azeemba

Why weren't crypto gpus running at 100%?

A single job posting for an engagement team doesn't really tell you if its a company strategy or not.

Directors hire on random ideas all the time and then pivot as needed

Crawler, AI or not, cannot afford to pay per visit. The entire model of crawling works because the incremental cost of each crawl is so low. Even fractions of a penny would be prohibitive.

But since we have the dissent opinions, it's not tricky at all to see why they are dissenting.

Alito and Thomas straight up believe that the constitution does not provide birthright citizenship and the executive order is valid. Gorsuch mostly agrees but makes an exception if the parents plan to stay in the US. Kavanaugh agrees that birthright citizenship is not provided by the constitution, instead he argues its a federal statute that congress can overturn (but the president cannot)

https://www.scotusblog.com/2026/06/supreme-court-strikes-dow...

Amusingly, these things still happen. JP Morgan found that their nickel in a warehouse was just painted rocks. This happened in 2023: https://www.businessinsider.com/jpmorgan-bought-nickel-recei...

The author of the article is named Lawless. Is that an inverted nominative determinism? https://en.wikipedia.org/wiki/Nominative_determinism

It actually just looks at your localization settings

Matlab and Mathematica are VERY different.

Matlab and Python are in the same ballpark. Easy syntax and large standard library. Matlab provides a lot more dedicated libraries for niche areas but the overall experience feels the same.

Mathematica doesn't really have a standard counterpart. Jupyter notebooks try to capture the experience but the support for symbolic expressions makes the Mathematica experience very different.

Just to remind everyone though, Microsoft won that lawsuit on appeal.

So the history here is that Microsoft lost its monopoly on its own poor decision making.

The funny thing is that an oxygen-rich environment is a hell-hole! Oxygen is insanely reactive and will corrode anything. Even early life on earth found oxygen toxic. It was released as a waste product by early life and they were so successful that all that oxygen accumulated resulting in the Great Oxidation Event (https://en.wikipedia.org/wiki/Great_Oxidation_Event).

That likely resulted in many species going extinct!

I think a large enough org that needs many different certificates should have an internally-trusted CA. That would then allow the org to decide their own policy for all their internal facing certificates.

Then you only have to follow the stricter rules for only the public facing certs.

The author is suggesting that websites care more about server side issues than client side issues. To the point that they don't realize that users stop using them.

I think that statement is way too strong and obviously not true of businesses. It might be true if hobbyist websites where the creator is personally more interested on the server side but it's definitely not true of professional websites.

Professional websites that have enough of a budget to care about the server side will absolutely care about the client side and will track usage. If 10% fewer people used the website, the analytics would show that and there would be a fire drill.

What I can agree with on the author is more of a nuanced point. Client side problems are a lot harder and have a very long tail due to unique client configurations (OS, browser, extensions, physical hardware). So with thousands of combinations, you end up with some wild and rare issues. It becomes hard to chase all of them down and some you just have to ignore.

This can lead to it feeling like websites don't care about client side but it just shows client side is hard.

I honestly love reading them. The fact that I got tagged means something went awry and it's fun seeing HOW that happened. Who misunderstood what or who is missing what context.

Then I can make bunch of people's work life easier by clarifying the state for everyone

The main point is that the conditional didn't actually introduce a branch.

Showing the other generated version would only show that it's longer. It is not expected to have a branch either. So I don't think it would have added much value

They are different court cases, usually popularized by techemails (https://www.techemails.com/)

The ICP-Brazil thing was just a surprise coincidence as I was finishing up the article. I had started collecting my notes on the topic when Entrust got distrusted. Fun threads to read through:

- https://bugzilla.mozilla.org/show_bug.cgi?id=1890898

- https://groups.google.com/a/mozilla.org/g/dev-security-polic...

DigiCert incident was pretty loud so that got my attention too: https://bugzilla.mozilla.org/show_bug.cgi?id=1910322

The animation looks awesome! Looks like the author used matplotlib (as they mention in a comment on the website): https://github.com/vankessel/sandbox/blob/master/graph/inter...

In the past, I have used manim to make mathematical animations: https://www.manim.community/ Manim is more flexible but that comes with some overhead of complexity and learning. Example of some animations using manim:

- List of videos using manim: https://www.manim.community/awesome/

- A blog post I made: https://azeemba.com/posts/degenerate-matter.html

NASA images was the first thing I thought of when I thought of public domain images. Looks like public.work doesn't include NASA images but this website does.

I can't tell if this is sarcasm or not but AI is next to useless in 3D graphics. Some studios are experimenting with using it for iterating on concept art. I can even believe it if some studios are using it for some background low priority textures.

But for actual production models, AI models do not produce usable model right now. Their topology is so bad that they could not be animated. Even if you just need a static model, the topology is so bad that it would not support making minor fixes to the model.

I think they might be making a joke about how JavaScript can act surprisingly when `+` operator is used with strings/arrays in combination with numbers

The article mentioned possibly being animal trap, hiding yourself or spiritual.

But honestly seem like just guesses.

This is cool, definitely new to me. I do wonder if this just moves the problem though. As the article explains:

One crucial requirement is that the counter must be initialized before the first reference to the flexible-array member. Another requirement is that the array must always contain at least as many elements as indicated by the counter.

This means as the array is initialized and updated, you must keep the counter updated with it. Its just as easy to screw that up than it is to screw up the general problem of accessing elements that are out-of-bounds

Consider posting this as a new post! It seems like a fun list to read through

Do you have an example?

I feel like the post doesn't relate to the title?

The post seems to be about changing the Leaderboard and doesn't comment too much about whether the actual real-life performance of LLMs is plateauing and what can be done about it.

I am having trouble understanding what the complaint is here.

The docs still mention bigger models with 128k tokens and smaller models with 8k tokens. It seems reasonable to optimize for big and small use cases differently? I don't see how we are being "robbed".

The poster was joking that it looks like there are less than 126 Firefox users

The problem here was neither media extensions nor embedded JavaScript though.

It was pdf.js handling of fonts

https://immich.app/ has built in auto-detection and tagging.

I am surprised by the over-engineered comment. Object auto-detection is kind of an advanced feature and I would expect a lot of extra infra required for it. Immich hides it because they just publish docker containers for you but I imagine it would feel more complex if you were to look at the internals.

The top answer highlights a lot of different things but spaced repetition does seem to get quite a bit of emphasis. I first learned of it from my buddy's blog post on how he uses anki cards to do space repetition for basically everything: https://www.petemillspaugh.com/anki

I tried to use them but haven't really stuck with it.