vovavili

>Pandas wins if you want to use what the vast majority of folks are using

Vast majority of skilled developers are now using Polars, unless they are constrained by lack of Narwhals support in their third-party library of choice (e.g. Great Expectations, SHAP). That's the more important trend to follow.

I was actually just recently looking if there is a self-hostable alternative to a HTAP system like TiDB + TiFlash with a Postgres-compatible syntax, but LLMs didn't really pick up on CedarDB yet. These guys know game.

There is a whole chapter in Designing Data Intensive Applications dedicated specifically to handling atomic transaction concurrency issues while avoiding the performance trade-off you're describing here, you will definitely find it interesting.

https://www.oreilly.com/library/view/designing-data-intensiv...

I honestly have no idea what you're trying to say.

>People don’t like the idea of their local elected officials bypassing the rules in the dark.

Others don't like technological progress to be reduced to the speed of bureaucracies who have no compelling incentive to be particularly efficient. Especially when the technology at hand might end up making quite a few of these bureaucracies much less staffed.

>Turns out meta's reputation among US consumers is sparkling compared to their overseas reputation...

I'd argue that this is more of a testament of how US regulatory regime facilitates consumer choice compared to paternalism of other major markets. I personally can't wait to have a free AI assistant like Muse available in my country.

I don't assume that. The supply of men on dating apps is greater than the supply of women, and they tend to be less picky and more open to short-term mate-seeking behavior. This means that the dynamic of these apps is more fundamentally shaped by aggregate preferences of women.

This app in particular attempts to overcome it by using a very clever sorting algorithm - https://en.wikipedia.org/wiki/Gale%E2%80%93Shapley_algorithm

Commercial dating apps are a market for lemons - as a woman, you can't quickly assess whether a partner will be compatible with you or not without investing substantial time and effort. Hypergamy is the expected equilibrium - if you're not sure about the long-term compatibility, you might as well go off of looks or cues of status alone. I would expect the outcome distribution of an app with a much better matching algorithm to be drastically less concentrated in the upper tail. Moreover, a state-run dating app lacks the perverse incentive to keep you swiping, frustrated and jealous.

As long as that government is benevolent enough to avoid demographic engineering. Singapore is a fine melting pot, but I'd probably avoid using a closed source government-run dating app in both my home (EEU) and my current (WEU) country.

Let's put it this way.

Singapore started from an unusually favorable, but still genuinely troubled, Asian starting position. Their government then proceeded to get a remarkable number of important things right, and turned it into one of the richest countries in the world. That achievement does not require pretending Singapore was a fishing village in a swamp in 1965. The more accurate story is impressive enough on its own: plenty of countries and cities with comparable or better initial advantages failed to achieve anything close to Singapore's subsequent trajectory. Look at how Western former great powers are doing, or even closer countries like Japan.

I'd argue the way Singapore uses state capacity deserves substantial credit for that.

It has achieved the highest global GDP per capita is just a couple dozen of years, starting as a small bit of swamp forcibly kicked out Malaysia. I think it's unfair to say that people are over-romanticizing these sorts of accomplishments.

Whoever governs Singapore must have that iron in him.

It's also known as being one of the very few countries actually capable of using state capacity to overcome serious social and economic problems. Godspeed.

What exactly makes Carbon absurd?

Sounds like a remarkably unempathetic perspective. Stay stubborn.

I am a young man who lives in Europe and I too cannot comprehend the alleged second and third order benefits of a "real" social system.

I actually greatly appreciate that I can put something cute on my sister's PC that comes from a developer that won't bundle it with malware. Seems like they fail miserably at being evil.

Why are you on a YCombinator subdomain if this represents your beliefs?

It's open-source so you can just compile it yourself.

Not phoning home, vertical tabs and peak new tab aesthetic is what got me in.

You might like Brave Origin then. If hidable crypto integration is what keeps Brave free, I am personally content with it.

It's more of a statement about excessive alarmism.

A consequence of a totally self-imposed handicap.

Nothing will ever top gwern.net

>though it does at least seem that the dividing lines won't be red vs blue this time.

You'd think. Every single person unreasonably skeptical of AI that I know of usually comes from blue echo chambers.

I would argue that elaborating on scattered bits and pieces found here and there throughout the Internet is enormously valuable. Fair attribution is the tricky part.

Nebius hit me with two LeetCodes, communicated a third round of system design and instead did a third round of LeetCode. It's very much a thing for companies worth applying for, especially if internal communication gets messy.

A lot of companies who claim to want highly intelligent and autonomous engineers by revealed preference actually want glorified slightly above average ticket pushers.

At least by the looks of it, Jane Street appears to be an odd one to genuinely value competence.

The theory behind it certainly is unreasonably strong - https://www.astralcodexten.com/p/prediction-market-faq

It seems trivially fixable if you RLHF an LLM to always count characters deterministically with:

  sum(1 for c in word if c == "r")

I wonder why haven't major labs done this yet.