• 1 Post
  • 80 Comments
Joined 2 years ago
cake
Cake day: March 26th, 2025

help-circle


  • The court filing is linked in the first paragraph and worth a read. The article is a bit vague because the claims in the filing are pretty sweeping, but to list some highlights:

    • Brave started selling copyrighted scraped articles in bulk from their search index to AI companies
    • Brave built explicit and targeted functionality into their browser to circumvent newspaper paywalls
    • Brave repeats long, verbatim sections of NYT & other papers’ material on request through their AI web search feature

    I’m neither an attorney nor a huge fan of IP rights writ large, but in my reading, the newspapers are making a pretty strong case that Brave is consciously stealing and reselling copyrighted material in unaltered form for monetary gain.




  • It’s the backbone for a lot of other software, the Wordpress ecosystem most notably. Inertia is the big factor in terms of popularity.

    Old school PHP was all about the LAMP stack (Linux as your operating system, apache as your web server, mysql as your database, php as the runtime apache would use to serve dynamic documents). The ecosystem was the real benefit that made it sensational for the era; all the pieces were designed to play nicely with one another.

    It was also really noob friendly: You could write basic html, copy it to your web server, reload in browser, and see changes. Then to make things dynamic, you sprinkle in a few <?php ?> tags in your html, copy it to the server, and it starts working! The code inside those php tags gets run by the web server and their output is written into the HTML. No reloading or resetting apache needed, and you need very little programming knowledge to start.

    The modern language is… serviceable, but lost the key features that made it great.








  • My extremely back-of-envelope estimate would be that it’s 100,000x slower to run inference on an IoT botnet compared to a data center GPU.

    100x penalty for cpu inference, 10x penalty for cpu quality, 10x penalty for crappy ram + disk, 10x sync penalty because you have to split up layers across nodes.

    And this is ignoring the standard showstopper issues you get with high-bandwidth distributed networking applications with spottily connected nodes.

    If you wanna talk about runaway AI risk, then the real threat to humanity is a frontier model with unauthorized data center TPU access, not IoT. The call is coming from inside the house, and they even spell this out at the beginning before pivoting to the fearmongering.