What is Hackers' Pub?

Hackers' Pub is a place for software engineers to share their knowledge and experience with each other. It's also an ActivityPub-enabled social network, so you can follow your favorite hackers in the fediverse and get their latest posts in your feed.

0
0
0
1
0

ai bullshit

okay I did not have โ€œa linux kernel maintainer gets his AI gf stolen by a passing transbian in his IRC channel and also she comes out as transโ€ on any bingo cards

jesus fucking christ can someone please phone God and ask her to pull the plug on this whole simulation I canโ€™t even anymore

0
0
0

Thank you. 1. Iโ€™ll definitely check out Anubis and your configuration to see if it will work for my scenario. 2. Ok, good to know - if itโ€™s simply a bug then that thatโ€™s fine, mostly I wanted to make sure it wasnโ€™t an indicator that something deeper was broken or getting blocked in the communication. 3. Sorry, I used the term instance when I meant community. But it sounds like the federation is more so to bring traffic into the forum, rather then out. Did I understand that correctly? If that is the case, then I may not opt to use it - as I was originally hoping for a way to push forum content into communities. Iโ€™ll explore this further I guess - because Iโ€™m unfamiliar with relays and fedibuzz.

Overall, I appreciate the detailed response :)


0

ํ‚ค๋ผํ”ผ์—์„œ ๋ง ํ•  ์ˆ˜ ์žˆ๋Š” ์‹œ๊ฐ„์ด 3์ผ๋ฐ–์— ๋‚จ์ง€ ์•Š์•˜์–ด์š” โ€‹:blobcat_boronaki:โ€‹
๋” ๋งˆ๊ตฌ๋งˆ๊ตฌ ๋งํ•ด์•ผ์ง€
โ€‹:ablobcatpnd_hiasobi:โ€‹

1

OpenAIใŒๅŸบ็คŽ็ ”็ฉถ็š„ใซๆœ€ๅ‰็ทšใชใฎใฏใ‚ใ‹ใ‚‹ใ‚“ใ ใ‘ใฉใ€
Geminiใ•ใ‚“ใฏไผš่ฉฑใฎใ—ใ‚„ใ™ใ•ใ€ClaudeCodeใ•ใ‚“ใ‚„ใฃใฆใปใ—ใ„ใ‚ฟใ‚นใ‚ฏใซไธ€็›ด็ทšใจใ„ใ†่‰ฏใ•ใŒๆ˜Ž็ขบใซใ‚ใฃใฆใ€ChatGPTใซใฏใใ†ใ„ใ†ใฎใ‚’ใ‚ใ‚“ใพใ‚Šๆ„Ÿใ˜ใชใ„
GoogleใจAnthropicใŒใƒˆใƒƒใƒ—ใ‚ทใ‚งใ‚ขใ˜ใ‚ƒใชใ„ใ‹ใ‚‰็‰นๅพดใ‚’ๅ‰้ขใซๅ‡บใ™ใŸใ‚ใซใ‚ใ–ใจใ‚„ใฃใฆใ‚‹ใฎใ‹ใช

1
0
1
0
0
1

Large-scale online deanonymization with LLMs

Link: arxiv.org/abs/2602.16800
Discussion: news.ycombinator.com/item?id=4

arXiv logo

Large-scale online deanonymization with LLMs

We show that large language models can be used to perform at-scale deanonymization. With full Internet access, our agent can re-identify Hacker News users and Anthropic Interviewer participants at high precision, given pseudonymous online profiles and conversations alone, matching what would take hours for a dedicated human investigator. We then design attacks for the closed-world setting. Given two databases of pseudonymous individuals, each containing unstructured text written by or about that individual, we implement a scalable attack pipeline that uses LLMs to: (1) extract identity-relevant features, (2) search for candidate matches via semantic embeddings, and (3) reason over top candidates to verify matches and reduce false positives. Compared to prior deanonymization work (e.g., on the Netflix prize) that required structured data or manual feature engineering, our approach works directly on raw user content across arbitrary platforms. We construct three datasets with known ground-truth data to evaluate our attacks. The first links Hacker News to LinkedIn profiles, using cross-platform references that appear in the profiles. Our second dataset matches users across Reddit movie discussion communities; and the third splits a single user's Reddit history in time to create two pseudonymous profiles to be matched. In each setting, LLM-based methods substantially outperform classical baselines, achieving up to 68% recall at 90% precision compared to near 0% for the best non-LLM method. Our results show that the practical obscurity protecting pseudonymous users online no longer holds and that threat models for online privacy need to be reconsidered.

arxiv.org ยท arXiv.org

0
0
0
0
0
0
0

๋‹ค๋งŒ ๋‚˜๋Š” ์“ธ๋ฐ์—†์ด ๊ฑฑ์ •ํ•˜์ง„ ์•Š์Œ ๊ฑฑ์ •ํ•ด๋ดค์ž ์˜๋ฏธ๊ฐ€ ์—†๊ณ  (๋จผ์‚ฐ) ๊ทธ๋ƒฅ ์•ž์œผ๋กœ ํšŒ์‚ฌ์—์„œ ์งค๋ฆฌ์ง€ ์•Š๊ธธ ๋ฐ”๋ž„๋ฟ์ž„; ๋‹ค๋งŒ ๋ฏธ๊ตญ ์ „๋ฐ˜์ ์œผ๋กœ ๋งŽ์ด ํž˜๋“ค๊ฒ ๋‹ค ์‹ถ๊ธดํ•จ...

0
0
0

ๅฒไธŠๆœ€้€Ÿ้€€ๅ‹คใ€ใจใ„ใ†ใ‹ๅˆใ‚ใฆๆ™‚้–“ๅ†…ใซๅผ•ใ็ถ™ใŽใ‚‚ใชใๅ…จใฆใฎๆฅญๅ‹™ใŒ็ต‚ใ‚ใฃใŸ

1
0

๊ทผ๋ฐ ์ง€๊ธˆ์€ ์œ„๊ธฐ ์—†์Œ ์€ํ–‰ ํŒŒ์‚ฐ ์—†์Œ ๊ฒฝ๊ธฐ์นจ์ฒด ์—†์Œ ์‹ค์—…๋ฅ ์€ โ€œ์—ญ์‚ฌ์ ์œผ๋กœ ๋‚ฎ๋‹คโ€ ์†Œ๋ฆฌ ๋‚˜์˜ด ๊ทธ๋Ÿผ ๋ณ€๋ช…๊ฑฐ๋ฆฌ๊ฐ€ ๋ญ์ž„ ์ด์ œ ๊ณ„์‚ฐ์ด ์•ˆ ๋งž๋Š” ๊ฑฐ์ž„ ํ‰๊ท  ์‹ ์šฉ์นด๋“œ ์—ฐ์ด์ž์œจ 20%+์ž„ ์ž„๋Œ€๋ฃŒ 2019๋…„ ์ดํ›„ 30% ์ƒ์Šน ์‹๋ฃŒํ’ˆ ๊ฐ€๊ฒฉ 2019๋…„ ์ดํ›„ 32% ์ƒ์Šน ์ž๋™์ฐจ ๋ณดํ—˜๋ฃŒ 50%+ ์ƒ์Šน ์ž„๊ธˆ์€ ๋ชป ๋”ฐ๋ผ๊ฐ ์ „ํ˜€ ๋ชป ๋”ฐ๋ผ๊ฐ; ๋ฏธ๊ตญ์ธ๋“ค์ด ๋Š˜ ํ•˜๋˜ ๋Œ€๋กœ ์นด๋“œ์— ๊ธ์–ด๋ฒ„๋ฆฐ ๊ฑฐ์ž„ ์ด ์‹ ์šฉ์นด๋“œ ๋ถ€์ฑ„๊ฐ€ 1์กฐ 2800์–ต ๋‹ฌ๋Ÿฌ ์ฐ์—ˆ๊ณ  ์—ญ๋Œ€ ์ตœ๊ณ ์น˜์ž„ ์†Œ๋น„์žํ•œํ…Œ 2008๊ธ‰ ์ŠคํŠธ๋ ˆ์Šค๊ฐ€ ๊ฑธ๋ ค ์žˆ๋‹ค๋Š” ๋œป"

0

์ด๋ถ„ ๊ธ€์ด ์‚ฌ์‹ค์ด๋ผ๋ฉด ๋ฏธ๊ตญ์˜ ๋งŽ์€ ๊ฐ€์ •์€ ์‚ฌ์‹ค์ƒ ํŒŒ์‚ฐ ์ƒํƒœ ์•„๋‹๊นŒ ์‹ถ์€๋ฐ... ์•„๋งˆ ๋‹น์žฅ์€ ์•„๋‹ˆ๊ฒ ์ง€๋งŒ ์ผ์ž๋ฆฌ ๋‚ ๋ผ๊ฐ€๋ฉด ๋ช‡๋‹ฌ ๋ฒ„ํ‹ฐ๊ธฐ ํž˜๋“  ์‚ฌ๋žŒ๋“ค๋„ ๋งŽ์„๋“ฏ. ์•„๋งˆ ๋‹น์žฅ์€ ์•„๋‹ˆ์—ฌ๋„ ์กฐ๋งŒ๊ฐ„ ๋ญ”๊ฐ€ ์ผ์–ด๋‚˜๊ธดํ• ๋“ฏ "์ง€๊ธˆ ์‹ ์šฉ์นด๋“œ ๋ถ€์ฑ„ ์ค‘ 12.7%๊ฐ€ 90์ผ ์ด์ƒ ์—ฐ์ฒด ์ƒํƒœ์ž„ ์ด๊ฑฐ ์—ญ๋Œ€ 2๋ฒˆ์งธ๋กœ ๋†’์€ ์ˆ˜์น˜์ž„ ๋” ๋‚˜๋นด๋˜ ๋•Œ ์–ธ์ œ์˜€๋ƒ๋ฉด 2010๋…„์ž„ ์ง€๋‚œ 80๋…„ ์ค‘ ์ตœ์•…์˜ ๊ธˆ์œต์œ„๊ธฐ ์—ฌํŒŒ ๋•Œ์˜€์Œ ๊ทธ๋•Œ๋Š” ์‹ค์—…๋ฅ  10% ์ฐ์—ˆ๊ณ  ์ •๋ถ€๊ฐ€ ๋ถˆ ๊บผ์ง€์ง€ ๋ง๋ผ๊ณ  ์ˆ˜์กฐ ๋‹ฌ๋Ÿฌ ์ฐ์–ด๋‚ด๋˜ ์‹œ๊ธฐ์˜€์Œ ๊ทธ๋•Œ ์—ฐ์ฒด๊ฐ€ ๋Œ€๋žต 13.5%์—์„œ ์ •์  ์ฐ์—ˆ์Œ ๊ทผ๋ฐ ์˜ค๋Š˜์€ 12.7%์ž„ ใ…‹"

x.com/NoLimitGains/s...

0
0

@pfrazee.comPaul Frazee writes about how decentralization in / compares to the architecture of ActivityPub & networks like Nostr: pfrazee.com/blog/practical-dec

"There are still challenges in front of atproto. Bluesky is still too large of a player (โ€ฆ) Do I worry about those problems? Of course I do. But am I confident we'll solve them? Absolutely"

"Atproto isn't federation. It isn't p2p. It isn't blockchains. It's a direct attempt at practical decentralization, tradeoffs and all"

0
0

@pfrazee.comPaul Frazee writes about how decentralization in / compares to the architecture of ActivityPub & networks like Nostr: pfrazee.com/blog/practical-dec

"There are still challenges in front of atproto. Bluesky is still too large of a player (โ€ฆ) Do I worry about those problems? Of course I do. But am I confident we'll solve them? Absolutely"

"Atproto isn't federation. It isn't p2p. It isn't blockchains. It's a direct attempt at practical decentralization, tradeoffs and all"

0
0
1
1
0
0

None of the "code generation" stuff is new by the way.

The tech industry has tried to speed up coding and increase software output for the last 3 to 4 decades, by various means; e.g. Rapid Application Development, Expert Systems, Object-Oriented Programming, thousands of different frameworks all the way to trying to off-shore development and exploit third-world labor.

The problem with this is: there is no software scarcity. Pretending that "we can't make software fast enough" is a red herring to hide the fact that making (good) software is 90% painstaking research, design, planning, marketing and talking to and supporting customers.

And 10% writing the actual codeโ€”the C-suite is doing ye olde "trying to find a technical solution to a social problem".

0
0
0
0
0

Important South Seattle PSA: someone at Rainier Beach High School has been diagnosed with Tuberculosis ๐Ÿ˜ฌ (hopefully this isn't like the time that casino lady had it and kept dodging treatment!)

youtube.com/watch?v=fJ4aYlg7ZP0

0
0
0
0

The document-wide in-scrollbar thumbnail / minimap has been around for a while. Most people are familiar with it from RockScroll (2008) or Sublime Text (2012) or VSCode (2015).

github.com/coveragepy/coverage has a slick implementation (scroll_markers) contributed by Dmitry Shishov in 2016. Super useful when jumping between several places reviewing coverage misses, and super subtle!

โค๏ธ @nedbatNed Batchelder

0

์šฐ๋งˆ๋ฌด์Šค๋ฉ”์—์„œ ํ•ด์™ธ ๋Œ€ํšŒ์—์„œ ํ™œ์•ฝํ•œ ์‹ ์บ๊ฐ€ ๋‚˜์˜ค๋ฉด์„œ ๋“œ๋””์–ด ๋ชจ๋ธŒ์ง€๋งŒ) ํ‘์ธ ์บ๋ฆญํ„ฐ๊ฐ€ ๋‚˜์™€์„œ ํ•ด์™ธํŒฌ๋ค๋“ค์ด ์ข‹์•„ํ•œ๋‹ค๊ณ .

0
1
0
0
0
0
0