HN user

avn2109

3,088 karma

Grad student and engineer. Recreational programmer. AleksNavratil.com

Posts4
Comments964
View on HN

Key point though is that Kalshi and similar prediction market betting apps are parimutuel, unlike traditional sportsbooks/Draftkings etc where punters bet against the house. This greatly changes (improves) the incentives for the house, e.g. they don't ban sharks and search only for dumb money punters, they try much less to prevent you from leaving with your winnings etc. The parimutuel house just takes a flat rake percentage off the top of the sum of all bids, so they just want to maximize the amount of money wagered overall instead of doing shady shenanigans.

I'm not a prediction market user/booster nor a sports gambler, but this is a key difference and a strict improvement from traditional sportsbook apps. If our society is going to have widespread legal gambling, we strongly want it to be parimutuel.

OSX's command-space spotlight search has been degrading functionally (at least on my machines) for literally years now. It peaked around ~2012, when it would reliably search the full text of all documents on my local hard drive quickly, and not do anything dumb like "search the internet by sending whatever I typed up into the search field to the cloud."

Nowadays it fails to reliably search the full text of documents on my local hard drive, tries to search the internet despite my best efforts to prevent this, and often even fails to find a file ~/Documents/foo.txt when I explicitly search for the string foo.txt. This is uniformly true on several work Macbooks and a couple personal macbooks too.

A truly astounding regression in functionality!

There is a new generation of bike locks which have the shackle wrapped in a composite coating that mostly destroys angle grinder cutting discs and similar cutting tools. This is likely the best current outdoor bike locking approach as it requires thieves to have either 1) lots of spare discs or 2) a torch or 3) picking skills, and very few thieves have any of those.

Study mode 12 months ago

LLM's are useful because they can recommend several famous/well-known books (or even chapters of books) that are relevant to a particular topic. Then you can also use the LLM to illuminate the inevitable points of confusion and shortcomings in those books while you're reading and synthesizing them.

Pre-LLM, even finding the ~5 textbooks with ~3 chapters each that decently covered the material I want was itself a nontrivial problem. Now that problem is greatly eased.

If you have any evidence for the claim that "many parts of DoD" are being exploded/defunded, it would be really interesting to see that. As far as I can see, just the opposite is true; the military industrial complex looks like it's increasing in size and scope.

> ... spends the normalized equivalent of America’s defense spending...

I'd be interested in seeing the numbers for that claim broken down if you can cite them. From napkin math it seems hard to make the budgets line up, unless we're doing a very large purchasing power parity adjustment?

Controversial take: On a population level, the negatives of smoking are at least partially mitigated by the "smoking suppresses obesity" side effect. Obesity is such a huge (pun not intended) public health issue in modern American society that any intervention reducing it is actually quite valuable.

It's at least arguable and perhaps even true that we as a society would be better off if everybody used tobacco at 1950's rates and therefore got skinnier, especially if we all just dipped Zyns or similar smokeless "low harm" nicotine formats instead of smoking.

That said, I personally use 0 tobacco/nicotine etc.

"... slapping a coat of culture war paint on it - and then selling it as a new scandal..."

Astounding level of misdirection/cope here, bordering on non-factual. Did we just read the same article? This is the textbook example of a DEI scandal and was so from the very beginning. I mean the "textbook" part literally, employment discrimination law textbooks will dedicate whole chapters to this scandal for decades at a minimum.

The million dollar salary thing is compelling. I would certainly switch careers from ML engineering for a million bucks of cash comp, especially in a low CoL location :) Also, the "30 years old" thing mentioned in the GP seems excessive, surely if they were really desperate to staff up, they could loosen that age limit.

> "...it never actually happened?"

The premise of the article is not at all that cargo cults never happened. Instead, the article acknowledges that cargo cults happened, but claims they are misunderstood and often ill-documented, and therefore unsuitable for software metaphors.

The "never happened" claim is waaaay stronger and likely no credible or serious observer would make the "never happened" claim.

By far the most interesting aspect of this tale is the sudden rush of evidence (news stories, internet videos etc) that insurers have abruptly started paying claims which were previously denied or mired in paperwork. The comments of e.g. [0] and [1] are filled with people claiming this. The system is clearly rattled by the appearance of the people's justice, in a "Let them eat cake" kinda way, and is trying to make a sudden show of not being so rapacious and unjust.

[0] mhttps://www.tiktok.com/@theredcoquette/video/744542308485049...

[1] https://www.tiktok.com/@joysparkleshine/video/74460838068822...

[2] https://www.wfsb.com/2024/12/05/anthem-backtracks-anesthesio...

Surely there is somebody with enough brains in the Pentagon to diversify their supplier base away from the moribund Boeing-esque incumbents, right? If they haven't learnt this lesson post-SpaceX, when are they going to figure it out?

A couple billion bucks is pocket change to the DoD, they literally "lose" it in their couch cushions, and it could eventually be the difference between "viable domestic defense aerospace industry" and "buy threaded steel nuts for $9000 each, with an 18 month lead time and 5000 pages of paperwork."