sakex

3% profit margin means you're better off selling the business and investing the cash into US bonds at current rates.

Another thing in that vein is features that were built because the previous approach was obtuse. For instance, find it harder to explain async/await in JS to people who never experienced callback hell, or why there is a borrow checker in Rust to people who never tried C++.

Too much immigration too quick, people got fed up from what I gather

I worked at google for 3 years and can confirm what you've heard. Obviously, every org is somewhat different.

While I agree we need to keep our immigration under control, this is not the solution.

Is this based on paged attention with hashing of the pages?

Incredible to see my high school mentioned here

Gaming GPUs only which are those we are all nostalgic about, but hardly the ones that matter now for Nvidia.

I also bought a laptop with an RTX 5090 (mobile) in New York last month. I paid half what I would have paid in my country. Especially when you consider how low the USD is trading at.

Levels in big tech are just a way to keep you motivated. You'll work harder to get a promo.

In the end it doesn't matter, you'll make more money by either leaving or getting a retention offer.

Exactly like Trump is doing to you.

I'd be surprised if they didn't scale it up.

There are new things being tested and yielding results monthly in modelling. We've deviated quite a bit from the original multi head attention.

Maybe add the date to the title, because it's hardly new at this point

What about MRI? Just had one. Sorry if it's a stupid question, I don't know much about this

The anomaly here are AI researchers

They won’t notice those extra 10 milliseconds you saved

Depends what you're doing. In my case I'm saving microseconds on the step time of an LLM used by hundreds of millions of people.

The point you're missing is that people were making the same kind of comments about Amazon and Uber not too long ago

I was laid off from Google in January last year alongside 150 people in my extended team. I managed to find a different team in Gemini, so now I'm part of Deepmind. I have very conflicting feelings because on one hand I really enjoy the work, the team, and the absolute genius of people I get to talk to; but on the other hand, I have some resentment for being so inhumanely laid off, I am sad for the people in my team who were not as lucky as me, and I know it can happen again any time.

I have some doubt about #2. Weren't Big Tech companies paying senior engineers $300K+ - in 2025-adjusted dollars - back in 2013?

Yes but big tech got bigger. Google had a 4th of its current workforce for instance, Meta a 10th, etc. It got much easier to get into those companies.

Such an accurate description of what I went through in Switzerland. Kids are mean, and I had to be mean to survive too. It stained my character in ways I am still trying to overcome more than a decade later.

Name and shame

That's something I wish I had when I started looking at einsums. It gets interesting when you start thinking about optimum paths (like in the opt_einsum package), sharded/distributed einsums, and ML accelerators.

I have a hard time imagining who would suffer from this more, but I have an even harder time imagining anyone would benefit from this.

Great article. I think what you should cover next is collective matmuls and sharding.

Using Jax, you should be able to get good performance out of the box

Shameless plug:

I have an implementation in Rust (CPU only) with added GRU gates

https://crates.io/crates/neat-gru

I agree to some extent, but I'd say it depends on the programming language. For instance, in Python, you absolutely need documentation because you don't know what type is expected, what different values can be passed or even the optional kwargs etc. It's also very common to pass strings as parameters when one should be using enums, which means you can't know what are the possible values without having the doc or diving into the code.

In Rust however you can get away with way less documentation because the type system and signatures are self explanatory. And if you're doing things right, the set of values you can pass to your function is limited.

Jax is really nice

There is no preemption in Javascript. It is based on cooperative multitasking[1] (your await statements and non-blocking callback) which is the opposite of preemption.

[1]https://en.wikipedia.org/wiki/Cooperative_multitasking#:~:te....