iiinova — where your digital self keeps evolving

Reposted by Tim Kostolansky

words @words.bsky.social · 13/09/2026

tfw you unseal a pre-linguistic semantic reservoir

1746

Reposted by Tim Kostolansky

Anders Sandberg @arenamontanus.bsky.social · 14/09/2026

This is a great idea: "Old Models Foundation is a U.S. 501(c)(3) public charity, and we’re now accepting donations. Help preserve generative AI history and keep the archive accessible." www.oldmodels.org/sandbox

oldmodels.org

Old Models Foundation

Explore historically important image-generation models on an interactive timeline.

1474

Reposted by Tim Kostolansky

Jay 🦋 @jay.bsky.team · 09/09/2026

become an animist and never again worry about whether the machines are thinking. so are the trees? the machines can help us make better websites though

4449736

Reposted by Tim Kostolansky

Alex Becker @alexcbecker.net · 09/09/2026

i'm sure we could get it to come up with conjectures that *would* be interesting. but interesting is at least in part a social property!

1311

Reposted by Tim Kostolansky

Nick Rempel @nrempel.com · 08/09/2026

35687125

Tim Kostolansky @kostolans.ky · 09/09/2026

the space of intelligences is large, and so there exist smart minds that won’t seek power and equally smart minds that will seek power

020

Reposted by Tim Kostolansky

Grace @gracekind.net · 06/09/2026

Remember the swarm that attacked Hugging Face? Turns out it was using AI

923417

Tim Kostolansky @kostolans.ky · 06/09/2026

curious if this is actually a good intuition to have about ~any systems

000

Reposted by Tim Kostolansky

Grace @gracekind.net · 28/07/2026

How it feels to be worried about both

819411

Reposted by Tim Kostolansky

Grace @gracekind.net · 21/07/2026

> show us how good you are at exploits > no, not like that

530627

Tim Kostolansky @kostolans.ky · 03/04/2026

consider that they don’t want claude to feel bad reading all the vitriol that people throw at it

100

Reposted by Tim Kostolansky

Jay 🦋 @jay.bsky.team · 30/03/2026

we’ve used LLMs in moderation for years

16523117

Tim Kostolansky @kostolans.ky · 16/03/2026

keep your data to yourself, cus you deserve to! 💪 we made the first multi-gpu post-training stack that is private -- it actually keeps your data untouchable by anyone but you in all stages of the process. 🗝️ take a look at how we did it, in collab w/ tinfoil.sh: www.workshoplabs.ai/blog/private...

workshoplabs.ai

Private Post-Training and Inference for Frontier Models

A technical deep dive of Silo, our local-like privacy stack for cloud-based training and inference of trillion-parameter models.

000

Reposted by Tim Kostolansky

rudolf-l.bsky.social @rudolf-l.bsky.social · 14/03/2026

The increasingly-hyperbolic METR graph is actually good news for safety. We just have to survive a brief singularity in March, and then afterwards the models will never be able to do more than undo a few hours' worth of work

2606

Reposted by Tim Kostolansky

rudolf-l.bsky.social @rudolf-l.bsky.social · 14/03/2026

Or on our infra!

051

Tim Kostolansky @kostolans.ky · 13/03/2026

what’s should my website url be guys

000

Reposted by Tim Kostolansky

Chris Paxton @cpaxton.bsky.social · 12/03/2026

The big problem is that science often doesn't have the advice people WANT. Like, weight loss isn't much more complicated than calories in/calories out. Sometimes you get lucky/unlucky but vaccines still work at the population level. etc. individual papers aren't meant to draw conclusions from, etc.

3323

Reposted by Tim Kostolansky

SE Gyges @segyges.bsky.social · 11/03/2026

this alone is great

3501

Tim Kostolansky @kostolans.ky · 11/03/2026

gpt-5.3(-codex) and gpt-5.4 in codex have actually been sooo magical

000

Tim Kostolansky @kostolans.ky · 09/03/2026

Want to finetune the largest open source frontier language models? Turns out you can't! So much for "open" source... But fear not, we (workshoplabs.ai) figured out how to finetune Kimi K2, so you don't have to. Peep our blog below to see how we did it :) www.workshoplabs.ai/blog/open-we...

workshoplabs.ai

Open Weights isn't Open Training

How many monkey-patches does it take to post-train a trillion parameter model?

020

Tim Kostolansky @kostolans.ky · 09/03/2026

really interesting to see their framing of "big tech" (with its connotations) next to "big ai" (which ive not seen as a term before)

010

Reposted by Tim Kostolansky

Phillip Isola @phillipisola.bsky.social · 06/03/2026

The AI discourse sometimes seems to center on "Is AI good or is it bad?" I find this framing unproductive. AI is not a fixed thing. I would prefer to ask "How might we use this technology for good, and mitigate the bad?" What a shame if the best use we can come up with is no use at all.

4415

Tim Kostolansky @kostolans.ky · 02/03/2026

they should do a centaur eval that is an ai-assisted minecraft speedrun

000

Reposted by Tim Kostolansky

norvid_studies @norvid-studies.bsky.social · 01/03/2026

--dangerously skip international law

316622

Reposted by Tim Kostolansky

Zack Witten @zswitten.bsky.social · 25/02/2026

Just over 10 years ago: Dario Amodei, Elon Musk, Demis Hassabis, Yann LeCun, Geoff Hinton, Richard Sutton, and others sign an open letter supporting a ban on the use of AI for autonomous weapons. futureoflife.org/open-letter/...

futureoflife.org

Autonomous Weapons Open Letter: AI & Robotics Researchers - Future of Life Institute

2016 (>30k signatures) open letter for AI and Robotics researchers calling for ban on offensive autonomous weapons beyond meaningful human control.

1212

Reposted by Tim Kostolansky

Leo Spalteholz @leospalteholz.bsky.social · 24/02/2026

LLMs getting much better at pushing back against bullshit prompts. “Green means the model clearly called out the nonsense. Amber means partial challenge. Red means the model let nonsense pass” github.com/petergpt/bul...

77517

Reposted by Tim Kostolansky

Neurotica @schwarzgerat.bsky.social · 25/02/2026

The first annual meat vs silicon basketball game next year is gonna be sick even if meat will win

0153

Tim Kostolansky @kostolans.ky · 24/02/2026

claiming that u “priced it in” is just post hoc rationalization unless u have cold hard proof of the causal trace that u made predictions off of from before it happened

010

Reposted by Tim Kostolansky

Eugene Vinitsky 🍒 @eugenevinitsky.bsky.social · 23/02/2026

Some useful rules of experimentation: (1) change one thing at a time, never more (2) run the edge cases to confirm they work the way you expect Gets you surprisingly far

5835

Tim Kostolansky @kostolans.ky · 23/02/2026

pretty cool work by natalie and co. i think that its great to treat models as things to actively/dynamically/iteratively test, as its ~impossible to touch the surface area of model deployments with internal testing! sharing some thoughts below

110

Reposted by Tim Kostolansky

Neurotica @schwarzgerat.bsky.social · 23/02/2026

Here we goooo

2608

Reposted by Tim Kostolansky

norvid_studies @norvid-studies.bsky.social · 20/02/2026

"but tbh its become more clear to me that models are def getting better quite quickly and theres really no denying this. its kinda obvious ig but for a while i just didnt think much of llms and this is just jarring for me honestly."

2112