This will be downvoted heavy, so just to remind Russia

Im curious if there is a deep need for entire codebase to be consumed in the first place?

It would be better to have the architecture support a more decoupled/modular design if you're going to rely heavy on LLMs.

That or let it consume high quality maintained documentation?

> Maybe it’s just more visible now but it seems like these companies are really accelerating in their evil lately.

Well I mean they have no incentive to behave any other way, if anything they are rewarded via the shareholders. Number goes up when they perform mass layoffs. That tells you everything you need to know.

Not particularly 'educational' at all, but "My dad wrote a Porno". Was recommended it by a friend and have been wetting myself laughing on the work commute.

Also "Stuff you should know" is a super popular one that always gets a listen.

Thank you for that insight and for sharing your opinion on a topic which is not for you.

One of the clearest instances of Bean Soup theory I've seen.

Maybe that's the name of the podcast. Wouldn't be a bad name for a podcast about books.

PFOA was used previously, but now it's changed to PTFE, this is supposedly less harmful, but its considered perfectly fine (afaik) as long as the temperatures dont exceed 260°C.. Which at that stage you're probably using the wrong tool.. Maybe some Iron cast skillet or something ?

PFOA is a surfactant that was previously used for emulsion polymerization of PTFE. It's not found in the finished product. It's since been replaced with other less well studied PFAS surfactants that might be less dangerous or might not, which seems to me an example of bad regulation. It would make more sense to regulate disposal of PFOA, which can be done safely with processes such as supercritical water oxidation.

PTFE itself is about as inert as you can get, assuming it's not overheated. If you use PTFE cookwear, I recommend getting an IR thermometer so you can learn how your cooking setup responds and control the temperature properly.

This is some next level spam posting. Not sure to be annoyed or impressed.

It's being delayed by ai companies from running on local consumer grade machines specifically by making the cost of entry too expensive. OpenAi buys 40% of wafers to ensure the price of memory stays high.

Hmm, never considered a targeted squeeze at consumer run models by way of slowing hardware proliferation. It "made sense" to try and box out other AI companies but I guess they also have a pretty strong vested interest in keeping VRAM low or preventing some kind of high-memory PCIe ASIC from getting cheap broad adoption.

Another thread suggested that OpenAIs primary play is to get big enough that it's too big to fail, funny to think that it's not a funding runway or algorithmic moat, just a hardware vault and the longer you can stop boats crossing it the more chance you get your fingers in all the pies.

> the U.S. strongly supports cross-border data flows that promote growth and innovation while protecting privacy, safety, and free expression

Yeah that will be a hard no from me. They're not exactly known for their positive attitude towards privacy. And free speech seems to depend on who's aligned with the administration.

"Does this mean I won't need to pay a skilled person?"