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/)L
Posts
4
Comments
118
Joined
3 yr. ago

  • deleted by creator

  • deleted by creator

  • getoffpocket.com, my guide to Pocket alternatives, just got a redesign

    Jump
  • deleted by creator

  • deleted by creator

  • I was paraphrasing and trying to be nice. Fine, you didn't say humans yearn for the workplace. You said humans existentially require the workplace.

    I think if AI replaces humans in the workplace, even with UBI, humans would cease to exist shortly thereafter as our lives will have become meaningless

  • According to technical experts, internet service providers across the country have begun implementing a rule that limits data transfers from sites using Cloudflare to just the first 16 kilobytes. This technique is relatively subtle but effective: very lightweight, basic websites can still load, creating a façade of normal internet function, while modern, media-rich sites are effectively broken.

    16 KB per website? What part of the normal internet is that small? What part of the indie web is that small?

    e.g. look at the smallest sites on https://512kb.club/

    Or is this just 16kb per request, which would make more sense with the following explanation:

    Analysts report that similar throttling is also being applied to other major western hosting providers popular with Russian users, including Germany’s Hetzner and the US-headquartered DigitalOcean... [they] are widely used by Russians to host private VPN servers, which allow them to bypass the Kremlin’s ever-widening blocklists.

    AFAIK, VPNs maintain a long-standing connection that would definitely use more than 16kb at a time.

  • deleted by creator

  • deleted by creator

  • deleted by creator

  • I see no reason to engage with, or trust anything created by, a bullshit generator. If Digg claims to "care" about the humans, then making the top comment into a brick wall (which has zero accountability) is a funny way of showing it.

    But then again, I'm sure their privacy policy also says they care about your privacy.

  • this would-be Reddit competitor, built for the AI era

    Oh no...

    The founders think that the internet is being flooded with bots and AI agents, which will create demand for online communities like Digg that foster real human connections.

    Okay, Digg has my cautious attention...

    Beneath posts, Digg is leveraging AI to summarize the article’s content.

    And they lost me.

  • deleted by creator

  • deleted by creator

  • deleted by creator

  • deleted by creator

  • OpenAI has gotten virtually unlimited funding for years. It has first dibs and deep discounts on Microsoft data centers.

    And somehow, despite every single trade restriction, multiple random startup companies in China (that don't even know how to secure their own databases) manage to make LLMs that outperform it.

    I'm not saying that because Chinese companies are uniquely cool. I'm saying that because this whole AI thing is uniquely stupid.

  • I figured as much. Even this line...

    M1's capabilities are top-tier among open-source models

    ... is right above a chart that calls it "open-weight".

    I dislike the conflation of terms that the OSI has helped legitimize. Up until LLMs, nobody called binary blobs "open-source" just because they were compiled using open-source tooling. That would be ridiculous

  • Locked

    Did lemm.ee already shut down?

    Jump
  • I've been able to run distillations of Deepseek R1 up to 70B

    Where do you find those?

    There is a version of Deepseek R1 "patched" with western values called R1-1776 that will answer topics censored by the Chinese government, however.

    Thank you for mentioning this, as I finally confronted my own preconceptions and actually found an article by Perplexity that demonstrated R1 itself has demonstrable pro-China bias.

    Although Perplexity's own description should cause anybody who understands the nature of LLMs to pause. They describe it in their header as a

    version of the DeepSeek-R1 model that has been post-trained to provide unbiased, accurate, and factual information.

    That's a bold (read: bullshit) statement, considering the only altered its biases on China. I wouldn't consider the original model to be unbiased either, but apparently perplexity is giving them a pass on everything else. I guess it's part of the grand corporate lie that claims "AI is unbiased," a delusion that perplexity needs to maintain.