Dude also used a LLM to generate descriptions for the packages he's serving from his package manager. And of course, it got them wrong, creating a headache for the actual package maintainers
The people who make up the 2% market share using Firefox overlaps with the people tired of AI - specifically LLMs - being shoved into everything.
Mozilla just released a LLM-driven website generator a few months ago. Why are we assuming they won't add something similar to Firefox?
If Mozilla wants to use machine learning, awesome. But how about we treat the techies who support them like adults and say "machine learning" instead of using the AI buzzword which is overloaded.
I doubt most people know that country TLDs are different from vanity TLDs. I know when I look up domains, they're usually all smooshed together and then the terms are in a giant block of ToS.
I want the creators I watch on YouTube to continue to get paid, both from YouTube and their sponsors. My contributions through premium are sliver of what they see, but if everyone stopped supporting them in that way, the total would be zero.
I back some of them on Patreon where I can, but it's not economically feasible for me to back them all in such a way.
Yeah, as someone who gets good mileage with YouTube Premium, I wish they offered a version just for ads without music. YouTube music sucks hard, but I can't justify paying for another music app on top of it.
The fact that Google started as a search company and yet search in their own apps sucks is boggling.
In YouTube Music, when you're building a tuner to create a station, you can't search at all. Instead, you get an endless scroll off bands and have to find the one you want that way. The order is random.
Like .. Pandora let you do the same thing with search back in the 00's
I agree with you, and normally like posts to just use the original headline, but in this case, it feels like the newspaper is pro-KOSA. They mention in the article hundreds of organizations support the legislation but don't mention the hundreds that have opposed it.
Perhaps something like "Microsoft president endorses online child safety bill night before Big Tech hearing [Bill is opposed by EFF over censorship concerns]" would be a better way to handle adding context.
At Giant, I'm pretty sure it's decided by the system based on some algorithm, not the employee. The one time I was audited, we were in the store for a long time and had removed a few items from the cart after adding them.
The audit consisted of the employee scanning ten random items and confirming we had scanned them too.
Yeah. I use DNS-level blocking too, but it's not something I can roll out to my non-technical family members. They understand turning off the browser extension if things don't work, but not adding a DNS whitelist and then waiting / clearing DNS.
Reddit expects to finish this year with ad revenue ... slightly over $800 million... Reddit had said two years ago it aimed to exceed $1 billion in ad revenue by 2023...
So they missed their two year goal by 20%. They had forecasted a 2.9x growth and achieved 2.3x
When it comes time to IPO, they'll just blame the economy and ad blockers, while showing how many users they forced into their app where ad blocking is harder.
Dude also used a LLM to generate descriptions for the packages he's serving from his package manager. And of course, it got them wrong, creating a headache for the actual package maintainers