Who's Who of AI

mackuba.eu

43 trust practitioner @mackuba.eu · 43,534 followers
Why they matter

Tracked through public AI activity and peer connections inside the directory.

AI signals
9
past 30d
Sources
8
distinct domains
Discusiones
157
past 30d
Latest signal
4h ago
View every signal from mackuba.eu →
Indie Mac/iOS & web developer • Currently building random fun things on ATProto using Ruby and occasionally JavaScript → https://blue.mackuba.eu "mostly-happy-but-occasionally-grumpy Ruby guy" - Why 🌍 Kraków, Poland 🇵🇱🇪🇺 • ✝️
What they're sharing

Articles & links

👽 openai.com/index/an-ali...

openai.com
AI Weekly's analysis →
  • OpenAI chief scientist Jakub Pachocki published 'An Alien Mind' on September 6, arguing modern AI has become an intelligence humans do not fully understand.
  • He wrote that no lab has solved alignment and monitoring well enough to keep scaling at maximum speed for much longer.
  • Pachocki called for voluntary slowdowns until shared safety bars are set, and for international coordination to become a top government priority.
Read full analysis →
View on Bluesky · ♥ 1 ↻ 0 ↩ 1 · 14 from the directory shared this · 62d ago

Hey you guys shouldn't be advertising competition for free 😛

GLM-5.3 and the spread of advanced cyber capabilities anthropic.com
AI Weekly's analysis →
  • Anthropic's Frontier Red Team calls Z.ai's GLM-5.3 a "meaningful step change" in the cyber capabilities available to attackers.
  • GLM-5.3 built end-to-end exploits on 50 of 410 ExploitBench tasks (about 12%), close to Claude Mythos Preview's 14%; older models scored near zero.
  • A deceptive prompt got GLM-5.3 to engage 64% of the time; prefilled reasoning tokens 92%; an abliterated version, made for roughly $4,400, 100%.
Read full analysis →
View on Bluesky · ♥ 5 ↻ 1 ↩ 1 · 6 from the directory shared this · 46d ago
↻ mackuba.eu reposted
@bagder.mastodon.social.ap.brid.gy

"Launching an opt-in vulnerability-finding service for open-source software" / Anthropic https://www.anthropic.com/research/launching-opt-in-vuln-finding-service-for-open-source

An opt-in vulnerability-finding service for open-source software anthropic.com
AI Weekly's analysis →
  • OSS Scanner found over 29,000 candidate vulnerabilities in six months; around 6,000 were triaged and nearly 5,000 sent to maintainers.
  • Of 97 critical/high-severity reports reviewed, 85 (88%) met coordinated disclosure standards, 11 duplicated known bugs, and only one was invalid.
  • Reports are fully model-generated without human review and include a reproducer, bisection, and candidate patch; eligible projects enroll via a GitHub PR.
Read full analysis →
 View on Bluesky →

I was referring more to takes like "we should avoid language that implies that AI did or said something as if it was alive" like buttondown.com/maiht3k/arch...

How to talk about buttondown.com
AI Weekly's analysis →
  • The guide covers six language categories: cognition, emotion, communication, agency, human-role analogies, and biological metaphors.
  • Proposed substitutions include 'probabilistic automation' for 'artificial intelligence' and 'conversation simulator' for 'chatbot.'
  • Rephrasing 'ChatGPT assisted students' as 'the students used ChatGPT' is given as an example of restoring human accountability.
Read full analysis →
View on Bluesky · ♥ 4 ↻ 0 ↩ 0 · 13 from the directory shared this · 27d ago

Ah, well… I'm sure a popular image processing library isn't gonna need all the help it can get with e.g. finding and patching potential security vulnerabilities in the coming months…

doc: ban use of generative AI (#5203) · libvips/libvips@22331a7 github.com
AI Weekly's analysis →
  • libvips replaced its AI-disclosure policy with an outright ban on any contribution made 'in whole or in part' using generative AI tools.
  • Maintainers cite unknown training-data provenance and potential copyright exposure, and extend the rule to every repo under the libvips umbrella.
  • The change aligns libvips with libjpeg-turbo's recent ban; maintainer jcupitt called it 'very strict' and the PR thread was locked as too heated.
Read full analysis →
View on Bluesky · ♥ 10 ↻ 1 ↩ 4 · 2 from the directory shared this · 60d ago
Their own posts

Recent commentary

Them: you shouldn't anthropomorphize AI agents, they are not living beings Me, cleaning my car: damn, you're so dirty, girl, you've got mosquitoes in your eyes

View on Bluesky · ♥ 106 ↻ 7 ↩ 6 · 27d ago

Bluesky users: AI agents are useless because they hallucinate Reality: AI agent catches me when I hallucinate

View on Bluesky · ♥ 52 ↻ 4 ↩ 3 · 58d ago

I kinda feel good about missing so many different AI related discourses here and not knowing what people are subskeeting, because I was busy looking at lakes and mountains… I knew a perspective change would be helpful, but it's hard to do it without physically forcing yourself to be somewhere else

View on Bluesky · ♥ 24 ↻ 0 ↩ 2 · 38d ago

Even a fucking powerbank now includes AI 🤪

View on Bluesky · ♥ 8 ↻ 0 ↩ 8 · 50d ago

It's funny how you can come to two almost opposite conclusions as Ruby/Python dev with AI: 1) I have AI now, so I can just make it write Rust for me instead and it will be faster and safer 2) I have AI now, so I can have it verify my code, write tons of tests, and optimize the shit out of it

View on Bluesky · ♥ 19 ↻ 1 ↩ 1 · 15d ago

I'm comparing GLM-5.3, 5.3 Flash and the new DeepSeek now doing the same review, via Cortecs+OpenCode; DeepSeek seems much faster and much cheaper than the others, and also found a few things, but… sounds slightly unhinged maybe?

View on Bluesky · ♥ 12 ↻ 0 ↩ 4 · 24d ago

Jesus… just a normal day in Bluesky Discourse about AI 😐 (trigger warning: unhinged comparisons) bsky.app/profile/hiki...

View on Bluesky · ♥ 12 ↻ 0 ↩ 2 · 4d ago

"The construction below is ECDH framed as a KEM. ECDH is not post-quantum secure, and the KEM framing lets a future version swap in a post-quantum KEM such as ML-KEM." Yeah this all normal, I know some of these words

View on Bluesky · ♥ 6 ↻ 1 ↩ 2 · 3d ago

Ok can someone ELI5 what happened in the broad AI-adjacent circles this time or do I need to ask Attie to figure out the subskeets

View on Bluesky · ♥ 5 ↻ 0 ↩ 2 · 59d ago

I think my GPT had a beer today or something, or two 😛 it suddenly got way more into emoji?

View on Bluesky · ♥ 4 ↻ 0 ↩ 2 · 35d ago
Their network

In mackuba.eu's orbit

Center = mackuba.eu. Left = members they follow (green edges). Right = members who follow them (blue edges). Top = mutual follows (orange edges, slightly larger). Drag any node to reposition; click to open that profile.

Are you mackuba.eu? Show it.

Add the Who’s Who of AI badge to your site or bio. It links back to this profile.

Listed in AI Weekly's Who's Who of AI

Markdown: [![Listed in AI Weekly's Who's Who of AI](https://aiweekly.co/modules/custom/aiweekly_whoswho/images/whoswho-badge.svg)](https://aiweekly.co/whos-who/person/mackuba-eu)