The lyf so short, the craft so longe to lerne.
02139
Joined January 2007
- Tweets5.8K
- Following622
- Followers1.1K
- Likes1.8K
Jonathan Ragan-Kelley retweeted
Wrote up a practical introduction to SIMD using a real example from Ghostty. SIMD has a reputation for being complex, but the common case follows a simple, repeatable shape and shouldn’t be scary! Everyone should know at least this much. mitchellh.com/writing/everyo…
Jonathan Ragan-Kelley retweeted
It genuinely feels to me like GPT-5.2 and Opus 4.5 in November represent an inflection point - one of those moments where the models get incrementally better in a way that tips across an invisible capability line where suddenly a whole bunch of much harder coding problems open up
@simonw This quote from the recent "Bitter Lesson of LLM extensions" post (sawyerhood.com/blog/llm-exte…) resonated in a way that felt like it belonged in your canon:
"An agent isn't just a[n] LLM in a while loop. It's an LLM in a while loop that has a computer strapped to it."
@simonw I'm curious about your thoughts on policies and issues around providers training on model queries.
It would be a privacy and PR nightmare any time people figured out how to exfiltrate private information memorized from training on chat logs or browser usage.
Jonathan Ragan-Kelley retweeted
Just figured out where these fake tariff rates come from. They didn't actually calculate tariff rates + non-tariff barriers, as they say they did. Instead, for every country, they just took our trade deficit with that country and divided it by the country's exports to us.
So we have a $17.9 billion trade deficit with Indonesia. Its exports to us are $28 billion. $17.9/$28 = 64%, which Trump claims is the tariff rate Indonesia charges us. What extraordinary nonsense this is.
Jonathan Ragan-Kelley retweeted
I just want to say that I hate you all for not buying the iPhone mini resulting in its cancellation and I’ll never forgive you for it.
Jonathan Ragan-Kelley retweeted
One time 20 years ago a computer science professor said something wrong (a technicality) in class, so I spent quite a lot of time proving them wrong in this video, and then just kept going, as if I don't have any self-control over my approach to projects:
youtu.be/Ae9EKCyI1xU
Jonathan Ragan-Kelley retweeted
Not being able to see beneath the presentation of the thing vs the underlying reality of the thing seems to be a recurring theme in tech journalism at the Times. LLMs have intent and agency precisely as much as crypto has intrinsic value. 🙄
nytimes.com/2023/02/16/techn…
Jonathan Ragan-Kelley retweeted
"I’m not worried about an onslaught of terrible papers—we’ve *already* got an onslaught of terrible papers. You learn very quickly to ignore them, just like you learn to ignore junk mail, stupid Netflix shows, and spam calls." experimentalhistory.substack…
Jonathan Ragan-Kelley retweeted
If you can get fifty brilliant PhDs to work on a boring or trivial problem (air conditioner optimization or playing some video game) for a few years, they'll probably solve it. You do that by telling them they are working on "artificial intelligence" which is The Next Big Thing.
Jonathan Ragan-Kelley retweeted
Continuing the *tutor tradition, @sirrice and Robert Ward release SQLTutor, which helps you visualize query plans and their execution cudbg.github.io/sqltutor/
ALT A SQLTutor visualization of a hash join
Jonathan Ragan-Kelley retweeted
💉 In six months, we could probably have a vaccine that prevents infection and transmission for all present and future covid variants. It works in theory and it works in mice. With bureaucratic obstacles, best case is 2024. @patrickc urges a policy fix: slowboring.com/p/we-could-ha…
