Again I’m in favor of choosing browsers on install, but lots of Chrome installs on Windows is not the same as being the default.
So much so that you even get this annoying popup from Edge when you try to download Chrome with Edge - which should be against the rules imo.
I agree with going after the Edge Lords and making things more fair…but I’m guessing Chrome is the most used we browser by a long shot even on windows so the “No platform independent browser can aspire to match Edge’s unparalleled distribution advantage on Windows." part feels like users are comfortable stepping over Edge’s corpse to download chrome anyway.
Over simplification but partly it has to do with how LLMs split language into tokens and some of those tokens are multi-letter. To us when we look for R’s we split like S - T - R - A - W - B - E - R - R - Y where each character is a token, but LLMs split it something more like STR - AW - BERRY which makes predicting the correct answer difficult without a lot of training on the specific problem. If you asked it to count how many times STR shows up in “strawberrystrawberrystrawberry” it would have a better chance.
I’d love to debate politics with you but first tell me how many r’s are in the word strawberry. (AI models are starting to get that answer correct now though)
Mods could just make a filter to remove everything new anyway. The concept of mods being unpaid volunteers means they get to fuck with reddit if they really want. They already had that issue with some subs just starting to allow porn during the first api protest. Sure reddit can just churn through to newer friendlier mods like the first time but they’re not going to be able to crush all the dissent and drama from moves like that.
But actually I think reddit has a bigger problem than protests. They tweaked their algorithm recently and it is going the way of facebook now, I’ve been getting 0 upvote day-old posts shown to me. They’re probably getting more engagement but I don’t think redditors are going to put up with that level of enshittification as easily as other social media where people are locked in by friends and followers.
Open AI has a projected revenue of 3 Billion this year.
It is currently projected to burn 8 Billion on training costs this year.
Now it needs 5 Gigawatt data centers worth over 100 Billion.
And new fabs worth 7 Trillion to supply all the chips.
I get that it’s trying to dominate a new market but that’s ludicrous. And even with everything so far they haven’t really pulled far ahead of competing models like Claude and Gemini who are also training like crazy.
They definitely do, especially legacy stuff that’s still kicking. Sweet, sweet tech debt that you hope won’t be a real problem until you’ve moved to a different company.
I’ve used word/onenote or FOSS equivalents the same way, they’re fine as a scratchpad for notes. As you said it’s nice being able to shove images in there. There are so many things that don’t belong anywhere else that I will forget after even a half hour break.
People in this thread don’t seem to understand how anti big business the FTC has been since Lina Khan was appointed. These reports are meant to be used by congress to help guide real policy. It’s one thing to just assume social media is violating privacy, it’s another thing to have a facts-based report on exactly what is currently happening.
Of course the FTC needs new laws to do any enforcement and there’s probably not enough anti corporation politicians to pass laws that give them real teeth on data privacy issues.
deleted by creator
https://youtu.be/bFM9HHB9JXI?t=714
It was a clip from this video.
The specific one I was watching was just talking about dead reckoning based on the same principle, but mapping seems like it could be a natural next step for the tech.
I watched some technical video the other day about a quantum guidance system and for a brief second I thought it might be a flat earther conspiracy because a handful of the field specific jargon in the beginning sounded made up.
I’m doing my part by writing really shitty foss projects for AI to steal and train on.
The nice thing is that a lot of environments that are secure enough to decide to airgap systems these days also make infecting the system and even moreso retrieving the emissions really difficult. Still a really cool field though and I’m glad this Dr. Guri is getting real weird with it.
type site:lemmy.world in front of your search if using google. You can combine multiple instances with the OR operator ie site:lemmy.world OR site:programming.dev this will force google to give you content only from your desired domains but lemmy.world posts will likely trample the other instances for a lot of stuff.
We’re becoming a little centralized (which I personally don’t find to be such a bad thing yet).
It’s non just Facebook either. Every big tech social media platform has headed in this direction of showing you stuff you don’t really want to see based on maximizing profit. For-profit social media seems to mostly be doomed to this outcome because it makes more money.
A company “accusing” someone of piracy isn’t proof. Access to the internet is almost essential these days. If you can prove a person is pirating prosecute them under the law with fines or even incarceration if warranted. But stripping internet access from someone shouldn’t be seen as an acceptable punishment for a free citizen anyway.
Whoever owns the network attached to the IP address also shouldn’t be responsible for actions of every user. Let’s ban an entire company, college, or government institution from the internet because an IP showed up on a list… dumb ruling.