epolanski

> 1. Get a little bit informed. It doesn't require a PhD, it could be done instead of a fraction of the time they spend scrolling bullshit on their smartphone.

The problem is that as soon as you start digging into it, you're essentially walking in a guilt trip.

The kind of "eating an hamburger pollutes more than driving an SUV for 50 miles" guilt trip.

And you realize most of the focus should be on materialism, consumerism and how/what you eat, not just cleaner transport.

But who really wants to do so? I myself am aware that me eating a beef steak per week has a gigantic impact on environment and resources. That switching to pork > chicken > eggs in that same meal would be better.

But I just don't want to enter the guilt trip. So I do a bit, but not as much as I could.

A carbon tax should help there. Make it so me eating a steak is not incentivized. But then good luck with farmers, the easiest like and empathy grabbing category of workers on the planet.

It doesn't seem to me like LLMs write stellar Rust, in fact it seems to be quite questionable and full of unsafes.

At that point, just write it in C and benefit from the faster toolchain and the more comprehensive suite of tools.

Only if you have to read the code, something that makes increasingly less sense as it seems like our job is increasingly putting more rails and tests, and reviewing code is just economically unfeasible, especially for 99.99% of projects out there.

The more I think about it, the less I can argue against "just have the LLM implement it in C".

Especially as LLMs already write stellar C.

> It's meaningless to say that selling something is profitable if you are ignoring the cost to produce it.

Don't enter that rabbit hole, you'd be surprised at how questionable are some accounting rules and practices, even in publicly traded companies.

The second half of the post could've been a pre-commit git hook.

> When has anyone ever done that?

Germans, Britons

That doctrine is far more of a partially implemented wishful thinking than anything the polish society or foreign policy really pursued.

As a society we're nowhere near the Germans when it comes to amending our mistakes.

Most poles will blame Ukrainians for their WW2 acts, but barely recognize our own fails between ww1 and ww2.

> This would be a new Poland. Wilson called for it to include ‘territories inhabited by indisputably Polish populations’ (this would become difficult to ascertain and haunt the future of the region)

I'm Polish, out of my 8 grandfathers: 2 come from modern day central Ukraine (Vinnytsia), 2 come from modern day Belarus (Ashmyany).

It's very hard to explain to people how weak the whole concept of nation state used to be in Eastern Europe till ww2.

From Eastern Germany to the Urals, every place you would go it was a deeply cosmopolitan mix of Germans, Jews, different Slav ethnicities and depending on the regions significant Romanians, Hungarians, Turks and some other minorities.

You would cross from one village to the other and the language would change, and in minor centers neighborhood and districts changed architecture and language. Different ethnicities also generally mapped to different professions and status. And those changes didn't really look like gradients, more like dispersed bubbles.

Ethnic tensions were always present, everybody literally hated and tolerated each other with Jews generally being the common scape goat.

My grandparents coming from modern day Ukraine vividly remember the violence and terror caused by UPA, the nazi-sympathizing Ukrainian militias. My mother's grandma could not forget till the day she lived (died in 2004) what she saw in her childhood village when Ukrainians came and started a massacre that saw her losing a brother, countless cousins, uncles and friends.

Of course, Poles have been the aggressors too, just a few decades before, when Poland attacked the Soviet Union in the 1920s and had its own share of ethnic violence against Ukrainians and Belarusians.

In any case, Eastern Europe is deeply plagued and poisoned by all of this past violence. No one is able to have a mea culpa moment, no one is able to admit their own guys were bloody murders often killing thousands of people like cattle. No one can admit their ancestors crap.

The same happens in the Russo-Ukrainian conflict. The Russians see in Ukrainians the accomplice of their biggest trauma (the Great Patriotic War against Nazi Germany).

The Ukrainians see in Russia the perpetrator of their biggest trauma (the 1920s Holodomor).

No side ever says sorry, no side ever takes responsibility, no side can move on. And Eastern Europe keeps being unhealed from its traumas and carrying this in modern age.

And a man that would rather see the world burn than step out (Putin) plays on those traumas he very well understands for his own personal lust of power.

The most stable systems are parliamentary democracies.

Sure, you might be switching governments often (and that's a feature, not a bug) at times, but things change slowly.

Which is what you need in a mature rich democracy.

The last thing you want is to elect hard to boot and control individuals like in presidential and semi presidential republics.

No there are no trivial details, and you don't need a mathematics major or category theory to understand what a functor is.

It's a pair of a type constructor F(a) and a map operation a -> b to get an F(b).

The two "non-trivial" details are the fact that the map operation has to respect identity, map(id) = id, and function composition: map(g ∘ f) = map(g) ∘ map(f)

None of those concepts is any difficult, nor requires any "math major" or non-trivial details, because the are not there.

Musk treating delusional fans like an ATM as always, lovely.

There's 1000 people that preordered their Roadster in 2017 for 250k $....

And they are rumoring now 2028 production, which in Tesla style will be deep in 2030s. If it ever happens.

It's not about one query but multiple ones, sometimes chained where it definitely makes a difference.

I can't read your blog (blocked by Cloudflare, Poland, Edge with no adblocker).

You sure it's that cool?

But it's cool /s

Wouldn't that incentivize betting on unregulated and untaxed markets?

No, you don't get it, it's information discovery, it's a financial derivative, it's arbitraging opportunity.

You should allow people to bet sums on whether you're gonna get shot tomorrow, so you can have the best possible information!

Of course there's no incentives for acting badly against you, and if there are it's again, just price discovery.

No there's no complexity behind monads except laziness and anti math circle jerk.

It's really basic concepts, much easier than the analysis you do in high school.

But people would rather sit through monad as boxes or burritos cringe tutorials.

He also used Claude to get the fastest inference engine out there (Dwarfstar) for Apple hardware and the extremely efficient Array data type in Redis.

The thing that makes monads difficult is laziness and the anti math circle jerk.

You start from a functor which is a data type that admits a map operation.

Then you add of, or pure, to make an applicative functor, and then you add flatmap to make a monad.

What a bold prediction /s

Apparently it's 2/3 times faster than the Go version.

Whatever one think, the fact that there's 4 vibe coded compilers, 2 of those with 100% coverage of the official one and all being much faster than the Go version should be a wake up call for the TypeScript team.

Despite the resources and having a huge team of people that are clearly good at programming languages and compilers, something is definitely not working as it should there if vibecoders can get much faster fully compliant compilers.

I'm sure they exist, but it's not the bulk of the work.

In southern Europe (Italy, Greece) that's well beyond 10% of the average pay.

And southern Europe isn't exactly the poorest region around.

For 99% of the tasks any model really would do, even GPT 5.4 from an year ago.

It's only for long running/very difficult cases that the differences are here, and they are small.

I honestly really don't care about models with new capabilities.

What I *really* need for my clients are good enough LLMs that can offer speed and maximum privacy/zero data retention, especially in Europe.

Right now I'm stuck in the loop of data -> pseudo anonymization on privately owned hardware (very expensive) -> third party LLM -> de-anonymize (map back to original content) -> show result.

It's expensive, error prone, and all I need to do are the same things I could do (or 99% of) with some Azure/AWS/Google APIs.

But what these scalers offer is still not good enough, especially for European users. We can't risk having a US judge or agency force those people to handle our data.

Fun fact: my SO works in a bank.

Even just an issue tracker like Jira would be an insane boost of productivity.

They live in a loop of emails and chats and meetings in a never ending sync to find out what's blocked, why, etc.

There are already an insane amount of "common sense" improvements to most office jobs that are missing, let alone automation.

> Because individual investors, and especially individual investors lumped together via a fund, have all the incentives for profit maximization and no incentives for quality of work or quality of workers' lives, except insofar as it is reflected in profits.

That's up to regulators to fix.

As an investor a company's duty is towards shareholders.

Source?

By far my favorite model.

At the beginning of the month I realized I had more tokens and more capabilities on Opencode at 8$/month via DS 4.1 flash, than I had on a 200$ account in June with the american providers.

Literally both more tokens and capabilities than Opus 4.8/GPT 5.5.

Even if you pay per token, you're still unlikely to outspend american providers.