• 1 Post
  • 14 Comments
Joined 1 year ago
cake
Cake day: July 6th, 2023

help-circle
  • Yeah, all training ends up being pattern learning in some form or fashion. But acceptable patterns end up matching logic. So for example if you ask ChatGPT a question, it will use its learned pattern to provide its estimate of the correct ouptut. That pattern it’s learned encompasses/matches logical processing of the user input and the output that it’s been trained to see as acceptable output. So with enough training, it should and does go from simple memorization of individual examples to learning these broad acceptable rules, like logic (or a pattern that matches logical rules and “understanding of language”) so that it can provide acceptable responses to situations that it hasn’t seen in training. And because of this pattern learning and prediction nature of how it works, it often “hallucinates” information like citations (creating a novel citation matching the pattern its seen instead of the exact citation that you want, where you actually want memorized information) that you might ask of it as sources for what its telling you.


  • I’m less worried about a system that learns from the information and then incorporates it when it has to provide an answer (ex. learning facts) than I am of something that steals someone’s likeness, something we’ve clearly have established people have a right to (ex. voice acting, action figures, and sports video games). And by that extension/logic, I am concerned as to whether AI that is trained to produce something in the style of someone else, especially in digital/visual art also violates the likeness principle logically and maybe even comes close to violating copyright law.

    But at the same time, I’m a skeptic of software patents and api/UeX copyrighs. So I don’t know. Shit gets complicated.

    I still think AI should get rid of mundane, repetitive, boring tasks. But it shouldn’t be eliminating creative, fun asks. It should improve productivity without replacing or reducing the value of the labor of the scientist/artist/physician. But if AI replaced scribes and constructionists in order to make doctors more productive and able to spend more time with patients instead of documenting everything, then that would be the ideal use of this stuff.







  • Servo Folk. It’s one of the actions by Mitchell Baker that I disapproved of. Remember that the Rust programming language came out of Mozilla, right? It was being designed to create a fast and secure web engine by a related team. This Web Engine was of course Servo, written in Rust. Mozilla than took parts of their work and incorporated it into the Gecko web engine that runs Firefox, which was the Quantum Update. That’s where you saw the major speed up in Firefox to catch up to and beat Blink in many cases. Mitchell Baker a couple of years later made a move to lay off the Rust and Servo folk and spin out those projects so that they wouldn’t be Mozilla’s problem anymore, discontinuing their funding. She then proceeded to give herself a huge raise all while Mozilla’s market share had fallen to ~3%. It ticked me off needless to say.

    Have you heard of Electron? It’s the use of Chromium’s Blink web engine to run web apps as individual programs. Applications like Signal, Ferdi, Atom text editor, VS Code (the most popular IDE for developers) all use electron. I asked myself for years why isn’t there a Gecko equivalent of Electron? The answer is that Gecko’s way too old and janky (cobbled together over decades since the Netscape Navigator days), making it too difficult to work with. But the Servo project, being a completely fresh web engine written in Rust, is looking to play that role as its immediate functional goal. It’s a smaller, more attainable goal before it becomes a full fledged web engine that competes with the likes of Gecko, Blink, and Webkit (Safari and also what Blink’s based off of) to run a full fledged browser. The Servo project was out in the wilderness for a while before coming back to life in 2023.

    https://servo.org/




  • Those complaint websites are tailored to the customers who suffer from the decline in competition. We are suffering from Google using its market position to kill our user experience and options. As I understand, it’s classic monopoly abuse.

    In the 20th century, the US broke up the Hollywood model where companies owned both the studios and the theaters (how you have 20th century Fox (or just 20th century now) and Fox theaters). Google owning 75% online advertising and 75% of web browser share is a clear conflict of interest and you can see it from how they’re pushing things like Manifest V3 via their browser (especially when you consider how Chrome is the default browser on their phones), now that it’s the only browser that developers are increasingly starting to support.

    If you follow that model, one thing that’s going to have to be done is to have Chrome/Chromium browser development be broken away from Google proper. Google can’t fund the developers any longer.



  • Look, if Lemmy, NPR, and PBS can happen, then it’s always possible to fork Firefox (or throw more weight behind the Servo folk who are moving towards developing the Rust web engine towards embedded applications to get it up to speed faster for general web browsing) if Mitchell Baker and search revenue approach to funding Firefox is getting in the way of having a fast, private, and secure browser for everybody.

    But enough woah is me and our obstacles are overwhelming on here. In this case, if we do nothing, we get nothing. Especially if you’re right that the Mitchell Bakers of the world are not behind us. I know we at least have an ally in the EFF.