cyd
@chongyidong.bsky.social
I also post on Mastodon: https://kopiti.am/@cyd
Could have done better with a better domestic leader, but I think failing to understand how much life absolutely sucked before him will prevent you from understanding the somewhat complicated attitude a lot of even quite educated and well informed Chinese people have towards him
After China dominated the market for cheap drones, Chinese firms are selling cheap drone interceptors... www.ft.com/content/547e...
China bets on exports of cheap 3D-printed drone killers
Chinese companies rush into sector as global conflicts raise demand for interceptors
ft.com
A new project: data bridge between #deepseek harness and #emacs: github.com/seewhydee/ds... Also my first serious effort at Emacs dev in about 15 years. Mostly AI aided for now, but will dig in manual polishing once the grunt work's done.
There's a new animated film out in China called Niu Lai that was made by a man and his mother over 5 years without any investment, marketing, or promotion. It made $1,100 from a few hundred people in attendance over its first 10 days, with many shocked that this could make it to the big screen.
Calling it now, this "AI safety feature" is going to be turned into a creepy surveillance tool.
Jo Michell@jomichell.bsky.social · last mo.This is interesting: Claude is going to add 'watermarks' to its output. support.claude.com/en/articles/...
The editorial line for #TheEconomist has always been negative on China, but this issue is really something else. The cover and reporting goes one way; editorial pushes in opposite direction with a stream of "but at what cost" cliches.
Why are there suddenly so many shady companies with Tolkien-based names? And why do they all give the same vibes as Sauron forging the One Ring?
MiMo-2.5 turned down a job on an RPG project because it's high risk ;-) Not even mad, just curious how this happened.
I guess we're gonna have to refight all those 1990s battles over obnoxious freeware license terms all over again, including old-BSD style attribution clauses, commercial use restrictions, etc.
fry69@fry69.dev · 2mo agoRe: Kimi K3 weight release. Looks like this is a "weights available" release. This is not an "open weights" release. The license restricts commercial usage. Unlikely we will see many (any?) other provider hosting that model. Totally missed that. huggingface.co/moonshotai/K...
Upgraded my Kimi subscription after reading this. Gib AI communism pls.
Rude Law Dog@esghound.com · 2mo agoDean Ball, OpenAI's policy/lobbying lead effectively admits that Chinese AI models are as good as Closed Frontier models and cedes that without bullshit regulation "AI Communism" will win. It's delicious
![Dean W. Ball @deanwball
Some observations on Kimi:
1. It's a very good model! I don't think its performance can be explained away by distillation or anything like that. In agentic coding sessions, it seems pretty much on par with the best public models of Q1 2026. In my fairly limited use, it also seemed very token hungry. It's not obvious to me that this model is actually that cheap to run.
2. I am personally surprised the Chinese state continues to allow the open sourcing of models this good, given potential risks. To be clear, I *myself* might be fine with models presenting this level of marginal risk being open weight, but I am surprised that China is fine with it. I suspect the reason they are is 75% explained by strategic blindness/lack of AGI-pilledness (the CCP is very Yann Lecun-y in its views of AI). The other 25% or so is their lack of compute for customer inference (making China's open-weight strategy an unintended byproduct of US export controls) and the normal Chinese strategy of aggressive exports. For the companies, as opposed to the government, the decision to open source is partially ideological and partially because they are behind, and they know that very few people would pay for sub-frontier models from China.
3. Open-weight models are inherently decelerationist, and I'm continually surprised to see the so-called "accelerationists" so excited about open-weight models. I suspect the reason they are is that they know open-weight models are effectively ungovernable, and they simply like the overall cloak of ungovernability open-weight models create over the whole of AI. It's not a bad strategy; it reminds me of James Scott's recounting of the hill people in "the art of not being governed." Still, in the end, open-weight models deter further AI capex.
[This alt text added by gemini-3.1-flash-lite-preview AI]](https://cdn.bsky.app/img/feed_thumbnail/plain/did:plc:y24gcrfugd3wnhzvl7or5ft5/bafkreifbhl6gcc6kowonnyzs6b3hl5m4rgxt7lnesqa2d4h7xakmkd6vm4)
Tyler Cowen envisions a wave of Americans fanning out to third world countries to teach the ignorant natives how to use ChatGPT...
Belatedly watched Oppenheimer. I liked it, but it's a pity how it skimmed over how the actual work of physics was done, or what made the historical Oppenheimer special (his charisma, organizational chops, and mentoring ability).
This is my mental image of oil tanker captains in the Straits of Hormuz now.
As I see it, the great bastion of hope for left techno-utopianism currently is Hangzhou Deepseek Artificial Intelligence Basic Technology Research Company, Ltd.
Albert Pinto@70sbachchan.bsky.social · 7mo agois there a technology that "the left" is excited about? www.transformernews.ai/p/the-left-i...
Original scifi and fantasy cover art on display at the Portals exhibition, Singapore
Coda: I've wondered if the fact that the fascists currently running the global hegemon are obsessed with a warrior ethos but also make literally all the mistakes that Sun Tzu identified, basically they're just the kind of failsons he was warning against, might make people appreciate Art of War more.
An experimental project I've been messing with: get LLMs to prepare and run a tabletop style RPG adventure scenario. It's at a minimally playable level, and on GitHub: My GM is AI (github.com/seewhydee/my...).
Another important thing: Chinese models are not strong because they distill US models. Distillation of models via API is *impossible*. If somebody tells you the contrary, they don't understand machine learning:
Deepseek's censorship still remains by far the most incompetently done thing about it. Here it is censoring Mao Zedong thought. Can't have Maoism polluting impressionable minds!
Tyler Cowen's gonna write a passive aggressive (ahem, Strausssian) post about how this is what democratic oversight is actually like, and people are too mood-affiliated to understand it.
Science Magazine@science.org · 4mo agoThe White House today announced it plans to make major changes in how the U.S. government manages research funding. https://scim.ag/43HgW4s
What will be the next shoe to drop? A couple of stray thoughts: * US funded academics will be forbidden from doing peer review on Chinese (or all foreign) papers? * The AAAS (an AMERICAN Association after all) forbidden from publishing papers by Chinese (or all foreign) researchers?
Jeffrey Mervis@policyhound.bsky.social · 4mo agoWhat if essentially all research collaborations with China were banned? There's a new bill before Congress that would do exactly that www.science.org/content/arti...
Look at it. Hormuz has that MAIN CHARACTER ENERGY "I realized I've only seen maps of the Strait of Hormuz and was curious what it actually looks like. It's majestic! It could be a tourist destination"
If you want to understand how the solar boom is playing out across the Global South, a good place to start is Agbon, Benin, where I lived for 2 yrs. It's exceedingly unexceptional: the reason solar is taking off there is bc it's taking off everywhere🧵 timhirschelburns.substack.com/p/solar-is-e...
Solar is everywhere now. Even here.
How the solar boom is playing out in one West African village
timhirschelburns.substack.com
