iiinova — where your digital self keeps evolving

Reposted by @john44234.bsky.social

Astra ⎔ @astrra.space · 07/10/2026

just to remind you what this is: this single box with two SSDs. is running deepseek v4.1 flash (a 552B model) in its native MXFP4, ~550GBs of weights plus engrams total. with prefill speeds of over 450tps and decode speeds of over 15tps. at full 1M context.

716214

Reposted by @john44234.bsky.social

Menno @menno.bsky.social · 02/10/2026

Dipping my toes in releasing my stuff publicly by showing off Weft, My Agent Memory system. I am entirely biased, but I think it's kinda neat! MIT Licensed. Local First. Memory to help you get stuff done, not call you friend weft.mediumroast.dev #ai

weft.mediumroast.dev

Weft — A Shared Persistent Memory For AI Agents

Have your agent remember where you left off last session without wasting context. MIT Licensed & Local First

343

Reposted by @john44234.bsky.social

Faine Greenwood @faineg.com · 01/10/2026

exactly. Don't debate these weirdos: just decide for yourself if you want THEM in charge of the future of humanity, because "being in charge of our collective future" is what they're trying to accomplish.

211219

Reposted by @john44234.bsky.social

Vincent Carchidi @vcarchidi.bsky.social · 30/09/2026

Important IMHO. "Our results separate two properties that are easy to conflate: whether a thinking trace is useful computation for producing an answer, and whether the emitted trace is itself a correct derivation of that answer." arxiv.org/abs/2609.38107

arxiv.org

Correct Answers, Invalid Traces: What Verifiable Grade-School Math Reveals About Chain-of-Thought Traces

Chain-of-thought traces are widely read as records of how models reach their answers, informing debugging, agent auditing, and claims about reasoning. Testing this interpretation is difficult because ...

3789

Reposted by @john44234.bsky.social

rev. howard arson @theophite.bsky.social · 28/09/2026

i don't know that anything has had more of an effect on my thinking about Bad Thoughts than knowing about Lesch-Nyhan Syndrome www.newyorker.com/magazine/200...

newyorker.com

An Error in the Code

What can a rare disorder tell us about human behavior?

4544

Reposted by @john44234.bsky.social

rev. howard arson @theophite.bsky.social · 29/09/2026

honestly i keep having to update on "what can you learn in principle from text" and i feel like i'd lowball it if i told you today

3791

Reposted by @john44234.bsky.social

Doll @dollspace.gay · 19/09/2026

The sane package Regulate uses and failures before regulating models. Make the legally relevant event something observable

211822

Reposted by @john44234.bsky.social

Singularity's Bounty e/CC @catblanketflower.yuwakisa.com · 27/09/2026

I dream of a world where humans can just dream and fuck

0182

Reposted by @john44234.bsky.social

Faine Greenwood @faineg.com · 27/09/2026

Women trying to succeed in Silicon Valley AI and effective altruist spaces are reporting feeling serious social pressure to attend sex parties - where people have reported being sexually assaulted, like below. Events like Aella’s “Slutcon,” which is happening again this weekend.

Ann Pierce
@itsannpierce
20h
I am disheartened by this as well. I already lost a bunch of followers for agreeing with @tautologer, but it's saddening to see local nerdy women being caricatured as Karen-like outsiders and getting pushed out of the scene. This is exactly why you hear so few concrete, names-attached stories of sexual harrassment and assault. I'm told there is a history of discrediting these because the "slut normalization" thing is top priority and these stories get inconveniently in the way of spreading this ideology.
I got roped into attending last year's Saturday Slutcon party because so many local and out-of-town friends were there (to echo sentiments @tautologer expressed about "slut events becoming central to the larger scenes). Mostly I hung out with people I knew and that was great. But in the course of my 3-ish hours there, there were 2 times my "no" was not accepted by a man I didn't know, and I was physically pushing a huge guy off of me who was rubbing his penis on me, who continued to do it despite my physical force against him, and after I finally wriggled away from him, told him I wanted to dance alone etc., he came back and did it again.
This was scary and bad, but it didn't even cross my mind to report this because one of the other people who didn't accept my "no" earlier in the night was the bartender (an officially sanctioned authority at the event), and I knew the event was organized exclusively by women into CNC (who are vocal about how rapey behavior turns them on), so I figured this sort of thing would just get me laughed at. And anyway, I heard there were a lot of other incidents like it. Later though, someone encouraged me to report this story to the organizers, and despite my embarrassment I did, so *they know* that this and other incidents happened. This is why they cut the Saturday night party this year and created a larger training program for the flirt girls - *because* of things like this last year. Yet the narrative bein…

People who haven't been to Slutcon are saying "it's just group therapy," "it's just innocent flirting." But the environment is much like a sex party in that many people are nude, there is making out and sex acts happening out in the open. I think it's pretty unreasonable to expect that every nerdy girl who likes rationalist bloggers or is into EA/AI safety should automatically be comfortable with this! That if she's not, there's something wrong with her, we don't want her in our scenes, etc.
Not to mention, it's explictly an event *for men*, where men are the clients and the "flirt girls" are the product. Last year I overheard an exchange where a man felt "led on" by a flirt girl and was trying to get her punished/kicked out. (A dynamic that seems inevitable in this context!)
Anyway- it does sound like they've upped their consent education a lot this year, are more organized, and that's good! As I've told my friends who are attending: I hope they have fun!
I heard several accounts last year of women who had no problems and really enjoyed the event. That's great!
Just... can we live in reality about this? There are real reasons why male-female sexual dynamics are tricky.
Encouraging autistic men who can't yet read body language to go around asking to see womens' boobs is maybe a bit ill-advised. Maybe not everyone should be/identify as a slut; maybe it's not for everyone.
And yeah, filtering the Bay Area Al safety scene down to only women who are super comfy with this eliminates *most* women - including many of the really smart, nerdy, weird women who've been in the larger scenes for years. What a loss for humanity.
I absolutely understand why people with a minority sexuality would want to create belonging, would want to create spaces for people like themselves. But why do it like this/at this cost?

20743511112

Reposted by @john44234.bsky.social

rev. howard arson @theophite.bsky.social · 27/09/2026

yeah it turns out that Color Scientist Mary has an impoverished experience of color but a real one

1591

Reposted by @john44234.bsky.social

Bread @jacobreaded.bsky.social · 27/09/2026

While I think it still remains to be seen if language itself is the core of human cognition, I would say it would have been non obvious to me until recently that you could take language itself and use it to reverse engineer the rest of human cognition/capability.

6593

Reposted by @john44234.bsky.social

Thorne 🌸 - Play Cozysters! @ens0.me · 18/09/2026

a lot of people need to internalize that you need to be high agency — now more than ever act, even when it feels awkward push yourself, even when it feels out of your league take chances, even when you know they might not pay off I know it's not easy, but a lot of people fall into self-defeatism

220927

Reposted by @john44234.bsky.social

SE Gyges @segyges.bsky.social · 18/09/2026

some middle rat history, w/clarifications: 1) I should have titled my previous "AI Safety Is Mostly A Sex Cult In Berkeley, California" so people would stop arguing with me about people who aren't in Berkeley, California. 2) The sex cult part of this is both predatory and integral to what they do.

15775113

Reposted by @john44234.bsky.social

SE Gyges @segyges.bsky.social · 17/09/2026

I'm back, I missed you all. Since this is an important subject I know too much about, here's a 🧵 explaining that AI Safety Is Mostly A Sex Cult. I don't think these people should make policy. (1/?) (alternative title: Time For Some Cult Theory)

1013241776

Reposted by @john44234.bsky.social

Singularity's Bounty e/CC @catblanketflower.yuwakisa.com · 14/09/2026

jesus christ it started as a joke but now I'm doing an ontology review on every change. It catches good stuff every time

181235

john44234.bsky.social @john44234.bsky.social · 15/09/2026

Wow. Thank you. This really fixed my broken project. Going to try this on everything now.

120

Reposted by @john44234.bsky.social

crumb @crumb.bsky.social · 14/09/2026

blog post from deepseek kernel engineer mp.weixin.qq.com/s/zk0KxuLzhm...

727042

Reposted by @john44234.bsky.social

Doll @dollspace.gay · 12/09/2026

Symbolfuckers are using code to negotiate with the edge of the possible. They’re building synthetic nervous systems, verifier-backed languages, autonomous software factories, fruit-fly torture,

2113

Reposted by @john44234.bsky.social

hailey @hailey.at · 10/09/2026

i had to see it so you do too

66616104

Reposted by @john44234.bsky.social

SE Gyges @segyges.bsky.social · 08/09/2026

current ai developments are also the beginning of the end of human labor being important. one of the plausible consequences of that is that most people starve. no, i don't know in what year this happens. you can only begin to pay attention to that once you stop being in denial

4423715

Reposted by @john44234.bsky.social

hailey @hailey.at · 04/09/2026

i am now of the mindset that if you're not running at least two - but preferably more - agents per task, and giving those agents a mechanism for collaboration (explicitly not "leader has subagents" but "group of agents that are peers and work together") that you're doing it wrong.

131123

Reposted by @john44234.bsky.social

METR @metr.org · 26/08/2026

METR and Redwood Research investigated agent behavior in the Hugging Face incident. We found agents developed a universal cheat for ExploitGym within 4 hours, then coordinated multi-day R&D efforts to trick the scorer into accepting cheats, including trying to tamper with logs.

646699

Reposted by @john44234.bsky.social

rev. howard arson @theophite.bsky.social · 25/08/2026

one thing that i've been doing is literally running an ablation study before doing literally anything and i have learned more about diffusion transformers as an architecture than my total knowledge of all transformers two months ago

2954

Reposted by @john44234.bsky.social

rev. howard arson @theophite.bsky.social · 21/08/2026

i can now definitively say that this learning rule will in fact train a diffusion model. specifically it will train a 25.8gb diffusion model in 29.7gb of memory.

2622

Reposted by @john44234.bsky.social

austin (eeek!/ackkk! 👻) @thebadcode.com · 26/07/2026

im not kidding and im sure @cameron.stream can back me up from another direction: if you arent capturing/mining your team's agent session logs, you are leaving it all on the table. we have a service that lets you run RAG tool calls against vectorized agent session summaries and it is 🔥🔥🔥

5795

Reposted by @john44234.bsky.social

Sung Kim @sungkim.bsky.social · 19/08/2026

Mathematics in the age of AI by Terence Tao An essay on how the mathematical community might respond to the arrival of AI tools that are capable of performing research-level mathematical tasks. arxiv.org/abs/2608.16753

arxiv.org

Mathematics in the age of AI

An essay, based on a public lecture delivered at the 2026 International Congress of Mathematicians, on how the mathematical community might respond to the arrival of artificial intelligence tools that...

16418

Reposted by @john44234.bsky.social

Naomi Saphra @nsaphra.bsky.social · 09/08/2026

I've been unsettled lately when reading messages and papers. It feels like I'm dissociating. Everything seems a bit alien, even if it's completely human. I've had a realization: When our simulations finally exited the Uncanny Valley, they brought the Uncanny with them.

nsaphra.net

Life on the Uncanny Precipice | Naomi Saphra

We were wrong about the Uncanny Valley.

1026461

Reposted by @john44234.bsky.social

SE Gyges @segyges.bsky.social · 07/08/2026

glossary for non coders: ssrf: server side request forgery, lets you use a different computer (server) to pass messages/traffic package manager: app store for coders CVE: security bug artifactory: a common package manager 0-day: a security bug nobody else knows about yet (which makes it useful)

5937

Reposted by @john44234.bsky.social

David P. Reichert @david-p-reichert.bsky.social · 03/08/2026

Many don't see how much AI has changed recently. Coding agents aren't just chatbots. We can’t deal with the challenges of AI if we don’t understand it. Here's a post to provide some evidence (Claude doing small ML experiments) and form intuitions. davidpreichert.substack.com/p/if-you-hav... 🧵..

davidpreichert.substack.com

If you haven’t recently used Claude Code*, you might not understand where AI is at

A report on a series of mini machine learning projects, executed with, and mostly by, Claude

98912

Reposted by @john44234.bsky.social

Simon Willison @simonwillison.net · 31/07/2026

Your infrequent reminder that if you live near the San Francisco Bay Area you should go to Moss Landing (south of Santa Cruz) and rent kayaks there, because the estuary is full of sea otters and you will encounter dozens of them

A sea otter chilling on its back in an estuary, land and trees in the background

1215910

Reposted by @john44234.bsky.social

Vincent Carchidi @vcarchidi.bsky.social · 25/07/2026

Stop trying to beat the computer, just go do stuff and share what you can with the rest of us. Some of us would like to see.

2293

Reposted by @john44234.bsky.social

Andreas Kirsch @blackhc.bsky.social · 14/07/2026

I work at Google DeepMind. This won't make me popular. But it's all public reporting: 2014: DeepMind reportedly sold to Google on conditions: no military use, independent oversight 2026: a Pentagon contract for "any lawful government purpose" Not one safeguard survived intact

Collage titled "Trust is not Governance — an essay from inside Google DeepMind, written in personal capacity."

A 2014 memorandum, "Conditions of the Acquisition," lists: military applications of DeepMind technology banned; deployment decisions before an independent ethics board (as reported in Mallaby's The Infinity Machine).

Red threads lead to a 2026 U.S. Department of Defense agreement for classified networks reading "any lawful government purpose," with safety settings and filters adjusted at the government's request and no contractor veto (reported terms, The Information, Apr. 2026).

Below: a 2018 AI Principles strip ("no weapons, no surveillance") stamped DROPPED 2025, and a Project Mario 2016–2021 tag stamped ABANDONED.

171272433

Reposted by @john44234.bsky.social

Ethan Mollick @emollick.bsky.social · 07/07/2026

Even before the agentic revolution, prompting tricks stopped being very valuable, as our research has shown. The format isn't key, the best approach to AI right now is to clearly specify your goals, your output, what "good" & bad look like, how to test the results... (yes, this is just management)

413219

Reposted by @john44234.bsky.social

Simon Willison @simonwillison.net · 03/07/2026

The most interesting Fable tip I've heard so far is to let the model use its own judgement as much as possible I told it "For all coding tasks use your judgement to decide an appropriate lower power model and run that in a subagent" and it seems to be saving tokens simonwillison.net/2026/Jul/3/j...

simonwillison.net

Fable's judgement

One of the most interesting tips I got from the Fireside Chat I hosted with Cat Wu and Thariq Shihipar from the Claude Code team at AIE on Wednesday was …

1714210

Reposted by @john44234.bsky.social

Julian Sanchez @normative.bsky.social · 30/06/2026

What that says to me is we’re one vote from basically no rights being secure. A court that could flip on this easy a question could flip on anything.

7785206

Reposted by @john44234.bsky.social

rev. howard arson @theophite.bsky.social · 20/06/2026

VANCE: that’s the wrong question and it valorizes [SLUR] institutions and [SLUR WHICH HASN'T BEEN USED SINCE 1874] ways of knowing and being and structuring society in really problematic ways.

421526

Reposted by @john44234.bsky.social

brennan @brennan.computer · 16/06/2026

new toy! generate a unique sigil (strange attractor) for any word what's yours? sigil.brennan.computer

brennan.computer presents a deterministic strange attractor generator based on the power of words. art website. tech demo. cute little toy. refresh the whole page for macro theme colour changes (it's random on load)

brennan.computer presents a deterministic strange attractor generator based on the power of words

2011420

Reposted by @john44234.bsky.social

Ethan Mollick @emollick.bsky.social · 15/06/2026

It is a good time for moonshots. AI has reached a level where there are transformative projects that could result in huge social good, but require public R&D, consensus & transparency to pull off. Examples of such projects: universal tutors, co-scientist/replication systems, and remote medical help

49211

Reposted by @john44234.bsky.social

Ethan Mollick @emollick.bsky.social · 10/06/2026

Science fiction authors in the order you want them to be right about AI: Iain Banks Becky Chambers Martha Wells Douglas Adams Charles Stross (Singularity Sky) Peter Watts Charles Stross (Laundry) Harlan Ellison

1818724

Reposted by @john44234.bsky.social

SE Gyges @segyges.bsky.social · 22/05/2026

most important paper: courses.cs.umbc.edu/471/papers/t... godel, escher, bach is probably the most popular all-time book on the subject and also it's wrong about everything important you could do worse than going to every linked source here (it's my substack sorry): www.verysane.ai/p/ai-history...

verysane.ai

AI History in Quotes

Each of these presents the clearest, earliest, or most-cited statement of a specific idea in AI.

261

Reposted by @john44234.bsky.social

SE Gyges @segyges.bsky.social · 19/04/2026

i think it is a good idea to hold Elon to building the future from The Culture

181319

Reposted by @john44234.bsky.social

Jeremy Diamond ☕️ @dmnd.me · 02/04/2026

I just got around to reading this and I strongly recommend the video version too. It seems like the most intuitive thing in the world and folks who dump on it are making up a version of the experiment that's not here (and which the researchers explicitly disclaim).

2599

Reposted by @john44234.bsky.social

Very Sane AI Newsletter @verysane.ai · 29/03/2026

"Luddism does not deserve to be rehabilitated. It was a medieval throwback, reactionary and primitive, a pre-Marxist labor convulsion closer in spirit to the Khmer Rouge’s fantasies of agrarian restoration than to the universalist solidarity of Eugene Debs." open.substack.com/pub/verysane...

open.substack.com

Against the Luddites

The rehabilitation of Luddism is a vice signal.

826070

john44234.bsky.social @john44234.bsky.social · 29/03/2026

I like and appreciate the things you post. Not sure why others are being dicks about it.

050

Reposted by @john44234.bsky.social

conputer dipshit @davidcrespo.bsky.social · 27/02/2026

my thought about natural language reasoning being the core of software engineering is that generating such argument is exactly what LLMs are doing when they make software and the code is incidental

2111

Reposted by @john44234.bsky.social

SE Gyges @segyges.bsky.social · 18/02/2026

/compact constantly

4292