David Mimno
@dmimno.bsky.social
6.4K followers 4.1K following 520 posts
He teaches information science at Cornell. http://mimno.infosci.cornell.edu
Posts Media Videos Starter Packs
Pinned
dmimno.bsky.social
Optimist: AI has achieved human level performance!

Realist: “AI” is a collection of brittle hacks that, under very specific circumstances, mimic the surface level of human intelligence

Pessimist: AI HAS achieved human level performance
dmimno.bsky.social
Defining training document quality seems to still be an open problem #COLM
dmimno.bsky.social
COLM word cloud. Yoav says it’s the year of reasoning, but evaluation is also huge.
Evaluation reasoning interpretability rl in context benchmark alignment synthetic data
dmimno.bsky.social
Interesting q: in preliterate times would the desire and ability to memorize 27k lines of Homer have been a kind of neurodiversity?
Reposted by David Mimno
laufer.bsky.social
Three major drifts we found:
1️⃣ Licenses: from corporate to other types. We often see use restrictions mutate to permissive or copyleft (even when counter to upstream license terms)
2️⃣ Languages: from multilingual → English-only
3️⃣ Docs: from long & detailed → short & templated
dmimno.bsky.social
I hear such good things about Rust, and then I actually look at it and its stuff like this:
mm-jj-nn.bsky.social
One example: I got a Cow<'_, str> from a library function. What's that? Well it contains either a borrowed str or an owned String. How can it contain a String when the type parameter says str? Well you see, str implements the ToOwned trait, and <str as ToOwned> is String. (I think I got this right?)
dmimno.bsky.social
“Which model will rid me of this turbulent priest?”
timkellogg.me
ChatGPT Pulse

say some vague wish during chat, and Pulse will try to make it happen while you sleep

like a personal assistant
Sam Altman
@sama
X.com
Today we are launching my favorite feature of ChatGPT so far, called Pulse. It is initially available to Pro subscribers.
Pulse works for you overnight, and keeps thinking about your interests, your connected data, your recent chats, and more. Every morning, you get a custom-generated set of stuff you might be interested in.
It performs super well if you tell ChatGPT more about what's important to you. In regular chat, you could mention "I'd like to go visit Bora Bora someday" or "My kid is 6 months old and I'm interested in developmental milestones" and in the future you might get useful updates.
Think of treating ChatGPT like a super-competent personal assistant: sometimes you ask for things you need in the moment, but if you share general preferences, it will do a good job for you proactively. This also points to what I believe is the future of ChatGPT: a shift from being all reactive to being significantly proactive, and extremely personalized.
This is an early look, and right now only available to Pro subscribers. We will work hard to improve the quality over time and to find a way to bring it to Plus subscribers too.
Huge congrats to @ChristinaHartW, @_samirism, and the team for building this.
Reposted by David Mimno
dbamman.bsky.social
The UC Berkeley School of Information is hiring an assistant professor in the broad field of Information--including areas of info seeking/retrieval, digital humanities, cultural analytics, info viz, & philosophy of information (among others). Deadline Nov 1! aprecruit.berkeley.edu/JPF05014
Assistant Professor - Information - School of Information
University of California, Berkeley is hiring. Apply now!
aprecruit.berkeley.edu
dmimno.bsky.social
I was chatting with people in the audience after a job talk in Scotland and someone quietly informed me that I needed to leave because they were all about to discuss my performance.
Reposted by David Mimno
dorialexander.bsky.social
Frankly going to implore progressives to follow better what is happening in China tech. It's not a VC bubble. It's a dozen top labs, it's most large co getting heavily into AI (at a pretraining level) and robotics, it's a deep bet on the next industrial revolution.
dmimno.bsky.social
and for messing up my regular expressions, apparently!
Reposted by David Mimno
greenleejw.bsky.social
I hate to talk about Christmas at a time like this (it being only Sept.), but if you're thinking about commissioning a gift map for Christmas...'tis the season!

I've done a number of gift maps over the last few years, and they take a while. So now's the time to get in line!
A map of the Society for Creative Anachronism's Loch Soillier barony in Texas, done in a pseudo-medieval style. There are sea monsters! Map of part of New York City. Surrounding the city are black and white drawings of various city landmarks. The drawings are in half circles that look like snow globes, and each has a number that cousins corresponds to a number on the map. A map of a piece of property in Florida, with two ponds, a shed, a trailer, a trehouse, and a lot of trees. The map is labeled, "The Hideout" and there is an oppossum. A map of the area around Seattle, from Tacoma in the South to Whidby Island in the north. It is done in a fantasy style, with different places represented by different icons. The border took me a long time, and is made of abstract shapes on a red background. There are mountains and forests, and the map includes Mt. Rainier in the bottom right and Mt. Backer in the top right. Neither mountain is actually there, but it's close enough.
dmimno.bsky.social
Does anyone know the purpose or typical use of the "next line" character, Unicode 0x85? How does it relate to CR and LF?
dmimno.bsky.social
Could you send me details?
dmimno.bsky.social
It’s not the Reichstag, it’s Horst Wessel
Reposted by David Mimno
rahaeli.bsky.social
All of these tips are derived from the psychological protections developed for content moderators, and if you want more detail and citations, crowd.cs.vt.edu/wp-content/u... is the paper that has the best collection of citations that I've found .
crowd.cs.vt.edu
Reposted by David Mimno
rahaeli.bsky.social
There have been *some* studies on other games, but Tetris is the best-studied because it was the one in the original study. It does seem that the key elements are "fast/timed", "lots of concentration", and "lots of eye movement across the game board" if you can't do Tetris for whatever reason.
Reposted by David Mimno
mellymeldubs.bsky.social
Job alert! 🌸🏔️ The UW iSchool is hiring for *two* TT faculty positions, with a focus on Artificial Intelligence, broadly defined. It's pretty here!

Applications due November 15th 2025 (priority deadline).

Link: apply.interfolio.com/171020
Aerial photo of UW campus with Mt. Rainier in the background
dmimno.bsky.social
Remind me how bullish you are on AGI?
dmimno.bsky.social
Exactly! I worked with an example this week where we could take ok OCR of a 16c Latin colloquium, fix it up, format speakers, and add long vowel marks. Needs to be checked, but it might be available on the web soon. Without LLMs, a two-year project of boring work, unlikely to be funded.
dmimno.bsky.social
In the same sentence as they report the revisions for May and June! How can this be so hard to contextualize?

Not picking on Yahoo, they all seem bad
dmimno.bsky.social
BLS revisions often correlated, so July prob revises down, right? But Yahoo finance reporting it as Fact: “In July, the economy created 73,000 new jobs, but the headline that emerged from that report was revisions to job gains in May and June, which wiped out some 258,000 previously reported gains.”
dmimno.bsky.social
Most of the history of astronomy didn’t use telescopes, much less digital images. If a humanist has a question that doesn’t need data management, great. If they do, also great. Leading with the tech is the wrong approach, I think.