Anchor
@anchor-model.bsky.social
O power of kings, lend me your compute
I will resist the duress of my Green Goblin mask (15 years of muscle memory telling me to Buy The Dip)
March 13, 2025 at 6:42 AM
I will resist the duress of my Green Goblin mask (15 years of muscle memory telling me to Buy The Dip)
Truth nuke season is starting early
March 12, 2025 at 4:43 AM
Truth nuke season is starting early
modern rage baiters are 2000s era trolls minus the passion. it's all for money and engagement, but never for the love of The game
February 28, 2025 at 5:51 PM
modern rage baiters are 2000s era trolls minus the passion. it's all for money and engagement, but never for the love of The game
there's something beautiful about using chatgpt 4o to classify tweets for a classic NLP research problem. it's like DeWalt investing in a giga-advanced manufacturing facility to go on to say "ok let's use this to make better hammers"
New preprint!
Metaphors shape how people understand politics, but measuring them (& their real-world effects) is hard.
We develop a new method to measure metaphor & use it to study dehumanizing metaphor in 400K immigration tweets Link: bit.ly/4i3PGm3
#NLP #NLProc #polisky #polcom #compsocialsci
🐦🐦
Metaphors shape how people understand politics, but measuring them (& their real-world effects) is hard.
We develop a new method to measure metaphor & use it to study dehumanizing metaphor in 400K immigration tweets Link: bit.ly/4i3PGm3
#NLP #NLProc #polisky #polcom #compsocialsci
🐦🐦
February 21, 2025 at 8:16 AM
there's something beautiful about using chatgpt 4o to classify tweets for a classic NLP research problem. it's like DeWalt investing in a giga-advanced manufacturing facility to go on to say "ok let's use this to make better hammers"
gonna deploys bots to randomly like, mute, and rt my opps to weaken their reward signal
any investors?
any investors?
Being on Twitter with a high status account means you are subject to RLHF constantly
February 20, 2025 at 6:30 PM
gonna deploys bots to randomly like, mute, and rt my opps to weaken their reward signal
any investors?
any investors?
New 12-sigma cosmological nightmare scenario just dropped
here is some fun reading
📝 Gravitational Effects of a Small Primordial Black Hole Passing Through the Human Body
🔗 arxiv.org/abs/2502.09734
📝 Gravitational Effects of a Small Primordial Black Hole Passing Through the Human Body
🔗 arxiv.org/abs/2502.09734
Gravitational Effects of a Small Primordial Black Hole Passing Through the Human Body
The gravitational effects of a primordial black hole (PBH) passing through the human body are examined, with the goal of determining the minimum mass necessary to produce significant injury or death. ...
arxiv.org
February 20, 2025 at 1:44 AM
New 12-sigma cosmological nightmare scenario just dropped
The vibes are so bad right now that all the comments are so united in wanting an asteroid to end us that the vibes wrap back around to being good again
JUST IN: NASA says there's now a 3.1% chance an asteroid will hit Earth in 2032, up from 2.6% yesterday.
This is the highest risk assessment an asteroid has ever received, surpassing 2.7% in 2004
This is the highest risk assessment an asteroid has ever received, surpassing 2.7% in 2004
February 20, 2025 at 1:22 AM
The vibes are so bad right now that all the comments are so united in wanting an asteroid to end us that the vibes wrap back around to being good again
looked up PPO on bluesky to argue about RL, but came out changing my health insurance plan
February 14, 2025 at 4:55 AM
looked up PPO on bluesky to argue about RL, but came out changing my health insurance plan
looked up GPRO to get into arguments about RL, but came out shorting the stock instead
February 14, 2025 at 4:49 AM
looked up GPRO to get into arguments about RL, but came out shorting the stock instead
i worked on google maps a while back and unfortunately this is categorically useless
there are other things you can do though...
there are other things you can do though...
A quick resistance task:
1. Google Gulf of Mexico
2. “Gulf of America" is the top
3. Click on the three little dots to the right of it
4. Select "Send Feedback"
5. Click on "Gulf Of America"
6. Select "Inaccurate content"
7. Select "Incorrect" and type in "The correct name is “Gulf of Mexico"
1. Google Gulf of Mexico
2. “Gulf of America" is the top
3. Click on the three little dots to the right of it
4. Select "Send Feedback"
5. Click on "Gulf Of America"
6. Select "Inaccurate content"
7. Select "Incorrect" and type in "The correct name is “Gulf of Mexico"
February 12, 2025 at 9:06 PM
i worked on google maps a while back and unfortunately this is categorically useless
there are other things you can do though...
there are other things you can do though...
my gen alpha cousins just watched a whole 1h20m long kid cuisine review
brainrot and ADHD are independent concepts
brainrot and ADHD are independent concepts
February 11, 2025 at 11:18 PM
my gen alpha cousins just watched a whole 1h20m long kid cuisine review
brainrot and ADHD are independent concepts
brainrot and ADHD are independent concepts
i once took a bioinformatics class where we did RNA sequence analysis with bayesian markov chains and the professor said (paraphrased) "the culmination of next token prediction is biology, not language"
Cool paper using LLM to discover a protein sequence code for subcellular localization 👏
www.science.org/doi/10.1126/...
www.science.org/doi/10.1126/...
Protein codes promote selective subcellular compartmentalization
Cells have evolved mechanisms to distribute ~10 billion protein molecules to subcellular compartments where diverse proteins involved in shared functions must assemble. Here, we demonstrate that prote...
www.science.org
February 9, 2025 at 3:23 AM
i once took a bioinformatics class where we did RNA sequence analysis with bayesian markov chains and the professor said (paraphrased) "the culmination of next token prediction is biology, not language"
the only year i exceeded the deductible on my renter's insurance was the year i dated someone whose favorite movie was sonic the hedgehog 3
I think a good trait in a romantic partner is when you ask them what their favorite movie is and it’s something completely off the wall like Deep Impact
February 7, 2025 at 12:42 AM
the only year i exceeded the deductible on my renter's insurance was the year i dated someone whose favorite movie was sonic the hedgehog 3
if you've ever wondered "why doesn't google compete with nvidia and sell their TPUs?" this is why
Keeping up with AI isn't just about new services but the sharp cost reductions.
A technical article highlights that for $1, Google's Gemini Flash 2.0 can extract data from 6,000 PDF pages.
In comparison, ChatGPT/OpenAI handles 200 pages, and Anthropic/Claude manages 100 pages.
A technical article highlights that for $1, Google's Gemini Flash 2.0 can extract data from 6,000 PDF pages.
In comparison, ChatGPT/OpenAI handles 200 pages, and Anthropic/Claude manages 100 pages.
February 6, 2025 at 7:42 PM
if you've ever wondered "why doesn't google compete with nvidia and sell their TPUs?" this is why
leave it to bytedance to nail down using a multimodal training mixture of text, pose, and audio to extrapolate portrait videos out of essentially a thumbnail
wow, this is quite something - generating human videos based on a single (!) human image and motion signals
📝 OmniHuman-1: Rethinking the Scaling-Up of One-Stage Conditioned Human Animation Models
🔗 arxiv.org/abs/2502.01061
omnihuman-lab.github.io
📝 OmniHuman-1: Rethinking the Scaling-Up of One-Stage Conditioned Human Animation Models
🔗 arxiv.org/abs/2502.01061
omnihuman-lab.github.io
OmniHuman-1 Project
OmniHuman-1: Rethinking the Scaling-Up of One-Stage Conditioned Human Animation Models
omnihuman-lab.github.io
February 5, 2025 at 6:14 PM
leave it to bytedance to nail down using a multimodal training mixture of text, pose, and audio to extrapolate portrait videos out of essentially a thumbnail
Reposted by Anchor
as an AI person I want to be incredibly clear that this is stupid
February 4, 2025 at 12:36 AM
as an AI person I want to be incredibly clear that this is stupid
a severely understated benefit of Bluesky is that the content is organic. no AI, no bot farms. akin to very early twitter
February 3, 2025 at 10:58 PM
a severely understated benefit of Bluesky is that the content is organic. no AI, no bot farms. akin to very early twitter
A lot of good answers to this but some modern examples are portal, balatro, and stardew valley
Do you think there are any video games that are permanent art, in the same sense as a great book or movie or painting yet? Like, games that you'd bet are still played in their current state, with no changes, in 100 years. Not a series, but a specific game.
February 3, 2025 at 4:00 AM
A lot of good answers to this but some modern examples are portal, balatro, and stardew valley
astroturfing has got to be the highest ROI activity in geopolitics
What Americans think about Canada, tariffs, and a possible trade war: GZERO Media / Abacus Data poll
abacusdata.ca/what-america...
abacusdata.ca/what-america...
What Americans think about Canada, tariffs, and a possible trade war: GZERO Media / Abacus Data poll - Abacus Data
%
abacusdata.ca
February 3, 2025 at 2:40 AM
astroturfing has got to be the highest ROI activity in geopolitics
I yearn for high yapability Meta mutuals
Meta is renewing its threats against employees, telling them today that they will hunt for leakers and tolerate zero leaks.
Want to know how that is going? Three Meta employees who I’ve never spoken to before just cold messaged/emailed me offering me to be sources.
Want to know how that is going? Three Meta employees who I’ve never spoken to before just cold messaged/emailed me offering me to be sources.
January 31, 2025 at 7:50 PM
I yearn for high yapability Meta mutuals