymore?
13.02.2025 21:24 โ ๐ 2 ๐ 0 ๐ฌ 0 ๐ 0
Does it cost anything to come back?
13.02.2025 21:17 โ ๐ 1 ๐ 0 ๐ฌ 1 ๐ 0
from: ใกใใฃใจๆใฎๆฅๆฌใฎๆฏ่ฒbot
Yurakucho Station, Tokyo, 1990
11.02.2025 23:42 โ ๐ 2 ๐ 0 ๐ฌ 0 ๐ 0
Bamboo, by Kuwayama Gyokushลซ, 18th century
Bamboo, by Kuwayama Gyokushลซ, 18th century
#Japanese #Art #JapanseArt #Ukiyo-e
30.01.2025 19:17 โ ๐ 34 ๐ 5 ๐ฌ 0 ๐ 0
If you are considering downloading DeepSeek, please read this first and share with others
You donโt owner your data. They keep it even if you delete and worse ..
29.01.2025 19:21 โ ๐ 13661 ๐ 4942 ๐ฌ 1668 ๐ 339
Your daily affirmation, from William Blake:
The tygers of wrath are wiser than the horses of instruction
Expect poison from the standing water.
You never know what is enough unless you know what is more than enough.
25.01.2025 17:03 โ ๐ 188 ๐ 14 ๐ฌ 7 ๐ 1
The reason A.I. is being shoved down our throats is shareholders want a return on their investment. Every company that integrates this crappy tech into their products knows you don't want it, but they've sunk too much money into it and don't care. Turn it off, don't use it, don't look back.
24.01.2025 02:20 โ ๐ 13860 ๐ 4131 ๐ฌ 83 ๐ 29
I am an ai scientist and I use ai every day. It makes many mistakes when discussing technical ai related questions. Many! It is good for me though sometimes because it makes me think about it's wrong answers to figure out why they are wrong.
24.01.2025 18:39 โ ๐ 0 ๐ 0 ๐ฌ 0 ๐ 0
Screenshot of a young person holding up a sign that says โBREAKING: ICE IN KENTUCKYโ
The algorithm-facing comment says, โHad to share these cute winter boots I just got!โ
People on TikTok are reporting ICE raids happening in small cities in various states.
To get past algorithmic censorship, theyโre writing down the information on paper and holding up and shuffling through the handwritten signs as they talk about pretend subjects like shopping, food, and animals.
23.01.2025 23:31 โ ๐ 21048 ๐ 5639 ๐ฌ 413 ๐ 496
accept with serenity
the things that cannot be changed,
Courage to change the things
which should be changed,
and the Wisdom to distinguish
20.01.2025 15:09 โ ๐ 4 ๐ 0 ๐ฌ 0 ๐ 0
Watching for it all the time
17.01.2025 03:52 โ ๐ 1 ๐ 0 ๐ฌ 0 ๐ 0
The Atlantic piece on Modi's India focuses on absolute problems while ignoring rates of improvement. To fairly assess his leadership, we need to compare development speeds pre and post-2014. Are roads, schools, jobs improving faster or slower? Raw problems without trends tell half the story
05.01.2025 17:08 โ ๐ 3 ๐ 0 ๐ฌ 0 ๐ 0
I THINK JAL also just said something else as well
22.12.2024 04:13 โ ๐ 2 ๐ 1 ๐ฌ 1 ๐ 0
๐๐ฏ
18.12.2024 02:39 โ ๐ 0 ๐ 0 ๐ฌ 0 ๐ 0
Could this be used for simultaneous translation do you think?
18.12.2024 02:19 โ ๐ 0 ๐ 0 ๐ฌ 1 ๐ 0
Do you need London based? I have students in D.C.
15.12.2024 04:51 โ ๐ 0 ๐ 0 ๐ฌ 1 ๐ 0
Wow!
14.12.2024 21:41 โ ๐ 1 ๐ 0 ๐ฌ 0 ๐ 0
๐ฒ๐ฝ - macroecology, evolution, conservation, spatial. #rstats, data. mammalogist. tidyverse and ๐จ instructor
Reproducible bugs are candies ๐ญ๐ฌ
MG author. Fledgling poet. Repโd by Nicole Eisenbraun @ Ginger Clark Literary Agency. WHERE THE LIGHT SHINES IN releasing fall 2026.
Website: jemattern.com
George S Pepper Professor of Public Health & Preventative Medicine; Biostats, Stats & Data Science, Lifelong learner & truth seeker; Views my own & not employerโs
Scientist who is passionate about correcting misinformation
https://www.youtube.com/backtothescience
IT-Analysis, Consulting and Development. Likes simplicity and appreciate sense of humor. CCC since 1998. He/Him. Interfacing: de/en/hr/hu
#TeamDatenschutz #BlueTeam
Other platforms: https://echobit.de
Write threads in a nice text editor, automatically chops it into chunks of less-than-300 chars, numbers them, and more. Add returns where you want a new post. Click button to publish your thread. RSS feed included. https://this.how/threadWriterForBluesky/
Research @OpenAI. I study Reinforcement Learning. PhD from UT Austin. Previously FAIR Paris, Meta US, NVIDIA, CMU, and IIT Kharagpur.
Website: https://hari-sikchi.github.io/
PhDing @UCSanDiego @NVIDIA @hillbot_ai on scalable robot learning and embodied AI. Co-founded @LuxAIChallenge to build AI competitions. @NSF GRFP fellow
http://stoneztao.com
PhD Student in Machine Learning at CMU.
๐ฆ twitter.com/steph_milani
๐ stephmilani.github.io
PhD at Machine Learning Department, Carnegie Mellon University | Interactive Decision Making | https://yudasong.github.io
PhD student | Interested in all things decision-making and learning
PhD at NYU studying reasoning, decision-making, and open-endedness
alum of MIT | prev: Google, MSR, MIT CoCoSci
https://upiterbarg.github.io/
In search of mathematics and ML content.
PhD@NYU
Reinforcement Learning PhD student, UPF Barcelona.
Uncertain in the face of optimism.
ahanadeb.github.io
ML Research @ Tzafon | Prev: Robot Learning & RL PhD @Technion
More data isn't all we need ๐ญ๐ฆพ ๐
PhD Student working on Generlization and State abstractions in #RL, #MetaLearning, and #AutoRL
amsks.github.io
RL researcher looking for DACs // What is this AutoRL anyway?
she/her
Currently: Leibniz Uni Hannover
Previously: Uni Freiburg (Master's) | Meta AI London (Intern)
Always & Forever: AutoRL.org
Stanford CS PhD working on RL and LLMs with Emma Brunskill and Chris Piech. Co-creator of Trace. Prev @GoogleDeepMind @MicrosoftResearch
Specifically
- Offline RL
- In-context RL
- Causality
https://anie.me/about
Unverified hot takes go to this account
For professional, see https://cvoelcker.de
If I seem very angry, check if I have been watered in the last 24 hours.
Now ๐บ๐ธ flavoured, previously available in ๐จ๐ฆ and ๐ฉ๐ช