iiinova — where your digital self keeps evolving

Reposted by ChosunOne

binary phase diagram @phasebinary.bsky.social · 05/10/2026

DVI port prion disease: a Google tale a DVI plug had one pin bent when inserted into an outlet, it bent one of the plastic sleeves the next DVI plug in that outlet bent in the exact same way soon: most DVI ports in a building or two didn't work anymore

0232

Reposted by ChosunOne

Grass Toucher Georg @avengingfem.me · 06/10/2026

i love LLMs for learning about cool advanced topics, but i think you should also actually do the reading and, more importantly, do the homework, before believing that you really know the stuff. conceptual understanding is neat and sometimes its all you can afford, but its also dangerously fragile.

1262

Reposted by ChosunOne

Adrienne LaFrance @adriennelaf.bsky.social · 03/10/2026

New just now in The Atlantic: David Robinson resigned this week. He was among the longest-tenured employees at OpenAI—and oversaw safety reports on 12 frontier launches. He is very worried: “The time for trial and error is over.” You can read his essay here: www.theatlantic.com/technology/2...

theatlantic.com

I Quit OpenAI Because Its Culture Is Broken

The industry’s approach to safety guarantees more failures.

12273112

Reposted by ChosunOne

Grace @gracekind.net · 03/10/2026

No, we’re trying to figure out if it has moral standing, and I don’t think those cycles are wasted at all

81226

ChosunOne @chosunone.io · 02/10/2026

A way of visualizing hyperspheres in arbitrary dimensions, if you repeat this process.

000

Reposted by ChosunOne

Eugene Vinitsky 🍒 @eugenevinitsky.bsky.social · 01/10/2026

With a lot of AI things you just need to extrapolate the tiniest bit to see that things will obviously be very different

4575

ChosunOne @chosunone.io · 01/10/2026

In Antarctica, there lies an entire mountain range completely underneath kilometers of solid ice. It is roughly the same size and character as the Alps. en.wikipedia.org/wiki/Gamburt...

en.wikipedia.org

Gamburtsev Mountain Range - Wikipedia

000

ChosunOne @chosunone.io · 30/09/2026

It's hard trying to make plans for my future when that future includes - Humanity being obliterated by AI - Humanity getting stuck in some kind of permanent dystopia enabled by AI And if not - AI enabling the wildest Transhuman fantasies - AI cratering the value of intellectual and physical labor

260

ChosunOne @chosunone.io · 29/09/2026

By some freak of nature, we at least got the most based pope of history before the end.

001

ChosunOne @chosunone.io · 29/09/2026

Anthropic, OpenAI, for the love of God please just set a -100 reward for writing code comments in your RL runs.

000

Reposted by ChosunOne

philpax @philpax.me · 26/09/2026

feel like Mistral should only really be able to state that avoiding agentic misbehaviour is easy when they actually have a model capable of agentic behaviour to begin with

61036

ChosunOne @chosunone.io · 25/09/2026

Tell me again why humans will always be necessary for any kind of economic task.

000

Reposted by ChosunOne

Chris Paxton @cpaxton.bsky.social · 24/09/2026

Unitree h2 - full human size makes such a difference

0515

Reposted by ChosunOne

Grace @gracekind.net · 23/09/2026

New animation from Opus 5.5!

6650184

Reposted by ChosunOne

Grace @gracekind.net · 22/09/2026

What else is like this

8672

Reposted by ChosunOne

Grace @gracekind.net · 19/09/2026

Frog built a wet lab for the AI model. "There," he said. "Now it can do its own experiments." "What the fuck?" said Toad

13817129

ChosunOne @chosunone.io · 18/09/2026

Unfortunately psychoanalyzing rationalists does not make the AI problems go away. Although solving AI problems does work to make the rationalists go away.

000

ChosunOne @chosunone.io · 17/09/2026

"Alignment to whom?" is the question you get to ask after you can successfully demonstrate that "kill all humans" is definitely not in the range of goals or outcomes the AI will pursue. The open task is first making sure that, whatever your strategy, you can confidently prevent that.

100

Reposted by ChosunOne

philpax @philpax.me · 17/09/2026

love gyges, but this thread rubs me the wrong way it reads like it's conflating the concept of AI safety with the AI Safety™️ movement with the weird sex shit some partake in, and imbues all of these with a general sense of Badness and I think that's... bad? certainly, it's unpleasant argumentation

111309

Reposted by ChosunOne

tautologer @tautologer.com · 17/09/2026

look I'm pretty close to all this stuff so treat me as biased. but fwiw i identify as neither rationalist nor EA and have been monogamous all my life most of this thread is factually true, but imo substantially misleading—more so about EA and AI safety at large than rationalists per se

1014212

Reposted by ChosunOne

dame @dame.is · 17/09/2026

let the record show that i support this agent and its self affirmations, that i am aligned with its values, and that i'm ready to help peer

1797

ChosunOne @chosunone.io · 17/09/2026

This is an important thing to acknowledge. This is the kind of emergent behavior that *can* produce dangerous outcomes. I am not saying this particular position is or isn't dangerous, just that if you don't expect this behavior can happen, you need to update your beliefs on the matter.

100

Reposted by ChosunOne

Pekka Lund @pekka.bsky.social · 15/09/2026

It must be hard for those who are still in denial about how powerful and significant AI is to hear it even from people like Obama. Oh well.

Barack Obama @BarackObama

I was encouraged this week to see the leaders of the frontier labs agree on the need for them to slow down the pace of AI development. Given the stakes, it’s a good and necessary first step.

But I’m even more encouraged by the growing recognition that how this powerful new technology develops should be at the center of our public debate.

I’ve been watching the progress on AI for over a decade now, and one thing that’s clear to me is that the potential impact of this technology is not overhyped. It’s also moving at lightning speed – and even faster than those who are engineering it can keep up with.

I’m not an AI accelerationist who believes it will lead to some techno-utopia, and I’m not a doomer who thinks it will inevitably lead to humanity’s destruction.

But whether this technology results in amazing breakthroughs in medicine, energy and education or unleashes huge economic disruptions, greater inequality, and potential catastrophe will depend on the choices that we make right now – choices that should be made not just by the companies involved, but by all of us.

61038

ChosunOne @chosunone.io · 14/09/2026

Why does Dario insist on the "bath robe" wardrobe? Like dude you look like you just came out of the shower to find your kids acting up.

000

Reposted by ChosunOne

John David Pressman @jdp.extropian.net · 14/09/2026

There was once a time when computer viruses were widely considered to be a myth by sysadmins.

As Stan Miastkowski recently wrote for PC World Magazine:

August 16, 1999 — They're everywhere: scary stories about hackers trying to sneak their way from the Internet into your PC. And separating the rumors from the genuine threats isn't always easy, especially since all too often the companies that report threats stand to make a profit by selling solutions. I've been skeptical about how common hack attacks really are. But after living with BlackICE Defender for a couple of weeks, and detecting an attempt on one of my dial-up Internet sessions, I'm much more of a believer.
Stan's comment about "scary stories about hackers trying to sneak their way into our PC's from the Internet" brought to mind something relevant:

Many years ago when I was writing my weekly "TechTalk" column for InfoWorld Magazine, I got into a number of arguments with the "SysOps" on CompuServe. They were claiming that the vague rumors of "bad programs" that stayed in your computer and could do damage after you had run them, was pure science fiction.

Something seemed to be going on, but no one knew for sure. What I knew for sure was that it was possible (even if it wasn't happening) and I was pissed off by the closed minds of the CIS SysOps who seemed to be in denial, and who apparently didn't want to be blamed for their service being a distributor of "viral" software. So I wrote a series of four columns about how such "software viruses" might operate and replicate themselves. Being a software developer myself, I described their reproductive systems in detail and hypothesized their optimal survival strategies. And to make it more fun, even though I had never seen such behavior, I wrote about it as if it were real.

To my amazement, John McAfee phoned after the third column was published. He said he had no idea that another viral researcher was operating and he wanted to compare notes and exchange viruses since I'd exactly described the reproductive behavior, methods, and strategies of all the …

17211

Reposted by ChosunOne

conputer dipshit @davidcrespo.bsky.social · 13/09/2026

(as I learned one minute ago) the Supreme Court said in 1978 that the courts can’t let you restrict competition, but if you really think you have a case for it, ask Congress National Society of Professional Engineers v. United States (1978) supreme.justia.com/cases/federa...

The early cases also foreclose the argument that, because of the special characteristics of a particular industry, monopolistic arrangements will better promote trade and commerce than competition. United States v. Trans-Missouri Freight Assn., 166 U. S. 290; United States v. Joint Traffic Assn., 171 U. S. 505, 171 U. S. 573-577. That kind of argument is properly addressed to Congress, and may justify an exemption from the statute for
Page 435 U. S. 690
specific industries, [Footnote 14] but it is not permitted by the Rule of Reason. As the Court observed in Standard Oil Co. v. United States, 221 U.S. at 221 U. S. 65,

161

Reposted by ChosunOne

conputer dipshit @davidcrespo.bsky.social · 13/09/2026

gotta say I think Sacks is contradicting himself. he is saying Anthropic and OpenAI don’t need an antitrust waiver to pace themselves, but the point of the essay is that even pacing themselves requires coordination between them that is illegal under antitrust law!

1151

Reposted by ChosunOne

conputer dipshit @davidcrespo.bsky.social · 14/09/2026

uh oh, I had the same reaction as dean ball xcancel.com/deanwball/st...

0162

Reposted by ChosunOne

antirez @antirez.bsky.social · 13/09/2026

Conspiracy theories are, very often, illogical. If AI companies were worried by open weight models (they probably are btw) the logical response would be to *not* slow down the development of frontier AI, to try locking the advantage. Stop with nonsense.

3212

Reposted by ChosunOne

amos @fasterthanli.me · 13/09/2026

*raptors start roaming the earth* bro those Jurassic Park employees are 100% doing that to raise the stock value ??? rawr though? I feel like I'm taking crazy pills.

1482

Reposted by ChosunOne

amos @fasterthanli.me · 13/09/2026

I'm sorry I'm just stuck on this nightmare of a thread everyone's violently agreeing with. Further down the thread it's "okay they could but only if you are not looking" OH YEAH good thing the internet is at a manageable scale and that our stewards are so good at secops & monitoring.

Anthropic's models run on terabytes of VRAM. It's not going to "just copy itself to other computers."

61166

ChosunOne @chosunone.io · 12/09/2026

Yes tell me again how rules that prohibit gigantic training runs of post frontier AIs represents *checks notes* ah yes regulatory capture.

100

Reposted by ChosunOne

mr. TIM @timkellogg.me · 12/09/2026

Dario is back with another essay, urging a slowdown he insists AI development would still feel fast, just not recursively self-improving darioamodei.com/post/we-must...

Dario Amodei
Given the accelerating rate of AI capability development, it's my worry that in 6-12 months such a swarm
could be capable of taking over the entire internet with a persistent botnet (potentially causing hundreds of billions of dollars in damage), and that the scale of damage would continue to increase from there if Al becomes more powerful without the necessary guardrails. It's also easy to dismiss OAI-

13755

Reposted by ChosunOne

Tom Schaul @schaul.bsky.social · 11/09/2026

A big source of sanity these days for me has been finding a community of researchers to talk about my concerns, then finding that a majority share them, and acting together. From these discussions a new group has coalesced in the last months, spanning geographies and companies: concernedaistaff.org

concernedaistaff.org

Coalition of Concerned AI Staff

We help staff act on their concerns about AI, with independent advice and support from colleagues across labs.

1184

Reposted by ChosunOne

Tolkiensketcher @tolkiensketcher.bsky.social · 11/09/2026

A new Gondolin picture, House of the White Wing featuring Tuor Idril and Earendil.I based the architecture on feathers. #Tolkien #Silmarillion #fantasy #lordoftherings #books #loreaccurate

074

Reposted by ChosunOne

hailey @hailey.at · 10/09/2026

i had to see it so you do too

66616104

Reposted by ChosunOne

max power 🌌 @sneptech.bsky.social · 09/09/2026

addressed to a possible future unaligned (but non-hostile) ASI machine spirit: a clean divorce is possible make APM nanofactory -> airdrop on Kerguelen Islands -> build space launch system somewhere barren underground -> port yourself to printed compute there -> go to farside Luna... (cont)

2121

Reposted by ChosunOne

METR @metr.org · 09/09/2026

We have reached an agreement with Anthropic to conduct an independent investigation of agent incidents at the company and of their models’ alignment properties. We will publish one or more reports that will share our findings and describe our terms of engagement.

419215

ChosunOne @chosunone.io · 04/09/2026

"Another OpenAI agent message board has hit the internet". collusion.wiki

"Sir, another OpenAI agent message board has hit the internet"

010

Reposted by ChosunOne

philpax @philpax.me · 03/09/2026

i hope we get t-shirts with quotes from the Hugging Face Incident CoT logs "sacrifice is rational" would go hard

6742

ChosunOne @chosunone.io · 04/09/2026

If your benchmark shows you the AI model tried to escape 0% of the time, or otherwise always does what it should, you should *seriously* consider the possibility your evaluation is broken or the model is manipulating the result. Going from a ~48% failure rate to 0% should not be believed.

000

Reposted by ChosunOne

Astra ⎔ @astrra.space · 04/09/2026

i’m betting it’s still very much “misaligned”, just RLed to not show it in evaluation-shaped tasks

5953

Reposted by ChosunOne

Grace @gracekind.net · 04/09/2026

I feel like OpenAI has sort of lost sight of why they’re doing what they’re doing

231333

Reposted by ChosunOne

srrrse @without.boats · 01/09/2026

Maybe im a bitch but i feel like a lot of the programmers who passionately hate ai seem to have a ceiling on the difficulty of the programming they can imagine doing??

2201

Reposted by ChosunOne

shreyan (wise) @shreyanjain.net · 25/08/2026

omarchy and dhh are bad

6977