News for hackers. Stuff that matters.

SegfaultSeagull

Re:Seven books I keep close because I love them on Friday August 14, 2026 @01:40PM (#49302047)
The Hebrew is תָּמ֣וֹת נַפְשִׁי֮ עִם־פְּלִשְׁתִּים֒, which literally translates to "You shall die, my soul, with the Philistines."
Re:First Robotic Satellite Servicer Launched on Monday July 27, 2026 @05:40PM (#49075863)
Can you tell us what is out of bounds, so we know not to ask those questions? :)
Time to get a second job buddy.
Re:Claude Platform on AWS on Tuesday May 12, 2026 @12:29AM (#48104197)
"Claude in the cloud" has a nice ring to it.
Social security and Medicare account for 36% of the Federal budget and are absolutely understood as a cost to run the government. Calling them transfer payments doesn't obviate the fact that they are costs to American taxpayers. Social security and Medicare are also the two Federal outlays that economists fear will bankrupt the United States.
The cost to operate the Federal government for one day is about $18 billion.
$400 billion
Re:Databases Were Not Designed for This on Sunday April 26, 2026 @01:07PM (#47911876)
This comment is savage and I’m here for it.
Re:AI-Assisted Cognition Endangers Human Development on Wednesday April 15, 2026 @02:42PM (#47783387)
It’s a bit ironic that the author includes an AI generated audio version of the article, you know, so we don’t have to read it.
Re:Lunar Flyby on Tuesday April 07, 2026 @07:35PM (#47682710)
Can someone ELI5 how it is one side of the moon is never seen on Earth? The moon orbits and also rotates, does it not?
Re:Sand from Different Beaches in the World on Friday March 27, 2026 @02:12PM (#47546238)
This is fantastic! Excellent share. Though I object vigorously to its exclusion of any beach in California. Scandalous!
Starring GitHub repos considered harmful.
Re:Missile Defense Is NP-Complete on Tuesday March 24, 2026 @09:41AM (#47502472)
Or perhaps they will learn they are outmatched, lack the resources and technological capabilities to compete, and deterrence will have been established.
This is great and congrats on the success. Many years ago I tried starting a cybersecurity reading group in my city since the startup I was working at was small and people there weren’t interested in that topic. I got a lot of very green, aspiring and non-professionals to show up. We couldn’t really agree on where to start and people had different ideas of where to focus or even how much they wanted to contribute. Mostly people wanted to hear a summary and didn’t really put in the kind of effort that I had hoped. It didn’t last long. Congrats again on making it 5 years and covering so much ground.
Re:Shall I implement it? No on Thursday March 12, 2026 @08:23PM (#47359178)
What if, stay with me here, AI is actually a communist plot to ensorcell corporations into believing they are accelerating value creation when really they are wasting billions more in unproductive chatting which will finally destroy the billionaire capital elite class and bring about the long-awaited workers’ paradise—delivered not by revolution in the streets, but by millions of chats asking an LLM to “implement it.” Wake up sheeple!
Re:Shall I implement it? No on Thursday March 12, 2026 @08:05PM (#47359041)
What if, stay with me here, AI is actually a communist plot to ensorcell corporations into believing they are accelerating value creation when really they are wasting billions more in unproductive chatting which will finally destroy the billionaire capital elite class and bring about the long-awaited workers’ paradise—delivered not by revolution in the streets, but by millions of chats asking an LLM to “implement it.” Wake up sheeple!
This comment should be flagged for casting aspersions on a minority group.
> I don't think we have to worry about cute animal pictures taking over HN.

Challenge accepted.

Re:Open Source Iran War Cost Tracker: 45.7B on Wednesday March 04, 2026 @11:59PM (#47257753)
"Estimated US military spending since April 20, 2024, the day Congress passed $26.38B in emergency security assistance for Israel. "

In April 2024, Congress passed a $95.3 billion national-security supplemental package that included funding for Ukraine, Israel, and Indo-Pacific security. The Israel portion totaled about $26.38 billion.

Key points about that money:

Total Israel-related funding: about $26.38B. Humanitarian aid: roughly $9.15B (for civilians affected by conflicts, including Gaza). Missile defense: about $5.2B for systems like Iron Dome and David’s Sling. Weapons procurement: about $3.5B for new weapons. Other military supplies/services: about $4.4B.

https://www.reuters.com/world/us/whats-us-houses-foreign-aid... https://en.wikipedia.org/wiki/Public_Law_118-50

Re:Open Source Iran War Cost Tracker: 45.7B on Wednesday March 04, 2026 @11:52PM (#47257714)
That only works if you treat Iran’s proxies as unrelated actors.

Hamas (Oct 7), Hezbollah on the northern border, the Houthis in the Red Sea, Iraqi militias hitting U.S. bases — that’s been Iran’s strategy for decades.

April 2024 may have been the first direct exchange, but the war with Iran’s proxy system was already underway.

Re:How Trump Decided to Go to War on Monday March 02, 2026 @03:09PM (#47223396)
"The Iranians presented the Americans with a seven-page plan with proposed levels of future nuclear enrichment, numbers that alarmed Mr. Witkoff and Mr. Kushner.

The Americans still wanted the Iranians to commit to zero enrichment, and proposed giving them free nuclear fuel for a civil nuclear program, but the Iranians refused, a U.S. official said. After the talks ended, Mr. Witkoff and Mr. Kushner told Mr. Trump they did not think a deal could be reached."

That NYT detail is the key piece for me.

If Iran was presented with an offer of zero enrichment plus guaranteed civil nuclear fuel — meaning energy without weapons capability — and still refused, that tells you something.

Enrichment isn’t symbolic. It’s the hard technical pathway to weapons-grade material. You don’t insist on retaining it unless you want the option to cross the threshold.

The Tehran regime has funded proxies and asymmetric violence for decades. A nuclear-armed version of that regime isn’t just a regional problem — it fundamentally changes deterrence math across the Middle East and likely triggers proliferation.

If intervention now prevents a regime with that track record from getting nuclear weapons, that’s not escalation for its own sake. It’s preemption of a much more dangerous equilibrium.

You can debate costs and risks — but pretending the enrichment fight was benign ignores what enrichment is for.

That’s fair — Feb 24, 2022 was a massive rupture in the European security order.

The argument for this being as important (or more) is different though.

Russia launched a conventional invasion to redraw borders.

The regime in Tehran has spent nearly five decades exporting violence through proxies — Hezbollah, Hamas, militias in Iraq, Syria, Yemen — destabilizing multiple regions simultaneously while avoiding direct accountability. It’s been a long-running asymmetric war, not a single event.

If this moment marks a direct confrontation with that system, it’s not a regional flare-up. It’s potentially the first real attempt to confront a state that’s institutionalized proxy terrorism as a core strategy.

Different kind of challenge. Possibly broader consequences.

Re:The Pentagon strongarmed AI firms before Iran strikes on Monday March 02, 2026 @01:44PM (#47222217)
I get why people are nervous about AI in the military. Autonomous weapons are a scary concept. But I think the framing sometimes skips a step.

Modern militaries are already software-driven. Targeting, logistics, satellite analysis, cyber defense — it’s all code. Machine learning is just the next layer on top of systems that already exist. It’s not like we’re going from swords to Terminators overnight.

Also, opting out doesn’t mean the technology goes away. If the U.S. or other democratic countries decide “we won’t touch this,” that doesn’t slow down China, Russia, Iran, etc. It just shifts the balance toward actors who are less constrained.

There’s a real argument that better AI could reduce civilian harm. If a system can process more sensor data than a human and flag inconsistencies or uncertainty, that can make strikes more discriminate, not less. Humans under stress make mistakes too. A lot of them.

I’m much more worried about governance than the raw tech. Keep humans in the loop. Make systems auditable. Make decision chains reviewable after the fact. That’s a policy problem, not a “ban the math” problem.

The military isn’t going to stop using advanced software. The real question is whether it’s built inside systems that answer to law and oversight, or outside of them.

Older »