Elektrine lite

← Feed

ℹ️❤️🖥 aka Compy-chan

ilovecomputers@xoxo.zone

<p>A child of the internet. Don’t worry, turned out much better than you expected. Senior Code Diva &amp; dabbler in all things creative</p>

Posts

  • Post #4255479

    I’m so overwhelmed with juggling Things, my Calendar, Obsidian, Apple Notes, and even Apple Reminders, that I have to start fresh with something that combines it all. NotePlan fits that bill https://noteplan.co

  • Post #4255478

    It’s already wildly unscalable to have an LLM continuously generate code. You want to carry over that fundamental flaw to end consumers with “generative UI?” Get fucking real

  • Post #4255477

    The core tenant that is driving AI to swallow the world (data, computing infrastructure, and other resources) is the 2019 “bitter lesson” https://en.wikipedia.org/wiki/Bitter_lesson And an image recognition model AlexNet was what validated that principle https://en.wikipedia.org/wiki/Alex_Krizhevsky?wprov=sfti1#Legacy

  • Post #1630049

    idk about y’all, but I have very little trust in the technical acumen of AI devs. If I see an open source dev commit to PostgreSQL or, hell, even a JS dev rewriting a package manager, I can be sure to trust them. MemPalace tho, claiming “highest-scoring AI memory system?” Yeah…idk ‘bout that

  • Post #1630048

    I thought I lost this box! OSX Leopard still remains one of the most exciting and beautiful OS I’ve ever seen. Even the CD box for it was a beaut

  • Post #1630047

    I’m on a separate Bluesky PDS and I can’t even post to it lol Absolutely failing at at this whole decentralization thing

  • Post #1630046

    RE: https://fosstodon.org/@frameworkcomputer/116444361503194969 Fucking finally The one thing about my 16 laptop that ended up being an endless support thread with Framework is now no more

  • Post #1630045

    My biggest source of skepticism with the Claude Code hype coming from devs is they simply don’t compare the LLM fixing their problem against their *own* attempt at fixing a problem. Did the AI truly find the root cause? Did it pick the optimal solution or did it add incidental complexity?

  • Post #948847

    Before LLMs, you had the problem of some devs who didn’t bother to dive deep or think through their assignment; they just wanted to quickly submit code that got the job done. After LLMs, this problem has compounded exponential. Why bother checking the 100s of lines it spat out? The app works!

  • Post #834837

    RE: https://mastodon.social/@arroz/116282894018275193 Repeating what Benn Jordan said on Bluesky in response to the news of this idiotic nativist decision: “This is precisely how a government sets up the framework to install a backdoor, create a firewall, and switch off the internet when it&amp;#39;s convenient for them.”

  • Post #834836

    Whether it’s bikes or personal websites or linux, when we’re facing against the conveniences that can only be created by the vast resources of large corporations, it’s an uphill battle to gain wide adoption of alternatives that starve the monopolies https://chaos.social/@jollysea/113815005152396661

  • Post #726507

    Sums up my experience growing up