Skip Navigation

InitialsDiceBearhttps://github.com/dicebear/dicebearhttps://creativecommons.org/publicdomain/zero/1.0/„Initials” (https://github.com/dicebear/dicebear) by „DiceBear”, licensed under „CC0 1.0” (https://creativecommons.org/publicdomain/zero/1.0/)N

nfultz

@ nfultz @awful.systems

Posts
3
Comments
176
Joined
3 yr. ago

  • Stubsack: weekly thread for sneers not worth an entire post, week ending 4th January 2026

    Jump

  • Stubsack: weekly thread for sneers not worth an entire post, week ending 28th December 2025

    Jump

  • Stubsack: weekly thread for sneers not worth an entire post, week ending 28th December 2025

    Jump

  • Firefox browser falls to AI. What do we do now?

    Jump

  • Waterfox lore - it got acquired by System1 of "Guy Who Runs Three Companies Called Fidelity But Not The Fidelity You Know Probably Doesn’t Care That There’s Already A Company Called System1 That Does That Same Thing As The System1 His SPAC Is Buying. Just saying." 1 and then went private again 2, presumably they bought it back after the stock predictably tanked. Subprime adtech is a strange place.

    IMHO should just bring back iceweasel, but what do i know.

  • Firefox browser falls to AI. What do we do now?

    Jump 
  •  
        
.comment-node :has(.text-danger) {
    font-family: "Comic Sans MS", "Comic Sans" ;
}

    Sorry (not sorry).

  • Stubsack: weekly thread for sneers not worth an entire post, week ending 21st December 2025

    Jump

  • Backgammon is "easier" than chess or go, but it has dice, so it not (yet) been completely solved like checkers. I think only the endgame ("bearing off") has been solved. The SOTA backgammon AI using NNs is better than expert humans but you can still beat it if you get lucky. XG is notable because if you ever watch high stakes backgammon on youtube, they will run XG side by side to show when human players make blunders. That's how I learned about it anyway.

  • Stubsack: weekly thread for sneers not worth an entire post, week ending 21st December 2025

    Jump

  • Stubsack: weekly thread for sneers not worth an entire post, week ending 21st December 2025

    Jump

  • Why is WSJ rehashing six month old whitepapers? Slow news week /s

    Anything new vs the last time it popped up? https://www.anthropic.com/research/project-vend-1

    EDIT:

    In mid-November, I agreed to an experiment. Anthropic had tested a vending machine powered by its Claude AI model in its own offices and asked whether we’d like to be the first outsiders to try a newer, supposedly smarter version.

    Whats that word for doing the same thing and expecting different results?

  • Stubsack: weekly thread for sneers not worth an entire post, week ending 21st December 2025

    Jump

  • https://www.theinformation.com/articles/can-ucla-replace-teaching-assistants-ai

    Miller’s team also recently used software from startup StackAI to develop an AI-powered app that writes letters of recommendation, saving faculty members time. Faculty type basic details about a student who has requested a letter, such as their grades and accomplishments, and the app writes a draft of the full letter.

    AI is “one of those things that you might worry could dehumanize the process of writing recommendation letters, but faculty also say that process [of manually writing the letters] is very labor intensive,” Miller said. “So far they’ve gotten a lot out of” the new app.

    Anyone using this thing should be required to serve on the admissions committee. LoRs aren't for generic B+ students that you don't even remember, just say no.

    I googled stackai, saw their screenshots and had ptsd flashbacks of mid 2000s alteryx. why do we keep reinventing no-code drag-and-drop box-and-arrow crap.

  • Stubsack: weekly thread for sneers not worth an entire post, week ending 21st December 2025

    Jump

  • I went down a punch-card history rabbit hole today on the empirical software engineering discord. TIL:

    We have been living in a world 10 columns too short. Think of all the HN headlines we could have had instead...

    I imagine it was like VHS vs Beta only with pocket protectors.

  • Stubsack: weekly thread for sneers not worth an entire post, week ending 21st December 2025

    Jump

  • Stubsack: weekly thread for sneers not worth an entire post, week ending 21st December 2025

    Jump

  • Stubsack: weekly thread for sneers not worth an entire post, week ending 21st December 2025

    Jump

  • https://kevinmd.com/2025/12/why-ai-in-medicine-elevates-humanity-instead-of-replacing-it.html h/t naked capitalism

    Throughout my nearly three decades in family medicine across a busy rural region, I watched the system become increasingly burdened by administrative requirements and workflow friction. The profession I loved was losing time and attention to tasks that did not require a medical degree. That tension created a realization that has guided my work ever since: If physicians do not lead the integration of AI into clinical practice, someone else will. And if they do, the result will be a weaker version of care.

    I feel for him, but MAYBE this isn't a technical issue but a labor one; maybe 30 years ago doctors should have "led" on admin and workflow issues directly, and then they wouldn't need to "lead" on AI now? I'm sorry Cerner / Epic sucks but adding AI won't make it better. But, of course, class consciousness evaporates about the same time as those $200k student loans come due.

  • Stubsack: weekly thread for sneers not worth an entire post, week ending 21st December 2025

    Jump

  • LA Unified School District forces unfiltered AI on kids

    Jump

  • Stubsack: weekly thread for sneers not worth an entire post, week ending 14th December 2025 - awful.systems

    Jump

  • I did it, I went and made a Official Public Comment IRL:

    In UCLA's Strategic Plan, Goal 1 is to "Deepen our engagement with Los Angeles" and Goal 5 is to "Become a more effective institution". By engaging with Los Angeles businesses, UCLA can get both better terms, prices, and services, and support the local economy. Buy Local, Spend Local.

    The federal government encourages this with Small Business Innovation Research and Small Business Technology Transfer grants, among other things. Furthermore, the State of California requires a portion of its spending go toward certified Small Businesses.

    And yet, the University apparently awarded a contract reportedly worth hundreds of thousands to millions of dollars to OpenAI. I have not found any documentation of an open Request for Proposals or competitive process for that award.

    My question is:

    If there was an RFP, where was it publicly posted, and if there was no RFP, why not, and were Los Angeles vendors or small businesses evaluated as alternatives, as recommended by UC policy and state law?

    Given the scale of this spending and the context of a budget crisis, transparency, compliance, and small-business participation are critical to our effectiveness and engagement.

    I’m asking for clarity on how this decision was made, how it aligns with procurement guidelines and University goals, and how DTS plans to ensure that local and small businesses are meaningfully included moving forward.

    Thank you.

  • Stubsack: weekly thread for sneers not worth an entire post, week ending 14th December 2025 - awful.systems

    Jump

  • My college used the green Russell Norvig text, which had (checking...) 12 pages on neutral nets out of 1000 pages. I liked the class well enough, but we used Java 1.3 and lisp would have been better.

  • Stubsack: weekly thread for sneers not worth an entire post, week ending 14th December 2025 - awful.systems

    Jump

  • New Time War lore dump

    I need to quit clicking, my yt recommendations are so cursed now.

    Love the Edgar Allen Poe thing at 3:30 though, wtf. Also where did he find the sephiroth at 4:19?

  • Stubsack: weekly thread for sneers not worth an entire post, week ending 14th December 2025 - awful.systems

    Jump

  • Is there even any young people we could plausibly call whippersnappers on orange site anymore, it feels like they're all well into their 30s/40s at this point.

    I miss n-gate but that was what, 8 years ago.

    But in fairness to actual whipper snappers, and to your point, the '56 Dartmouth Workshop forward privileged Symbolic AI over anything data driven up through the first AI winter (until roughly the 90s and the balance shifted) and really warped the disciplines understanding of its own influences and history - if 70s RMS was taught anything about Neural Nets, it's relevance and importance would probably have been minimized in comparison to expert systems in lisp or whatever Minsky was up to.