iiinova — where your digital self keeps evolving

Reposted by Sireesh Gururaja

William B. Fuckley @opinionhaver.bsky.social · 30/09/2026

There are two academic disciplines, computer science and sociology. Biology? Computer science. Physics? Computer science. Economics? Computer science. Computer Science? Actually, Sociology.

1428742

Sireesh Gururaja @siree.sh · 24/09/2026

Like, France and Germany have an OSS version of Google docs already!

020

Sireesh Gururaja @siree.sh · 24/09/2026

This is a great set of points, and I think the org/movement side could really benefit from similar work to host e.g. alt office stacks that reduce depends on e.g. Microsoft and Zoom

121

Reposted by Sireesh Gururaja

Cella @cellllla.bsky.social · 21/09/2026

Please join us at Carnegie Mellon University and online on October 2nd, 2026 for the Community-Centered AI Infrastructure Summit, a statewide convening to consider the opportunities, challenges, and governance of AI infrastructure development in Pennsylvania. RSVP here: communityaisummit.org

Flyer for the Community-Centered AI Infrastructure Summit on Friday October 2, 2026 from 9-5pm at Mellon Institute Auditorium, Carnegie Mellon University, Pittsburgh, PA. Join residents, community organizations, policymakers, researchers, technologists, energy experts, and advocates from across Pennslyvania for a public convening. Organized by CMU, Data & Society, PA Climate Equity Table, and Center for Coalfield Justice, with support from the Block Center.

22916

Sireesh Gururaja @siree.sh · 21/09/2026

That this produced a bunch more text for LMs to train on may have been a side benefit, but I doubt it was the goal. Especially because the kinds of language models that Google seemed interested in pre-ChatGPT didn't even generate text, and I'm not sure how much value SEO text would have provided

140

Sireesh Gururaja @siree.sh · 21/09/2026

This also to say that I think your original post had it kinda backwards: the new SEO guidelines seem designed for the more natural language search that Google wanted to encourage through BERT

130

Sireesh Gururaja @siree.sh · 21/09/2026

Can't speak fully authoritatively about this, but I believe this pivot to natural language is when neural approaches started to be widely deployed at Google. In particular, BERT (an early LLM) started being used in search around this time. I wouldn't be surprised if they started using neural ASR too

blog.google

Understanding searches better than ever before

How new advances in the science of language understanding will help you find more useful information in Search.

120

Sireesh Gururaja @siree.sh · 21/09/2026

Congrats, and can't wait to eventually read it!!

010

Sireesh Gururaja @siree.sh · 19/09/2026

Also really excited to start using my new investing strategy, buy low sell high!

060

Sireesh Gururaja @siree.sh · 18/09/2026

Yes!! Even just the ability to have a portable ID between conferences would be great. If you haven't already, also worth looking at @mariaa.bsky.social and @eugenevinitsky.bsky.social's bluesky client Lea, which does cool conference things also

030

Sireesh Gururaja @siree.sh · 18/09/2026

I would love a more streamlined way to collaboratively figure out who is at the conference (so far I've relied on tags, and usernames, I.e. X is at conference!), and to be able to both cluster and reach out to people based on common interests.

240

Sireesh Gururaja @siree.sh · 18/09/2026

Yes, and while I've generally disliked them (not that many people on there, generally jankily built), I really acutely felt their lack when at a conference that didn't have one (this past year's CHI) like @djpardis.com, I've also built custom feeds and found them useful!

120

Sireesh Gururaja @siree.sh · 09/09/2026

hindley milner tripe system

030

Sireesh Gururaja @siree.sh · 04/09/2026

They really need to start allowing you to allocating points for different options, so many of us are from multiple places!

040

Sireesh Gururaja @siree.sh · 04/09/2026

It remains wild to me that people have just stopped entertaining the idea that X is maybe not a place to post after just...all of what's going on there 💀

050

Sireesh Gururaja @siree.sh · 03/09/2026

This was a really fun read!

010

Reposted by Sireesh Gururaja

Joseph Fasano @josephfasano.bsky.social · 24/08/2026

Back to school. ❤️

282566549

Reposted by Sireesh Gururaja

e.w. niedermeyer @niedermeyer.online · 21/08/2026

If you want people to engage with a technology you care about, you have to find ways to frame it around providing them with some agency. You also have to be ready for people to just say "fuck you, no" and be OK with that. You have to learn to center people and not the technology itself.

1141461

Sireesh Gururaja @siree.sh · 12/08/2026

Real domain!! 💃🤩 So happy to see this in public!

120

Reposted by Sireesh Gururaja

P(aul) Frazee @pfrazee.com · 05/08/2026

My impression is that AI coding is on a "pick two of three" triangle: scope, ship speed, and correctness. Small tools are (speed + correctness) fast prototypes are (speed + scope) and big projects are (scope + correctness). Every project that tries to be all three degrades to "fast prototype."

2343939

Sireesh Gururaja @siree.sh · 04/08/2026

It has been *wild* to me to see the way that people in my department have fully gone back to business-as-usual posting on Twitter for their papers. There were a couple of brief blips where people tried bluesky and LinkedIn, but that's nearly all gone now 😔

161

Reposted by Sireesh Gururaja

New-Cleckit Dominie @ncdominie.bsky.social · 28/07/2026

This is just to say I have OCRed the text thet was in tne arch:ve and whicb yov were prabaldy seamhing for "breahtast" Fargme ms fheg mem debo:w5 sb ?mccf mmb 5c oc1b

7778512571

Sireesh Gururaja @siree.sh · 24/07/2026

It's finally out! At a time when basic research funding through the NSF and NIH in the US is under threat, it's important to understand what "basic" means in the context of more available DoD funding. For a lot of academics, the DoD's "basic" might not be be as basic as you think!

1115

Sireesh Gururaja @siree.sh · 23/07/2026

I also thought plane! The use of "plain" singular feels odd in the context of a whole country - if it were plural, I'd absolutely parse it as "plains" instead of "planes". Spain has plains, sure, but what would "the plain" of Spain be?

020

Sireesh Gururaja @siree.sh · 23/07/2026

40 minute headways in Pittsburgh are fine, apparently?

Google maps directions where the transfer instruction is "Walk, then wait for up to 1455 minutes"

020

Sireesh Gururaja @siree.sh · 22/07/2026

The queer one really perfects the format! The way those women plant and twist the knife...ooff

020

Reposted by Sireesh Gururaja

Mateusz Fafinski @calthalas.bsky.social · 18/07/2026

Remember, both the Iliad and the Odyssey were originally audiobooks.

18373501318

Sireesh Gururaja @siree.sh · 05/07/2026

Great thread from a book that has lived rent-free in my head since it argued that computers are inherently and fundamentally conservative. I didn't even know it was out of print!!

"Yes, the computer did arrive "just in time." But in time for what? In tie to save—and save very nearly intact, indeed, to entrench and stabilize—social and political structures that otherwise might have been either radically renovated or allowed to totter under the demands that were sure to be made on them. The computer, them, was used to conserve America's social and political institutions. It buttressed them and immunized them, at least temporarily, against enormous pressures for change."

0183

Sireesh Gururaja @siree.sh · 05/07/2026

Yes, let's!! I'm more on the individual+human involved evaluation angle, and I think this would be complementary

030

Sireesh Gururaja @siree.sh · 05/07/2026

Pierce was also one of the authors of the ALPAC report, which is credited similarly to the Lighthill report for contributing to the AI winter in the US.

020

Sireesh Gururaja @siree.sh · 05/07/2026

You may already have seen this, but also worth reading JR Pierce's 1969 piece "Whither Speech Recognition". The writing is similarly...pointed, and there's also some interesting foresight about the shape of what language models _did_ end up doing for speech recognition.

Quote from Pierce's "Whither Speech Recognition". It reads: "It would be too simple to say that work in speech recognition
is carried out simply because one can get money for it. That is a
necessary but not a sufficient condition. We are safe in asserting
that speech recognition is attractive to money. The attraction is
perhaps similar to the attraction of schemes for turning water
into gasoline, extracting gold from the sea, curing cancer, or going
to the moon. One doesn't attract thoughtlessly given dollars by
means of schemes for cutting the cost of soap by 10%. To sell
suckers, one uses deceit and offers glamor."

140

Sireesh Gururaja @siree.sh · 04/07/2026

And the brief (hopefully not vehicular) detour to gin and tequila along the way!

110

Sireesh Gururaja @siree.sh · 03/07/2026

Working on exactly this kind of thing!! Would love to talk about it

110

Sireesh Gururaja @siree.sh · 26/06/2026

We can always look to you for a timely feed 🫡

010

Sireesh Gururaja @siree.sh · 23/06/2026

Extremely hype to read this 👀

031

Sireesh Gururaja @siree.sh · 23/06/2026

I wish I could come to this!!! It looks so good!

010

Sireesh Gururaja @siree.sh · 18/06/2026

This whips, and I cannot wait for interactive kubernetes tutorials like I'm learning JavaScript for the first time 🤩

110

Sireesh Gururaja @siree.sh · 12/06/2026

Especially relevant considering how many schools now have the GPU resources that you'd need for it. I think one of the better reasons is probably a liabiility/PR thing — you don't have to worry content moderation or what people are using the service for

030

Sireesh Gururaja @siree.sh · 10/06/2026

Excited to watch this! Are you drawing on Weizenbaum re: computation generally, or more about the way the predictions are tied to the data?

110

Sireesh Gururaja @siree.sh · 10/06/2026

I think it'd be great to do this, but fully controllable by the sharer - temp links if desired, also able to turn it off. BUT doing it based on all the data - a true "this is someone else's for you feed" would be really neat, esp if you could pin it rather than it taking over your for you feed

010

Sireesh Gururaja @siree.sh · 08/06/2026

If you haven't already, I'd also def recommend taking a look at the European Language Industry Survey, which I believe they publish every year, and which has an archive of past results: elis-survey.org

elis-survey.org

ELIS 2023 LANGUAGE INDUSTRY SURVEY

ELIS focuses on the industry trends, expectations and concerns of a wide range of language industry stakeholders, including language service companies, independent language professionals and others.

120

Sireesh Gururaja @siree.sh · 08/06/2026

I recently helped out with a paper about this! +1 to the other Moorkens rec, I found this work (among others) really helpful to contextualize LLMs/neural stuff as another front in the effort to make translation into commoditized piecework: benjamins.com/catalog/ts.0...

benjamins.com

“A tiny cog in a large machine”: Digital Taylorism in the translation industry

Translators have worked with the assistance of computers for many years, usually translating whole texts, divided into segments but in sequential order. In order to maximise efficiency and inspired by...

120

Reposted by Sireesh Gururaja

Brian Waismeyer @yetanotherbrian.bsky.social · 08/06/2026

good ML: * carefully designed for narrow contexts (or with awareness of context limits) * designed by ML and topical experts, with expert curated data and expert reviewed outputs * designed for specific needs and consumers, and kept up to date for both * clear training / comms on use and limits

1598

Sireesh Gururaja @siree.sh · 05/06/2026

The implicit compositionality born of English's status as a bastard language is a gift that calls for audacity, chutzpah, moxie, panache, etc

031

Reposted by Sireesh Gururaja

🌶 David Gray Widder @davidthewid.bsky.social · 03/06/2026

I am beginning new research, and would like recommendations on your favorite writing on the "strange bedfellows" phenomenon. Unlikely coalitions fighting for a common goal, and the ethical quandaries therein.

172