piinbinary

I wonder if either (a) high frequency trading firms are going to start running their algorithms on iPhones, or (b) Geekbench doesn't accurately reflect the real world performance in some way.

Assuming Geekbench is a good reflection of real world performance (in e.g. games, video encoding, encryption, decompression, compilation, databases, etc), I'm curious to know what AMD and Intel are doing so wrong that a cellphone chip could outpace their desktop chips. I don't think the A20's lithography is vastly better, nor does the chip have more area. Where's the performance difference coming from?

It's more productive to explain why the idea doesn't work, since people can learn from that.

A bit off topic, but I think I'm starting to get model fatigue. These come out 10x faster than new Javascript frameworks were coming out 10 years ago (at least new models are far easier to adopt).

This makes me want to find a side project for an excuse to give Go another try (I last used it professionally pre-generics).

I do still wish it had discriminated unions (algebraic data types) and some better error handling ergonomics.

Definitely.

But I might not read it super carefully when recipe planning at the start of the week, which can result in scheduling mistakes.

Another pet peeve (which this doesn't solve): Recipes that include "marinate overnight" or some such in the middle (which can be easy to miss), instead of telling you up front that you have to start the day before.

Some of my pet peeves about regular cookbooks are:

- Steps that actually include many different steps.

- Steps that don't repeat the quantities, making me scroll back and forth between the step and the quantity list.

The diagrams in this post very cleanly solve both.

This has some lessons that apply much more broadly than to just airports (e.g. apps that make you jump through hoops and expect you to know what 2fa or SSO mean)

> Most generative AI tools, however, are trained on copyrighted and licensed content, and their output can include content that infringes those copyrights and licenses

To some extent, it feels like the genie is out of the bottle on this. There's so much LLM-generated code out there, and I'm sure plenty of it could be argued to infringe a copyright or license (though I think the legal bar for counting as infringement is set too low), that there's no way to go back and undo it.

That said, OpenJDK might be afraid that someone will decide to make an example of them because they are a high-profile target.

The part of this that's the most interesting to me is the fact that it is worthwhile for the compiler to expend the effort looking for this optimization opportunity. I would expect (x << 2) & -4 to be a fairly rare pattern, and even then removing the & -4 only saves one or two assembly instructions.

ChatGPT tells me that it would take a very large array to detect distance with any accuracy

One day I want to build something like this, except for sound. It would be great to get a heading and distance for where a sound is coming from.

This could be both for small scale things (e.g. which part of this is squeaking?) or large scale (e.g. is that booming noise coming from the construction a few blocks away?)

I wonder if it would ever start to make sense to burn an AI model into ROM, replacing a large portion of an inference machine's RAM with ROM. (Probably not, since I'm sure those machines do dual-duty and run training when the inference workload slows down)

Is there a URL for that language? It is unfortunately a bit un-googleable!

I have this book and like it.

I also wish there was a book that guided you through the process of implementing a language with accurate garbage collection, similar to how Crafting Interpreters teaches you to implement a language. Perhaps it could start with a shadow stack + simple mark-and-sweep and then move on to stack maps + generational GC.

Prior discussion: https://news.ycombinator.com/item?id=29104047

I have a suspicion these new prices will stick around, even after the RAM shortage ends.

Speaking of which, what's the timeline of the RAM shortage ending? I have no sense for whether it is going to be (for example) 6 months or 3 years.

Only as

> the act of writing out a problem to a model still forces the same sentence-level precision described earlier

(model referring to LLM here)

but not as writing for writing's sake

Writing is as good, or even better https://www.paulgraham.com/words.html

How do the linear types here relate (or not relate) to the Borrow Checker in Rust?

This 404s for me

The README references Electroboom's implementation

It would also be fun to have a full computer algebra system (like maxima) on a phone

I bet if you took one of Taalas' cards which consumes 200 watts for 14,000 tokens/second [0] and slowed it down by a factor of 10, it would actually be quite reasonable to power by bicycle.

[0]: https://www.sdxcentral.com/news/chip-designer-taalas-bets-on...

I've been going through Nora Sandler's Writing a C Compiler book and writing a compiler in Python. I'm excited to start the chapters on optimization - those seem like the most fun algorithm problems.

I recommend the book. It certainly isn't easy (maybe 3x harder than Crafting Interpreters), but I've learned a ton (eg how to deal with operations on different sizes of types, or the trick of using pseudoregisters to avoid having to figure out registers up front).

https://github.com/jmikkola/writing-a-c-compiler-python

I'd be very curious to know what class of vulnerability these tend to be (buffer overrun, use after free, misset execute permissions?), and if, armed with that knowledge, a deterministic tool could reliably find or prevent all such vulnerabilities. Can linters find these? Perhaps fuzzing? If code was written in a more modern language, is it sill likely that these bugs would have happened?

How does the benchmark tell how many branches were mispredicted? Is that something the processor exposes?

> a luxury apartment building goes up, surveys the market, and sets its rents 30% higher for the privilege of living in a new building with a gym for dogs or ball pit or whatever. Then the older buildings say, "Well, we can raise our rents 20% and still be the best deal in town," and so on.

I think that might not be the right cause and effect relationship. The actual cause is increased demand. This creates both the increased pricing of existing stock and an incentive to build new stock.

Back when I lived in SF, there was one bus route (the 6, I believe) that I could use to get to work. The bus was so slow due to frequent, long stops and traffic lights that I could keep up with it on foot by walking briskly. I only bothered taking it when it was raining because it didn't get me to work any faster than walking.

That's an interesting point.

I'm also curious how bus stops interact with timed lights. Presumably each time the bus stops, it gets kicked back to the next cycle of green lights (which might be a low-single-digit minute delay).

Hopefully there's a traffic engineer in the audience who can give the real answers.