taladar

@taladar@sh.itjust.works

• ⁨Comment⁩ on ⁨Netflix will show generative AI ads midway through streams in 2026⁩ ⁨⁨4⁩ ⁨days⁩ ago⁩:

  A good rule of thumb is that if you have heard of a brand but don’t remember anything positive about them they should probably be dead to you.

• ⁨Comment⁩ on ⁨Netflix will show generative AI ads midway through streams in 2026⁩ ⁨⁨4⁩ ⁨days⁩ ago⁩:

  And imagine if products that couldn’t get by on their own merits without ads wouldn’t exist at all. How much more productive and happy our society would be if we got rid of useless products and the negative feelings ads induce when we don’t have those useless products at the same time.

• ⁨Comment⁩ on ⁨Netflix will show generative AI ads midway through streams in 2026⁩ ⁨⁨4⁩ ⁨days⁩ ago⁩:

  Sounds a lot like gig economy for everyone.

• ⁨Comment⁩ on ⁨Algorithm based on LLMs doubles lossless data compression rates⁩ ⁨⁨4⁩ ⁨days⁩ ago⁩:

  And how do I get the prompt that will reliably generate the data from the data? Usually for compression we do not start from an already compressed version.

• ⁨Comment⁩ on ⁨Max pivots back to HBO Max as WBD rethinks ability to compete with Netflix⁩ ⁨⁨4⁩ ⁨days⁩ ago⁩:

  it ends in X and the CEO is the dumbest fuck alive? Must be one of Elon Musk’s companies?

• ⁨Comment⁩ on ⁨Audible unveils plans to use AI voices to narrate audiobooks⁩ ⁨⁨4⁩ ⁨days⁩ ago⁩:

  Seems more like a proof of concept project for that paper than something they are pursuing seriously judging by the GitHub location in some example folder that hasn’t seen any significant updates in over a year. If it is so great I would assume they would pursue it more actively and replace existing models with it two years later.

• ⁨Comment⁩ on ⁨Company Regrets Replacing All Those Pesky Human Workers With AI, Just Wants Its Humans Back⁩ ⁨⁨4⁩ ⁨days⁩ ago⁩:

  Except that your context aware search engine would tell you when there is no result and AI will just make shit up.

• ⁨Comment⁩ on ⁨Company Regrets Replacing All Those Pesky Human Workers With AI, Just Wants Its Humans Back⁩ ⁨⁨4⁩ ⁨days⁩ ago⁩:

  Probably depends on the language in the target market, a lot of European languages are not that common in countries with cheap labor.

• ⁨Comment⁩ on ⁨Audible unveils plans to use AI voices to narrate audiobooks⁩ ⁨⁨4⁩ ⁨days⁩ ago⁩:

  Most of these languages dont even have enough professional voice actors to cover the bandwidth.

  And you think anyone is training AI voice models for those languages? Have you even seen how long it takes even large companies like Google to support the languages with hundreds of millions of speakers?

• ⁨Comment⁩ on ⁨Palantir CEO Alex Karp praises Saudi engineers and takes a swipe at Europe, saying it has 'given up' on AI⁩ ⁨⁨4⁩ ⁨days⁩ ago⁩:

  This was using gdscript to make a basic character controller.

  Sure, if you make something that likely has literally dozens of copies of the exact thing in the training data it can probably do well.

• ⁨Comment⁩ on ⁨A VPN Company Canceled All Lifetime Subscriptions, Claiming It Didn’t Know About Them⁩ ⁨⁨4⁩ ⁨days⁩ ago⁩:

  I propose “the new dark ages” might be more appropriate.

• ⁨Comment⁩ on ⁨A VPN Company Canceled All Lifetime Subscriptions, Claiming It Didn’t Know About Them⁩ ⁨⁨4⁩ ⁨days⁩ ago⁩:

  Lifetime means lifetime

  No, actually that is part of the problem, they shouldn’t even be allowed to advertise ‘Lifetime’ without explicitly stating whose lifetime.

• ⁨Comment⁩ on ⁨A VPN Company Canceled All Lifetime Subscriptions, Claiming It Didn’t Know About Them⁩ ⁨⁨4⁩ ⁨days⁩ ago⁩:

  True but that is a situation that doesn’t really apply very often in the “if you hit the lottery” situation mentioned in the post you replied to.

• ⁨Comment⁩ on ⁨AI Could Be the Most Effective Tool for Dismantling Democracy Ever Invented⁩ ⁨⁨5⁩ ⁨days⁩ ago⁩:

  Half of it wouldn’t even work if the news media would do their job and filter out crap like that instead of being lazy and reporting what is going on on social media.

• ⁨Comment⁩ on ⁨Palantir CEO Alex Karp praises Saudi engineers and takes a swipe at Europe, saying it has 'given up' on AI⁩ ⁨⁨5⁩ ⁨days⁩ ago⁩:

  it can code pretty effectively

  As long as you are okay with 95% of your code not doing what you intend it to do while using libraries that don’t exist sure. Otherwise it really can’t. I am a programmer myself and have quite frankly wasted many more hours than I should have giving it the benefit of the doubt on that one and it just produces utter horseshit as code.

• ⁨Comment⁩ on ⁨Palantir CEO Alex Karp praises Saudi engineers and takes a swipe at Europe, saying it has 'given up' on AI⁩ ⁨⁨5⁩ ⁨days⁩ ago⁩:

  Mainly the purpose of those is to diffuse responsibility.

• ⁨Comment⁩ on ⁨OpenAI says its latest models outperform doctors in medical benchmark⁩ ⁨⁨5⁩ ⁨days⁩ ago⁩:

  So they created a test so broken and warped that no actual professional can understand it but their AI performs well on it?

• ⁨Comment⁩ on ⁨Palantir CEO Alex Karp praises Saudi engineers and takes a swipe at Europe, saying it has 'given up' on AI⁩ ⁨⁨5⁩ ⁨days⁩ ago⁩:

  AI is here to serve humanity, that’s where the value is, not to serve only a select few.

  Mostly right now it is here to torture humanity with low quality slop. It literally can’t do the things people claim it is capable of doing, no matter for whose benefit. It is just that some people benefit even if the output quality is significantly lower as long as they have enough people convinced that it is a viable replacement.

• ⁨Comment⁩ on ⁨Noncoders are using AI to prompt their ideas into reality. They call it 'vibe coding.'⁩ ⁨⁨5⁩ ⁨days⁩ ago⁩:

  And we naively thought Excel files with VBA scripts used as a database was the worst that could fall into that category a decade or two ago.

• ⁨Comment⁩ on ⁨Noncoders are using AI to prompt their ideas into reality. They call it 'vibe coding.'⁩ ⁨⁨5⁩ ⁨days⁩ ago⁩:

  ChatGPT Karma system that only allows limited interaction while you haven’t jumped through enough hoops?

• ⁨Comment⁩ on ⁨Noncoders are using AI to prompt their ideas into reality. They call it 'vibe coding.'⁩ ⁨⁨5⁩ ⁨days⁩ ago⁩:

  There was one just yesterday where AI printed “Listening on 127.0.0.1:1234” while actually listening on “0.0.0.0:1234”.

• ⁨Comment⁩ on ⁨Audible unveils plans to use AI voices to narrate audiobooks⁩ ⁨⁨5⁩ ⁨days⁩ ago⁩:

  Let AI pay for them and AI listen to them too. That way we can pay for and listen to actually good ones.

• ⁨Comment⁩ on ⁨Audible unveils plans to use AI voices to narrate audiobooks⁩ ⁨⁨5⁩ ⁨days⁩ ago⁩:

  Two narrators with one reading the male and one reading the female characters is usually okay but the full cast dramas are the worst.

• ⁨Comment⁩ on ⁨Audible unveils plans to use AI voices to narrate audiobooks⁩ ⁨⁨5⁩ ⁨days⁩ ago⁩:

  Sound effects, music […] improve the experience

  Actually hard disagreeing on that. I absolutely hate the audio drama versions of audio books and prefer the narrator only ones since they are much clearer and require a lot less focus to listen to and work in more contexts (background noise,…). Sound effects and music (while something is read, intro or outro style music is okay) distract from the actual content.

• ⁨Comment⁩ on ⁨Audible unveils plans to use AI voices to narrate audiobooks⁩ ⁨⁨5⁩ ⁨days⁩ ago⁩:

  It’s the publishers that hold any real power

  It might be time to finally change that, especially considering what a piss poor job they have been doing for decades at their own part of the production of media.

• ⁨Comment⁩ on ⁨Audible unveils plans to use AI voices to narrate audiobooks⁩ ⁨⁨5⁩ ⁨days⁩ ago⁩:

  So, would your opinion change if it turns out they’re going to use purchased voice rights to have a single narrator perform the whole book and then use AI to turn the narrators voice into a full voice cast?

  It would make me hate it even more because I already hate the existing full cast of humans audio dramas 99% of the time and actually prefer a single (or low number of) narrator approach.

• ⁨Comment⁩ on ⁨USDA Reverses Course, Commits to Restore Purged Climate Webpages in Response to Farmers’ Lawsuit⁩ ⁨⁨5⁩ ⁨days⁩ ago⁩:

  In debates it is sometimes referred to as gish gallop. Throw a huge number of things out there without regard for the likelihood of success with each of them to overwhelm the opponent.

• ⁨Comment⁩ on ⁨Avoiding AI is hard – but our freedom to opt out must be protected⁩ ⁨⁨6⁩ ⁨days⁩ ago⁩:

  But they literally can’t ask you for it if it is about high volumes of data that only become useful if you have all or close to all of it like statistical analysis of rare events. It would be prohibitively expensive if you had to ask hundreds of thousands of people just to figure out that there is an increase in e.g. cancer or some lung disease near coal power plants.

• ⁨Comment⁩ on ⁨The worst tech event ever: looking back at Google I/O 2024⁩ ⁨⁨6⁩ ⁨days⁩ ago⁩:

  Honestly, I don’t need innovation in every area, I would settle for existing systems getting more robust and streamlined and better standardized and compatible to each other (e.g. in the space where IoT devices all have their own app make some standard to get rid of that).

• ⁨Comment⁩ on ⁨[paper] Evidence of a social evaluation penalty for using AI⁩ ⁨⁨6⁩ ⁨days⁩ ago⁩:

  This apparent tension between AI’s documented benefits

  That is one hell of an assumption to make, that AI is actually a benefit at work, or even a documented one, especially compared to a professional in the same job doing the work themselves.