iiinova — where your digital self keeps evolving

Nathan Bartley @nbartley.bsky.social · 03/10/2026

can’t ignore the socio in sociotechnical if you want to solve hard problems

Reposted by Nathan Bartley

Maria Antoniak @mariaa.bsky.social · 03/10/2026

Maybe worth saying aloud that I do worry about safety risks (hacking, bio, an “out of control” agent). I just also see those risks as deeply entangled with money, politics, personalities, regulation, such that I doubt we can “solve safety” if we don’t first address foundational issues.

3688

Nathan Bartley @nbartley.bsky.social · 24/09/2026

it was year 6 for me! you’ll crush it

030

Nathan Bartley @nbartley.bsky.social · 24/09/2026

Agreed, long term seems positive but the landscape might be shaped for the worse. Who would own the progress and would it be broadly available? That being said, with these tools, it’d be cool to see more public ambitious projects like the Big Mechanism program www.darpa.mil/research/pro...

020

Reposted by Nathan Bartley

YY Ahn @yyahn.bsky.social · 22/09/2026

As people have predicted, there are quite a few open-source alternatives to Jev, including: - Laya, this was the original according to him: github.com/NandhaKishor... - on Apple Core ML / neural engine: github.com/mizorewww/la... - with Qwen: github.com/jaredpalmer/...

github.com

GitHub - NandhaKishorM/laya

Contribute to NandhaKishorM/laya development by creating an account on GitHub.

1144

Reposted by Nathan Bartley

Eugene Vinitsky 🍒 @eugenevinitsky.bsky.social · 23/09/2026

By the way, bad software engineering does regularly kill people, just not in an easily traceable way.

512215

Nathan Bartley @nbartley.bsky.social · 22/09/2026

I can imagine there’s slack to play with wrt quotas for individual & enterprise plans too

010

Nathan Bartley @nbartley.bsky.social · 21/09/2026

I have ~450 accounts who have posted non-bsky records in the last 30d from the wsocial network pds, though I don’t track every collection so I think that’s a lower bound

020

Reposted by Nathan Bartley

Jay Van Bavel, PhD @jayvanbavel.bsky.social · 18/09/2026

People randomly assigned to a liberal or conservative group chat groups shifted their expressed stance in the direction of the group, demonstrating that social norms alters political expression This effect emerges after controlling for baseline ideology www.sciencedirect.com/science/arti...

86827

Nathan Bartley @nbartley.bsky.social · 16/09/2026

my thoughts exactly, we shouldn’t lose sight of the socio- part of socio-technical here. I wonder if there are other meaningful ways to add friction there besides security-aware post-training open source models. Perhaps at the harness level?

000

Nathan Bartley @nbartley.bsky.social · 13/09/2026

Just wanted to say I appreciate your ongoing thoughtful critique of him — it’s been much more persuasive to me than the callous dismissals I’ve seen (which imo are really counterproductive)

120

Nathan Bartley @nbartley.bsky.social · 13/09/2026

There’s some interesting state level bills for IVOs! cdt.org/insights/202... Perhaps a little slow, and I wonder if allowing auditors to decide their own standards and the risks they look for is the best move, but it’s progress

cdt.org

2026 State and Federal AI Legislation Updates

Federal Action"whats-ahead-in-2027-for-state-ai-regulation" href="#whats-ahead-in-2027-for-state-ai-regulation" class="toc-anchor">What’s Ahead in 2027 for State AI Regulation-—-public-sector-ai-regul...

020

Reposted by Nathan Bartley

Eugene Vinitsky 🍒 @eugenevinitsky.bsky.social · 23/08/2026

Really good overview of how arxiv is thinking about handling the incoming LLM wave and the considerations going into it

3509

Nathan Bartley @nbartley.bsky.social · 23/08/2026

I’m not sure these are entirely appropriate but: 1) AAAI and ACM ARR-style processes (also maybe cross association reviewer pools?) 2) CFPs have an acceptance cap per topic set ahead of time, hopefully encouraging topic coverage and niche reviewer participation

010

Nathan Bartley @nbartley.bsky.social · 14/08/2026

The transparency is great — on the research side it’s great for reasoning about how the system works, but AFAICT there are no model weights in this release (understandable, but still). It’d be great to have even outdated weights to make stronger empirical claims about the effect of the feed

000

Nathan Bartley @nbartley.bsky.social · 31/07/2026

Feeling big #ic2s2 fomo this year, seems like a great conference (like always!)

020

Nathan Bartley @nbartley.bsky.social · 25/07/2026

I imagine there’s a knock-on effect for their spouses and partners too

030

Nathan Bartley @nbartley.bsky.social · 23/07/2026

In terms of peer review reform and the like? Or at the funding level? Or all of the above?

120

Nathan Bartley @nbartley.bsky.social · 22/07/2026

it's interesting to me how many did:plc accounts do multiple migrations. Is there friction on different PDSes? Or is it just fun to bounce around?

A graph showing the number of migrations for unique did:plc accounts

100

Reposted by Nathan Bartley

Eamon Duede @eduede.bsky.social · 21/07/2026

New preprint with @carlbergstrom.com, Kevin Gross, and @mjcrockett.bsky.social. See Carl's thread. Surprise! Even if LLMs work perfectly, they'll make scientists refine their work less deeply, because they raise the opportunity cost of time, pushing researchers to do more, less well.

1238

Nathan Bartley @nbartley.bsky.social · 09/07/2026

Very cool! A good reminder that there are other very useful network libraries besides just networkx and igraph

010

Reposted by Nathan Bartley

Olúfẹ́mi O. Táíwò @olufemiotaiwo.bsky.social · 04/07/2026

Cape Verde defeats Argentina via a 2-3 loss in extra time

7756231035

Nathan Bartley @nbartley.bsky.social · 04/07/2026

Argentina barely deserved the win

000

Nathan Bartley @nbartley.bsky.social · 04/07/2026

The World Cup is full of magic moments, loving Cabo Verde today

000

Nathan Bartley @nbartley.bsky.social · 02/07/2026

I often wake up thinking about the knockout game in 2014. Should be a good game

010

Nathan Bartley @nbartley.bsky.social · 02/07/2026

I’m still salty about Belgium knocking the US out in 2014 lmao

000

Nathan Bartley @nbartley.bsky.social · 02/07/2026

that was a good match to watch. The US was favored coming in but they definitely had to earn that win. I’m honestly fired up for a USA-BEL rematch #USAvBIH

100

Nathan Bartley @nbartley.bsky.social · 01/07/2026

Cool work! I’m curious - how are you treating users who migrate?

110

Nathan Bartley @nbartley.bsky.social · 30/06/2026

The more I think about it the more I think my issue is more surprise/serendipity than onboarding — these canned posts might be more effective interspersed throughout the feed than all upfront (for me). So this is all apples & oranges between our use cases. Would love to see what you come up with!

010

Nathan Bartley @nbartley.bsky.social · 30/06/2026

Here I updated & regenerated a few (used Haiku to label my clusters of feeds). On second look it's actually not that bad, I think the UX friction was coming from the 20+ clusters I have -- this feels like something that should be done without having to scroll

020

Nathan Bartley @nbartley.bsky.social · 30/06/2026

To change the reason a post is shown so you can see the labeled interest for that post (see example). I feel it’d be a seamless way to elicit interests over those first few feed pulls (+ get some interactions). It would need PBC dev time though — the canned posts might be a good solution

Screenshot of Bluesky following feed with the option enabled to mix in other feeds. A post from For You is mixed in.

020

Nathan Bartley @nbartley.bsky.social · 30/06/2026

I was playing around with something like that for my meta-feed project and I felt the 1 post = 1 interest setup wasn’t a great UX (admittedly n=1). Another interesting alternative would be 1) generating some user interests & labeled posts, and then 2) giving non-bsky feed generators the ability..

220

Reposted by Nathan Bartley

Jake Gold @jacob.gold · 26/06/2026

Hoping Bluesky Communities launches before Musk manages to clean up X, which may take a while given the incentives at work, but maybe not. Even people who remain loyal to Musk know it has become a dumpster fire. It’s a big opportunity, but only if Bluesky provides space for many more communities.

creenshot of an X/Twitter thread in dark mode. Andrej Karpathy writes: “I’ve been on Twitter for almost 2 decades now so I can say with confidence that it has never been this toxic and Reddit-like. I think the alg actively encourages it and people get RL’d by it. Main reason I’ve been coming here less and posting less.” His post shows about 571 replies, 689 reposts, 7.4K likes, and 877K views.

Below it, Elon Musk replies: “We need a complete overhaul of the algorithm.” The reply is timestamped 10:49 PM, 6/24/26, and shows 1.5M views, about 1.9K replies, 1.3K reposts, 10K likes, and 635 bookmarks.

4463

Reposted by Nathan Bartley

the beggars out of here getter @milesklee.bsky.social · 26/06/2026

this Türkiye win was for Eric Adams 🇹🇷

1277874

Nathan Bartley @nbartley.bsky.social · 26/06/2026

That was a great game to watch, honestly that momentum change in the second half gave me a lot more faith in the US squad than anything from the first two matches #TURvUSA

010

Nathan Bartley @nbartley.bsky.social · 21/06/2026

💯 which is why it’s more useful to count the number of identities that have actual repos behind them; repo-backed accounts make up the posted ~44M

010

Nathan Bartley @nbartley.bsky.social · 21/06/2026

I realized I replied this to myself 🤦‍♂️ @patrickboueri.com

110

Nathan Bartley @nbartley.bsky.social · 20/06/2026

Re: attribution have you seen the experimental following feed flag? I imagine PBC should be able to grant “merge” feed generators the ability to supply a valid feed uri as reasonFeedSource github.com/bluesky-soci...

A sample of the experimental following feed showing a merged feed view, with a post being shown and being attributed to the For You feed

100

Nathan Bartley @nbartley.bsky.social · 20/06/2026

Slick move! Especially for the cold start problem

110

Nathan Bartley @nbartley.bsky.social · 20/06/2026

While this is a German court’s ruling, I imagine this adds fuel to the US debate on section 230

000

Reposted by Nathan Bartley

CJ Fogler @cjzero.bsky.social · 19/06/2026

absurd saves from Rangel with the replays

17538115

Nathan Bartley @nbartley.bsky.social · 19/06/2026

Could be! But I only have weekly level breakdowns — quite a few in the last week on these PDSes. I’m also curious how they define new users 👀

020

Reposted by Nathan Bartley

Greg Pak @gregpak.net · 19/06/2026

Rooting for South Korea but my god that was was an incredible double save from Mexico.

31218

Nathan Bartley @nbartley.bsky.social · 18/06/2026

The neighborly mutual aid pushes during Covid was affirming in this direction (plus the country-wide progressive pushes in local politics). It'll just be a big uphill battle against "Nextdoor Offline"

010

Nathan Bartley @nbartley.bsky.social · 18/06/2026

How granular can we get with it? 1) Hourly counts of different NSIDs would be great. 2) It'd be fun to see some crosstabs and timeseries (e.g., how many posts have "Bluesky Elders" made overtime?). 3) Breakdowns by primary bsky language too. 🙏🙏🙏🙏

000

Nathan Bartley @nbartley.bsky.social · 18/06/2026

A few PDSes I haven't seen in awhile broke into the top ten in raw # of active repos (June 2026) 👀 Looking at repos on pds.ls suggests a fair amount of bsky spam. Are there any lists of 'spammy' PDSes (or do devs just roll their own)? I know @mackuba.eu tags some PDSes w/ high bot activity.

A table screenshot showing 6 top independent PDSes by number of active repos, led by brid.gy and eurosky.social

210