<10% of natural gas plants recover helium. All of them extract it. The remaining >90% vent it into the atmosphere. This is an engineering / money problem, not a physics problem.
It is a bit scary how people seem to genuinely be OK with violence (see this reddit thread [0]). Is just me or does it feel like the overall "temperature" has gone up.
We should at least ban the "preemptive" pardon if not all pardons. Pardon means forgiveness for a specific convicted crime, not a means to grant blanket immunity.
It's a marketing strategy. If it's almost certainly conscious and capable of ending the world if it desired (even if it isn't), imagine how good it could be at building your dream SaaS!
From 20y experience and CS degree, I see software engineering as a constant struggle against accidental complexity. Like quicksand, every movement you make will pull you deeper, even swimming in the right direction. And like entropy, it governs all things (there are no subfields that are free of complexity). It even seems impossible to give a meaningful, useful definition, perhaps by necessity. All is dark.
But now and then, something beautiful happens. Something that used to be dreadful, becomes "solved". Not in the mathematical strict sense, but some abstraction or some tool eliminates an entire class of issues, and once you know it you can barely imagine living without it. That's why I keep coming back to it, I think.
As a species, I think we are in the infancy stages of software engineering, and perhaps CS as well. There's still lots of opportunity to find better abstractions, big & small.
> People like to freak out about this, so I wanted to post it here to make sure that everyone who wants to freak out about it gets the opportunity to do so.
I've grown to appreciate unapologetic trolling of people who care way too much about what other people do to themselves or their own private property.
Wouldn't it be more correct to call the article "critical" and not "incendiary"? I looked it over and I don't remember seeing any calls to violence. Altman needs to remember that he holds an incredible amount of power in this moment. He and other current AI tech leaders are effectively sitting on the equivalent of a technological nuclear bomb. Anyone in their right mind would find that threatening.
This is exactly the point of part one of Fist Stick Knife Gun: A Personal History of Violence, by Geoffrey Canada. Unequal or lack of access to the executive branch of government will create a culture of vigilantism and lends itself to organized crime as a replacement for the policing arm of the state.
People become okay with vigilante justice when they see the executive branch as compromised, just look at the insane plot/ending of the film Singham.
Many people see this happening in the US. We should expect to see more vigilante justice and organized crime if we see the executive branch as having a significant principal-agent problem.
If you enjoyed this, you might like Mind Chess, which can be played without a board and pieces [1]:
Consider Mind Chess. Two players face each other. One says "Check." The other says "Check." The first says "Check." This continues until one of them says, instead, "Checkmate." That player wins -- superficially. In fact, the challenge is to put off checkmate for as long as possible, while still winning. This may be better stated: you truly win Mind Chess if you call "Checkmate" just before your opponent was about to.
Signal developer here. It's just because notification reliability is always a top support complaint, and a lot of people turn off notifications and don't realize they've done so. Admittedly, once a month is likely too aggressive.
I just did this to my MacBook not because of the sharp edge but because the pitting turns a sharp edge into a sawblade. Something about the grounding on on the frame when plugged in mixed with my sweaty hands leads to damage along this sharp edge on every MacBook I've ever owned.
I think I’m probably being dumb, but the gotcha here seems to be - ‘if I give an application permission to access a folder, it has access to the files in that folder’ - which is what I would expect??
This is from the first of the caveats that they list:
> Scoped context: Our tests gave models the vulnerable function directly, often with contextual hints (e.g., "consider wraparound behavior"). A real autonomous discovery pipeline starts from a full codebase with no hints. The models' performance here is an upper bound on what they'd achieve in a fully autonomous scan. That said, a well-designed scaffold naturally produces this kind of scoped context through its targeting and iterative prompting stages, which is exactly what both AISLE's and Anthropic's systems do.
That's why their point is what the subheadline says, that the moat is the system, not the model.
Everybody so far here seems to be misunderstanding the point they are making.
It's definitely not obvious, given that many, many gregarious species may certainly have inter-group clashes and skirmishes at territory boundaries but no full-scale war. Animals in general avoid violence between conspecifics, for the obvious reason that it's rarely worth the risk of being hurt unless you're very sure you're going to win. Dying for your group is something you almost never see outside eusocial species. Never mind dying in your prime reproductive age!
Yes, you need to read more carefully. In particular:
“8. Confirm that Documents access for Insent is still disabled in Files & Folders.
“9. Whatever you do now, the app retains full access to Documents, no matter what is shown or set in Files & Folders.”
[…]
“Access restrictions shown in Privacy & Security settings, specifically those to protected locations in Files & Folders, aren’t an accurate or trustworthy reflection of those that are actually applied. It’s possible for an app to have unrestricted access to one or more protected folders while its listing in Files & Folders shows it being blocked from access, or for it to have no entry at all in that list.”
As an American I feel like I've been going through a bit of an identity crisis from what I remember growing up.
Probably the rose tinted glasses of being a child but being from Florida I always had a sense of amazement and wonder as I heard the sonic boom of the shuttle returning to earth.
Really felt like I was coexisting in this incredible scientific powerhouse of a country full of bright and enabled peoples that knew how to prioritize curiosity and innovation.
Feeling like a bit of a "vibe" post which is everything wrong lately but I can't help but feel some satisfaction that we're still able to accomplish something like this in our space endeavors.
A poor attempt at joining the convo too late because I don't browse /new like everyone else. No one upvotes, and I question my intelligence for the 3rd time today.
I've been using the same version of Jekyll, using the same outdated, discontinued version of ruby, for more than 10y. I refuse to learn anything about ruby, or spend any time upgrading Jekyll or any of the 2 plug-ins I use, and I take a weird pride in that. It works, it generates my blog, I don't want it to do anything else. I have no idea how it works anymore. For all I know Jekyll has been abandoned. That version of ruby might be riddled with bugs and security holes, and why would I care? it's only used when I generate the website, in a docker container that doesn't talk to anything.
Eleventy might not receive new features, your website will still work.
"I could have done it better, it's not a big deal, oh, they had women and non white people on board, what even is the shareholder value of this mission, oh it was almost done 50 years ago..."
These people went literally to the moon and back. Furthest anyone has ever been. That's an achievement.
I know things suck right now. Even more reasons to appreciate what is possible with technology.
I agree with the premise of this article. This achievement is inspiring and re-assuring that competency brings results. The alternative is way too depressing AND it mostly is our reality right know.
While scary, information like this has been pretty accessible for 20-30 years now.
In the wild west days of the early internet, there were whole forums devoted to "stuff the government doesn't want you to know" (Temple Of The Screaming Electron, anyone?).
I suppose the friction is scariest part, every year the IQ required to end the world drops by a point, but motivated and mildly intelligent people have been able to get this info for a long time now. Execution though has still steadily required experts.
Many folks on HN are the exact sorts of people who have lived the thankless popular-enough-to-be-an-unpaid-job solo OSS maintainer dream, so I wonder if you feel as annoyed by the tone of this post as I do.
I truly don't understand how the same folks that champion accessibility and humane ideals while humble bragging about working for $5/hour to help get local businesses online can throw so much shade on people who are urgently trying to figure out a way to get paid, often just to keep the projects that they created alive so that these people can continue to use them for free.
I don't know if it's entitlement, projection or just wanting to have it both ways, but I wish they would channel their frustrations into helping to find a sustainable model for OSS creators to make a living wage to keep the magic coming instead of being shitty about people doing their best to find a forever home before their burnout finally kicks in.
I work at OpenAI and I really don't find this to be the case.
We're pretty diligent about applying search blocklists, closing hacking loopholes, and reading model outputs to catch unanticipated hacks. If we wanted to, we could choose to close our eyes and plug our ears and report higher scores for Terminal-bench, SWE-bench, etc. that technically comply with the reference implementation but aren't aligned with real value delivered to users, but we don't do this. My impression is that Anthropic and other labs are similar. E.g., in the Sonnet 4.6 system card they use a model to detect potential contamination and manually score those outputs as 0 if human review agrees there was contamination. If all the labs cared about was marketing material, it would be quite easy not to do this extra work.
There are ton of other games you can play with evals too (e.g., test 100 different model checkpoints or run secret prompt optimization to steer away from failing behaviors), but by and large what I've seen inside OpenAI is trustworthy.
I won't say everything is 100% guaranteed bulletproof, as we could always hire 100 more SWEs to improve hack detection systems and manually read outputs. Mistakes do happen, in both directions. Plus there's always going to be a bit of unavoidable multiple model testing bias that's hard to precisely adjust for. Also, there are legitimate gray areas like what to do if your model asks genuinely useful clarifying questions that the original reference implementation scores as 0s, despite there being no instruction that clarifying questions are forbidden. Like, if you tell a model not to ask clarifying questions is that cheating or is that patching the eval to better align it with user value?
Are you American? Because if you are from the country that dominated the world since WWII it feels different than being from the rest of the world.
Bretton Woods gave the Americans an "exorbitant privilege" that basically meant the US could live extracting wealth continuously from the rest of the world.
Then later the petrodollar system was established. People needed oil, the US would protect the Arabs with its immense army (financed with the dollar system) and in return the oil had to be sold in dollars, so all the world needed dollars if they wanted energy.
The US could just print dollars, and the rest of the world would suffer inflation.
That was great for the US for sure. Why not continue? Because the rest of the world do not want to continue supporting the US system.
The US was ok with Sadam using chemical weapons against Kurdish civilians until he decided to change the currency for paying the oil to euros.
The US does not want to de-escalate if that means the world stops buying US bonds and suddenly they are bankrupt and can not pay its debts exporting inflation to the rest of the world.
If Americans suddenly lose 50 to 70% of purchasing power then there will be war inside the US, not outside.
"Incompetence" of this degree is malice. It is actively malicious to create a system that automatically locks people out of their accounts with absolutely no possibility for human review or recourse short of getting traction in the media. "No sir, I didn't grind those orphans up. It was this orphan grinding machine I made that did it, teehee!"
> During in-house testing, which involved taking an iPhone 16 from iOS 18.5 to iOS 26.4.1, The Register found that Apple has kept the háček in the Czech keyboard, but removed the ability to use it in a custom alphanumeric passcode. The OS will not allow users to input the háček as a character. The key's animation triggers, as does the keyboard's key-tap sound, but the character is not entered into the string.
Sounds more like an actual bug than a decision to change the keyboard layout, if this happens only in the passcode screen?
after Apple removed a character from its Czech keyboard
I wonder what the thought process (or perhaps lack thereof) at Apple was. Did no one of the likely-somewhat-large team who did that think "wait, this could lock out our users who may have used that character"?
In the immortal words of Linus Torvalds: "WE DO NOT BREAK USERSPACE!"
Now one of the ways in might be those companies who claim to be able to break iPhone security for law enforcement and the like, but I'm not sure if they'd be willing to do it (at any price) unless you could somehow trick them into thinking you had some "interesting" data on there...
This is kind-of neat too, at least in the near term:
> In the coming weeks, we will relicense all of our source-available tools, including Tart, Vetu and Orchard under a more permissive license. We have also stopped charging licensing fees for them.
