nednobbins

When I use LLMs for search, I always ask for sources and then follow up.

It's more like if played a song on Guitar Hero enough to be able to pick up a guitar and convince a guitarist that you know the song.

Code from ChatGPT (and other LLMs) doesn't usually work on the first try. You need to go fix and add code just to get it to compile. If you actually want it to do whatever your professor is asking you for, you need to understand the code well enough to edit it.

It's easy to try for yourself. You can go find some simple programming challenges online and see if you can get ChatGPT to solve a bunch of them for you without having to dive in and learn the code.

Interesting.

It looks like that's been changed since the NPR broadcast.

https://redstagfulfillment.com/universal-postal-union-treaty/

The bullshit is that anon wouldn't be fsked at all.

If anon actually used ChatGPT to generate some code, memorize it, understand it well enough to explain it to a professor, and get a 90%, congratulations, that's called "studying".

Nobody builds cars under slave like conditions. It’s just not possible. Modern car factories are highly automated plants that require skilled operators. In the case of the VW Xinjiang, that was QC inspectors. There’s no way a hole in the wall car factory using outdated labor practices can come close to competing against modern production.

It doesn’t much. The data has already been collected. All this does is make it harder for Italians from accessing their frontend.

The frontend is likely collecting additional data and we know there’s a layer of censorship built into it but that’s limited to your query strings and some metadata.

Yes.

Your response got that there is limited legal recourse, even if it’s true. The main hope is messaging but it’s a long shot.

The Deepseek-R1 paper shows us that training good LLMs can be done by anyone. That means you don’t need NVidias top of the line chips and you don’t need to pay a premium to some company that got access to those chips.

If it turns out that they lied about the hardware they used, it means that Nvidia and the big AI companies still enjoy a monopoly.

It probably won’t be that bad. One thing that China has shown is that it values practical long term strategy over ideology.

If Canada proposes an trade deal that benefits China, they are unlikely to let the past get in the way, especially since the leader behind many of the past decisions is out.

They’re pretty clear about what they’re asking the DNC to do. It’s in the first paragraph of the article.

Before the election, people were told not to protest so the DNC wouldn’t lose power. Is there any time when are people allowed to tell their leaders that the party has lost its way?

Africa has a number of factors in its favor that make it a potential economic powerhouse. It has some of the largest natural resource reserves in the world, it has a huge population, it's conveniently located on or near several important trade routes.

It's also cursed with some pretty bad natural infrastructure. The rivers in Africa don't provide good access between the center of the continent and the coasts.

China had about the same GDP a Sub-Saharan Africa in the early 1990's https://data.worldbank.org/indicator/NY.GDP.MKTP.CD?end=1996&locations=CN-ZG&most_recent_value_desc=true&start=1960 It's taken China 35 years to get from there to it's current spot as 2nd largest economy in the world. And that was for an economy that was growing at nearly 2x the rate of the rest of the world for most of that period. https://data.worldbank.org/indicator/NY.GDP.MKTP.KD.ZG?locations=CN-1W

That's not to say that Africa can't become an economic powerhouse but it will take a lot of work and time. It would take sustained investment and reinvestment in Africa over several decades.

Because the human brain doesn't intuitively count the way we're taught in school.

Our brains are very good at understanding 1, 2, sometimes 3 and, "many". That's the data we get from smart chips, young children and isolated pre-literate societies.

Counting bigger numbers requires abstract systems. Our brains can do that but it's much harder and we don't grasp it as well.

The practical offshot of this is that while it's intuitively obvious that a small space like a garage will quickly fill up with toxic gasses, it's far less intuitive that a "very big" outside can get saturated by a "pretty big number" of cars.

We make food "from scratch" on a regular basis.

We've found a few different sources for fats. I'll focus on the pork-fat ones.

The most common source is to just collect fat any time we make pork things. The advantage is that it's cheap and easy; just let it cool and add it to a jar in the fridge. The disadvantage is that it will have a lot of other flavors (especially salt).

Sometimes we just by processed lard. That's basically the opposite end of the spectrum. It's very pure and has no flavor besides the fat itself.

Often we'll wet render our own fat. Traditionally that would be the trimmings off of other cuts. Unless you're butchering a pig (or have bought into a fractional pig through something like a CSA) those bits usually aren't available. Typically we'll just buy cuts that are very high in fat. For pork, that would be pork belly. We'll just buy an uncut slab and wet-render it. Trim any meat you want to cook with (belly is the part that bacon is made of) throw the rest in a pot of water an simmer it for a few hours. The fat layer that collects on top is almost pure lard.

We've also found that duck fat is a great substitute for lard. It has a similar smoke point to lard (slightly higher). It tastes different from lard but it's also good enough that the flavor itself will improve meals. Duck breasts are about 50% fat if you buy them with skins. You can also buy duck fat on its own.