airforce1

This one is critically acclaimed and is not a sokoban: https://store.steampowered.com/app/2721890/oo/

> After reading the linked article, and the comments here I still have zero clue about the game

A video is worth a thousand words, and there's a video at the very beginning of the article. Did you watch the video?

You control a character on a grid and you have to push sausages around the grid in order to grill them (some of the floor tiles are grill tiles). That's the core game. But the sausages roll and you can't let a given side touch a grill more than once. And the grid is space constrained - you can accidentally push a sausage off the grid and it will fall into an abyss and you have to start over. The puzzles are very difficult because there is so much complexity that stacks:

- Your character can strafe and push things, but your character is also 2 tiles wide and can pivot and swing a fork (and the swing action can push things)

- sausages only roll along one axis, otherwise they slide

- sausages can be stacked into the 3rd dimension which means there's also gravity

- if a sausage falls on your character's head you can move it around and rotate it

- etc.

I don't see how this is different from current human poaching practices. i.e. It appears to be currently legal to hire an employee from company A who has been "tainted" by company A's [proprietary AI secrets/proprietary CPU architecture secrets/etc] in order to develop a competing offering for company B. i.e. It's not illegal for a human who worked at Intel for 20 years to go work for AMD even though they are certainly "tainted" with all sorts of copyrighted/proprietary knowledge that will surely leak through at AMD. Maybe patents are a first line of defense for company A, but that can't prevent adjacent solutions that aren't outright duplications and circumvent the patent.

I think there could be a market for "permissive/open models" in the future where a company specifically makes LLM models that are trained on a large corpus of public domain or permissively licensed text/code only and you can prove it by downloading the corpus yourself and reproducing the exact same model if desired. Proving that all MIT licensed code is non-infringing is probably impossible though at that point copyright law is meaningless because everyone would be in violation if you dig deep enough.

The problem with CLIs is that unless it's a super well documented CLI like `gh`, the LLM will have a hard time figuring out how to use it from `--help` alone unless it's a really simple tool. If you want to do something complex, like create a JIRA issue, you either need to put the full issue schema in `--help` so that the LLM knows how to pass an issue or else you can use MCP which bakes tool schemas into the protocol.

> and then compile the ONNX to the native format of the device.

I'm assuming you are talking about https://github.com/onnx/onnx-mlir?

In your experience, how much faster is a "compiled" onnx model vs. using an onnx runtime?

It's a big deal, just not as big a deal as misleadingly implied. "The capitol building was bombed!" (implying Washington DC) vs "The capitol building [of Alaska] was bombed!" would both be big deals, but one is a much bigger deal than the other.

A state (not federal) house representative and her husband were murdered.

A state (not federal) senator and his wife were attempted murdered, but both survived and are expected to recover.

Your comment frames it as if 2 members of federal congress were assassinated which would have been a much bigger deal. State politicians being killed is still shocking and tragic, but try to be precise in your language as to not mislead.

It was an honest mistake, especially for someone who rarely travels.

It could happen to anyone in a country where possessing lock picks is not a criminal act. For example, your sibling might get you some picks in credit card form factor one year for Christmas. You put them in your wallet and forget about them. You travel a bit within the USA and nobody cares. Then years later you travel to Japan and are whisked away to jail because of a thing you forgot about in your wallet. The Japanese don't understand why an innocent civilian would ever have such a thing; therefore you must be a nefarious criminal.

> What reason does this guy have to lie?

It might not be 100% lies, it might be "based on a true story". The temptation to embellish/frame yourself as the faultless protagonist is instinctive and there are hundreds of examples of people doing it. Narrative shifts are super common in cases where facts are initially sparse and then more come to light... we don't have the whole context.

Surprised this story has not been flagged as it's essentially political flamebait - an uncorroborated, unverifiable account from a single person trashing the current US administration and causing everyone to pile on their hot takes and equally unverifiable and possibly embellished anecdotes.

...unless you run afoul of any of their many obscure laws, even unintentionally. I had a relative travel to Japan with his family. He's into locksport (watches Lock Picking Lawyer, etc). He had some lock picking paraphernalia on his person that he forgot about since he just carries it around 24/7 without thinking about it. Long story short, they were discovered in a metal detector at some point and Japanese security whisked him away to an interrogation room. He tried to explain locksport and youtube but the Japanese police were incredulous. He spent a full day in Japanese detention (leaving his wife and kids stranded in Tokyo without him) and at one point it was looking like he might be facing more serious charges, but then luckily someone from an American military base was able to bail him out somehow.

Yeah but we currently have a $2T deficit. We could do all sorts of good for "less than 10% of the DoD's annual budget", but we are already spending far more money than we bring in. The DoD's budget needs to be reduced by like 50% (along with a lot of other departments/programs) to have even a hope of getting debt under control, let alone introducing new "think of the children" expenses

Well, years later the website is still operational and the company still seems functional after Musk cut 80+% of the staff, which to me is pretty mind blowing. I'd call that pretty successful. If I, as the end user, can't tell the difference between pre-80% and post-80% cut twitter then what value was that 80% bringing to the organization, exactly?

Then why is everyone fleeing to escape the high taxes and crime? California is losing house seats at the present.

Also, I don't know if you can really credit the left's supermajority for the success of SV... CA's politics have pivoted over the years. Look at an election map of CA in 1980 and you'll see: https://en.wikipedia.org/wiki/1980_United_States_presidentia...

If your backend is JS and it's too slow for you, then obviously porting it to a machine code binary will speed it up significantly. If you are happy with your backend performance, then does it matter?

In my experience it is pretty difficult to make WASM faster than JS unless your JS is really crappy and inefficient to begin with. LLVM-generated WASM is your best bet to surpass vanilla JS, but even then it's not a guarantee, especially when you add js interop overhead in. It sort of depends on the specific thing you are doing.

I've found that as of 2025, Go's WASM generator isn't as good as LLVM and it has been very difficult for me to even get parity with vanilla JS performance. There is supposedly a way to use a subset of go with llvm for faster wasm, but I haven't tried it (https://tinygo.org/).

I'm hoping that Microsoft might eventually use some of their wasm chops to improve GO's native wasm compiler. Their .NET wasm compiler is pretty darn good, especially if you enable AOT.

My initial thought is that it is a troll post as well. All interviewers had to say was "ok, but your program needs to be able to take input from a file and write output to a file" and OP would have been caught with his pants down.

Instead OP is a hero who bests the interviewers at every turn with his cleverness.

Doesn't free market capitalism automatically fix this though?

In the example of a car company with zero women employees, if the market doesn't want "black, angular, high-powered cars", then they will lose market share to companies that produce cars that the market does want.

And if "getting input from lots of people with various backgrounds makes a company much better and more profitable" is a true statement, then capitalism will prove it because the most diverse companies will naturally become better and more profitable than non-diverse companies.

What's immoral is changing the works of artists but then passing it off as the artist's original.

Nothing wrong with changing the works of an artist or deriving new art from existing art so long as you make it very clear it is no longer the artist's original work. That's what ticks me off so much about this.

Some people may buy 2022 versions of Roald Dahl's literature thinking it's his original work when it is not.

I guess writers need to start having sha-256 hashes of all their works chiseled into their tombstones so that alterations can be detected post-mortem.

As far as I know he never picked his bowley. https://youtu.be/qV8QKZNFxLw

I like this 3d printed one too: https://youtu.be/7hUonUE1hEY

Similar principle - separating the input and testing phases, but uses an interesting binary ball bearing approach.

I think GP is pointing out that the ability to command a following or convince people to follow in your footsteps does not a genius make.

Brenton Harrison Tarrant commands a following and influenced a lot of copycats (his most recent disciple being the perpetrator of a certain shooting in Buffalo). So... can we categorize Tarrant as a genius purely on his influence of [others] that follow him?

I think the video game "Dead Cells" was created by a co-op

So are you arguing that it's impossible for skilled artists to make bad art? You're saying a master chef never makes meals that objectively taste bad, even if they are experimenting with completely new flavors and techniques? I think some people in this thread have raised Picasso to Godhood status. In my reckoning, Picasso was a skilled artist in an ocean of other skilled artists so he was desperate to do something different to try and stand out from the crowd. So he started experimenting with new "flavors and techniques". Assuming he is human, we can safely assume some of his experimental creations fell flat and objectively turned out bad. To deny this is to uphold Picasso as a God who is incapable of producing anything but perfection.

Really good article. Succinctly articulates and categorizes types of online social interactions I've observed over the past decade. I'm guilty of occasionally engaging bad faith communications myself. I'm trying to do better. For a while now I've felt this uneasiness that the escalating social discord we've been seeing over the past ~5 years could have disastrous consequences. I'm starting to think in terms of urgency, this year-over-year escalation of social discord needs to be addressed even before climate change.

Trolling is absolutely a form of bad faith communication. However, you needn't look far on twitter to find that trolls are pretty high hanging fruit compared to ideological flamewars perpetuated by non-trolls fighting to score points for their respective ideological teams.

> News reports what happened.

selectively reports what happened, and only if it tells the right narrative and/or generates clicks, yes.

Mass shooting where 3 innocents die = clicks and correct narrative and will therefore be reported.

DUI crash where 5 innocents die = boring, wrong narrative; no report

Civilian kills police officer = boring, wrong narrative; no report

Police officer kills civilian (esp. with different color skin) = clicks and correct narrative and will therefore be reported

> Nothing brings out creativity better than a limiting environment.

This is so true as a life principle!

I once read a book about the development of laser guided bombs [1], and in it it talked about how Texas Instruments made the first viable prototype [2] on a shoe-string budget even though they were competing against a large government contractor with a huge budget.

Because of the budgetary constraints, TI couldn't afford multi-million dollar wind tunnels for testing, so they built scale models of the bomb and dropped them into swimming pools, performing mathematics transforms on the results using the drag coefficients of water vs air. They came up with a bunch of other innovations as well, such as the shuttlecock seeker head and made a superior product that vastly outperformed the high-budget contractor.

[1] https://www.amazon.com/Weapons-Choice-Development-Precision-... [2] https://en.wikipedia.org/wiki/Paveway