HN Best Comments
前往频道在 Telegram
Comments from https://news.ycombinator.com/bestcomments Source code: https://github.com/border-radius/hn-best-comments
显示更多4 183
订阅者
-124 小时
-127 天
-2130 天
数据加载中...
吸引订阅者
七月 '26
七月 '26
+14
在4个频道中
六月 '26
+60
在6个频道中
Get PRO
五月 '26
+64
在5个频道中
Get PRO
四月 '26
+137
在9个频道中
Get PRO
三月 '26
+61
在9个频道中
Get PRO
二月 '26
+93
在4个频道中
Get PRO
一月 '26
+134
在8个频道中
Get PRO
十二月 '25
+61
在6个频道中
Get PRO
十一月 '25
+126
在9个频道中
Get PRO
十月 '25
+227
在6个频道中
Get PRO
九月 '25
+429
在6个频道中
Get PRO
八月 '25
+1 000
在6个频道中
Get PRO
七月 '25
+249
在5个频道中
Get PRO
六月 '25
+45
在2个频道中
Get PRO
五月 '25
+48
在1个频道中
Get PRO
四月 '25
+25
在0个频道中
Get PRO
三月 '25
+23
在1个频道中
Get PRO
二月 '25
+604
在2个频道中
Get PRO
一月 '25
+303
在2个频道中
Get PRO
十二月 '24
+25
在5个频道中
Get PRO
十一月 '24
+24
在2个频道中
Get PRO
十月 '24
+711
在3个频道中
Get PRO
九月 '24
+59
在4个频道中
Get PRO
八月 '24
+60
在3个频道中
Get PRO
七月 '24
+53
在1个频道中
Get PRO
六月 '24
+47
在3个频道中
Get PRO
五月 '24
+65
在7个频道中
Get PRO
四月 '24
+88
在2个频道中
Get PRO
三月 '24
+133
在4个频道中
Get PRO
二月 '24
+110
在4个频道中
Get PRO
一月 '24
+153
在8个频道中
Get PRO
十二月 '23
+84
在2个频道中
Get PRO
十一月 '23
+130
在7个频道中
Get PRO
十月 '23
+122
在2个频道中
Get PRO
九月 '23
+58
在0个频道中
Get PRO
八月 '23
+107
在0个频道中
Get PRO
七月 '23
+67
在0个频道中
Get PRO
六月 '23
+78
在0个频道中
Get PRO
五月 '23
+334
在0个频道中
Get PRO
四月 '23
+326
在0个频道中
Get PRO
三月 '23
+51
在0个频道中
Get PRO
二月 '23
+36
在0个频道中
Get PRO
一月 '23
+294
在0个频道中
Get PRO
十二月 '22
+1 360
在0个频道中
Get PRO
十一月 '22
+71
在0个频道中
Get PRO
十月 '22
+36
在0个频道中
Get PRO
九月 '22
+55
在0个频道中
Get PRO
八月 '22
+43
在0个频道中
Get PRO
七月 '22
+89
在0个频道中
Get PRO
六月 '22
+203
在0个频道中
Get PRO
五月 '22
+42
在0个频道中
Get PRO
四月 '22
+23
在0个频道中
Get PRO
三月 '22
+23
在0个频道中
Get PRO
二月 '22
+38
在0个频道中
Get PRO
一月 '22
+47
在0个频道中
Get PRO
十二月 '21
+36
在0个频道中
Get PRO
十一月 '21
+75
在0个频道中
Get PRO
十月 '21
+44
在0个频道中
Get PRO
九月 '21
+14
在0个频道中
Get PRO
八月 '21
+46
在0个频道中
Get PRO
七月 '21
+71
在0个频道中
Get PRO
六月 '21
+172
在0个频道中
Get PRO
五月 '21
+287
在0个频道中
Get PRO
四月 '21
+15
在0个频道中
Get PRO
三月 '21
+16
在0个频道中
Get PRO
二月 '21
+28
在0个频道中
Get PRO
一月 '21
+34
在0个频道中
Get PRO
十二月 '20
+639
在0个频道中
| 日期 | 订阅者增长 | 提及 | 频道 | |
| 11 七月 | 0 | |||
| 10 七月 | +2 | |||
| 09 七月 | +2 | |||
| 08 七月 | 0 | |||
| 07 七月 | 0 | |||
| 06 七月 | +4 | |||
| 05 七月 | +1 | |||
| 04 七月 | 0 | |||
| 03 七月 | +1 | |||
| 02 七月 | +1 | |||
| 01 七月 | +3 |
频道帖子
Re: Train sim created by just one person is being called the best ever made
Almost every time I see a halfway polished “solo developer” game, they did not do all the work themselves. Especially, they usually hire out the music, maybe other sounds, and much of the artwork. Sometimes they also have freelancers doing marketing and such. Sometimes even some paid help writing the software.
I highlight this not to bring those developers down, but because I think it’s important people understand how these things actually come to be, so they aren’t discouraged to try themselves by thinking they ought to actually be doing 100% of the work solo. That’s pretty rare.
topgrain2, 1 day ago
| 2 | Re: Apple sues OpenAI, accuses ex-employees of stealing trade secrets
It gets even worse. The person not only kept the laptop and used an exploit to download confidential Apple documents, they bragged about it to a contact who was still working at Apple who was also feeding him information:
> Liu allegedly kept an Apple-issued laptop after departing the company and exploited a vulnerability to download dozens of confidential Apple documents while he was working at OpenAI. He also maintained a relationship with Yu-Ting "Alyssa" Peng, an Apple employee who continued to give him updates on Apple's projects, vendor decisions, and engineering details. When Liu learned he still had access to Apple's systems, he texted Peng "LOL, I found out I can access the [network storage], so funny."
This is how you behave when you think you're so much smarter than everyone around you that consequences don't apply to you.
Whenever I leave a company I make sure everything that belongs to the company goes back to them and I wipe any access credentials or authenticator codes that might be on any of my devices. I can't imagine being so brazen that you'd keep the company laptop and then start using an exploit to download confidential information for your new employer.
Doing it at a the company that most aggressively enforces secrecy is even crazier.
Aurornis, 5 hours ago | 296 |
| 3 | Re: Muse Spark 1.1
Lot more details in the linked report https://ai.meta.com/static-resource/muse-spark-1-1-evaluatio...
From Terminal-bench-2.1 details,
> We use a bash-tool-only agent harness to evaluate 89 Terminal-Bench 2.1 tasks from the official repository, where resources are capped at 6 CPU cores and 8GB RAM.
This disqualifies the results. Each terminal bench task has a cpu upper limit and RAM upper limit. Overriding either is disqualification.
For reference, in tbench-2.1,
1. 0 out of 89 task allow 6 cpu cores (highest is 4, and i think only 1 task)
2. 8 out of 89 tasks allow 8GB RAM
This kind of shady benchmarking (I was talking about it just yesterday in a different context https://news.ycombinator.com/item?id=48838212) takes all joy out of building a harness to improve benchmark performance of a model because no matter what you do, you won't beat the headline (cheating) number. This is presumably why this model is not in the official benchmark leaderboard https://www.tbench.ai/leaderboard/terminal-bench/2.1
As an ex Meta employee, this is a little sad but not massively surprising. 'Number go up' is the core performance evaluation metric until PSC is done and you move on.
GodelNumbering, 2 days ago | 367 |
| 4 | Re: Why American ambulance rides are so expensive
I don't think the question is "Should ambulances be a thing?" though. It's a question of "Should someone in a situation where they need an ambulance have to balance the potentially life-threatening impact of saying no versus the potentially financially ruinous impact of saying yes?"
The (fairly obvious) answer to that no one should be in that situation. It's horrible. Society should find a better way to pay for ambulances. Most of the world has accepted that some system to spread the cost among everyone is better than putting people in that situation.
onion2k, 21 hours ago | 336 |
| 5 | Re: EU Parliament greenlights Chat Control 1.0
The EU is a farce, an undemocratic virtue signaling organization, and this is why:
- The Parliament voted against the first reading of this proposal twice in 2026, the first time they only supported limited cases for it, while the second time they actually defeated it fully.
- The Commission didn't care, and kept the proposal on the table by refusing to withdraw it.
- Once the Commission does that, the proposal goes on second-reading (despite the first-reading having defeated it) and it is established in a very PERVERSE way in EU law that to AVOID passing the proposal in second-reading you need ABSOLUTE majority which is incredibly hard to pursue (you would think that we would need an absolute majority to PASS a proposal that was previously defeated on first-reading, not instead needing absolute majority to DENY a previously defeated proposal that was again forced to the table).
- Furthermore, absences in practicality count as "No" on the rejection. So of course they scheduled the vote in the summer when notoriously there will be many absences.
By never withdrawing a defeated proposal they can effectively and in practicality pursue any agenda they want (it requires a massive mobilization effort to find absolute majority to defeat any proposal, especially when absences for any reason effectively count against rejection).
In PRACTICALITY, the Commission can pursue any agenda whenever and however they want, and throw the votes down the drain.
EU's democracy is lipstick on a pig.
fosk, 16 hours ago | 354 |
| 6 | Re: How the terrorist group Boko Haram uses frontier AI
> We saw in a movie how motorcycles can jump over bridges. We used
AI to learn how to do this. We gave it information, like what
motorcycles we use and the distance we need to jump and so on and it
gave us steps on what we have to do. We practiced a lot and kept
asking questions. We dug holes and filled them with broken glass and
fire to practice. 18 of us died in the process. Eight of us managed to do
it. The next time we attacked, we could jump.
Now listen, I'm not saying we need to give these guys more AI, but it clearly isn't yielding bad outcomes for us here.
"You're absolutely correct! For it to be a good practice ground you need to fill the trenches with broken glass and light the whole thing on fire"
arjie, 7 hours ago | 351 |
| 7 | Re: Apple sues OpenAI, accuses ex-employees of stealing trade secrets
Some pretty damning stuff:
> OpenAI also instructs new hires on how to avoid scrutiny when they leave Apple. For example, Mr. Tan warns them not to tell Apple that they have taken jobs at OpenAI, so they can stay at Apple as long as they can.
> Apple says it discovered a pattern of OpenAI recruits emailing themselves confidential information when leaving Apple, including Tan.
> OpenAI apparently used confidential Apple hardware information when approaching Apple suppliers, and tricked one company into using a "specific trade secret metal-finishing technique" for an OpenAI device by claiming it had Apple's permission to do so.
> Liu allegedly kept an Apple-issued laptop after departing the company and exploited a vulnerability to download dozens of confidential Apple documents while he was working at OpenAI.
Non-competes and the like are gross but what's described here isn't just "bring your expertise to OpenAI" it's "here is how to steal secrets on your way out" which is even grosser.
joshstrange, 3 hours ago | 356 |
| 8 | Re: My thoughts on the Bun Rust rewrite
I have learned so much reading Andrew’s code and as I said in the original post: Bun would never have happened without Zig.
> The post claims they were fuzzing their Zig code, while during our calls the whole Bun team told us that they were not fuzzing anything. This appears to be an outright fabrication.
Fuzzilli integration: https://github.com/oven-sh/bun/pull/24826
Merged PRs fixing issues Fuzzilli found in Bun’s Zig code:
- https://github.com/oven-sh/bun/pull/28926
- https://github.com/oven-sh/bun/pull/28934
- https://github.com/oven-sh/bun/pull/29255
- https://github.com/oven-sh/bun/pull/29210
- https://github.com/oven-sh/bun/pull/29199
Searching “Fuzzilli” shows more PRs: https://github.com/search?q=repo%3Aoven-sh%2Fbun+is%3Apr+Fuz...
Jarred, 1 day ago | 407 |
| 9 | Re: Building a real-time AI tutor for 5-year-olds
I think teaching a child to trust an LLM from a formative age is horrifically irresponsible.
If anything, an app should be made where a child learns to correct an LLM's mistakes and learn that it isn't trustworthy.
Actually, better, don't put an LLM in front of children. At all.
EDIT: If a use case is for children who can't afford good education, then use an LLM to make educational materials for children, review them, and make them available for free. After all, the contents are ripped off from human educators anyway.
TonyAlicea10, 8 hours ago | 469 |
| 10 | Re: EU Parliament greenlights Chat Control 1.0
Stupid parliamentary trick: Hold the vote on the day before the summer break - ensuring that many people have already returned to their home countries. Then use a sort of "reverse" parliamentary trick: the default is that this legislation is accepted. They needed an absolute majority - not of voting members, but of all members - to reject it.
Result: 314 against, 276 in favor, 17 abstentions, 113 absent
The EU is well on the way to becoming a totalitarian government.
ETA: It is shocking that 276 members of parliament would vote to support this. Are so many so naive? Or being paid off?
bradley13, 1 day ago | 540 |
| 11 | Re: The glass backbone: Why the Army's logistics will break in the next war
One of the most interesting innovations in the Ukraine war is their internal market place for drones, letting each drone group decide which drones they want to procure and use in battle.
It is not a top-down decision, production and supply as other armies use for their weapons logistics.
silvestrov, 1 day ago | 548 |
| 12 | Re: Train sim created by just one person is being called the best ever made
The dev is a person from Indonesia (Rizky Nova) who's device has 16GB of ram.
Being able to use the Unreal Engine for free to develop this is awesome. This couldn't have happened 10 years ago.
culi, 14 hours ago | 602 |
| 13 | Re: GPT-5.6
GPT-5.6 Sol sets a new SOTA on ARC-AGI-3: 7.8%
Sol is the first verified frontier model to ever beat an ARC-AGI-3 game
https://arcprize.org/results/openai-gpt-5-6
meetpateltech, 20 hours ago | 608 |
| 14 | Re: GPT-5.6
Funny to see that they did not include Fable 5 in their GeneBench and LifeSciBench comparisons because "it does not answer advanced biology questions and refuses the majority of questions in this eval".
Winner by default!
eig, 19 hours ago | 643 |
| 15 | Re: GPT-5.6
There is so much less drama involved with the Codex world. You don't realize how oppressive CC is until you've escaped it. Outages, weird restrictions, degradation, accelerated usage, etc etc etc.
postalcoder, 17 hours ago | 658 |
| 16 | Re: The glass backbone: Why the Army's logistics will break in the next war
A very insightful, and correct, piece.
I'll quote in full the following, which I think gets to the heart of the matter. If you have no push, you can't apply pressure to the point.
> The notion that amateurs talk tactics and professionals talk logistics is frequently discussed in military academies and war colleges, yet it is rarely reflected in the Army’s budget requests or modernization priorities. The outdated concept of the tooth-to-tail ratio, which implies the logistical tail is a bureaucratic waste that must be minimized to support the combat teeth, must be fundamentally reexamined. In modern warfare, the tail is the primary target. If the tail is severed, the teeth are rendered useless.
kayo_20211030, 17 hours ago | 673 |
| 17 | Re: EU Parliament greenlights Chat Control 1.0
Not only was Chat Control 1.0 already rejected twice by the European Parliament but:
- This vote took place on last day of the session when many MEPs had already left for Summer vacation - 112 MEPs of 719 didn't vote.
- The vote was called only two days before as an "Rule 170 - Urgent procedure" - 73 MEPs missed the vote making it "urgent". Normally it takes months of procedure to come up for a final vote.
spikels, 12 hours ago | 634 |
| 18 | Re: Postgres rewritten in Rust, now passing 100% of the Postgres regression tests
I spent a couple years managing a Postgres cluster with a petabyte of data. I wrote a couple blog posts from my work then[0][1]. I also wrote dozens of posts on the Postgres internals[2]. I've also given talks on how to generate fractals with SQL[3] and how to write a lisp interpreter in SQL[4].
[0] https://www.heap.io/blog/testing-database-changes-right-way
[1] https://www.heap.io/blog/analyzing-performance-millions-sql-...
[2] https://malisper.me/table-of-contents/
[3] https://www.youtube.com/watch?v=xKoYIvMFnoQ
[4] https://www.youtube.com/watch?v=MPSMH8w7nfw
malisper, 6 hours ago | 667 |
| 19 | Re: GPT-5.6
Codex has arguably been better than Claude Code for months now, but it's flown under the radar because it just didn't capture the same viral marketing effect and OpenAI in general has had more optics / PR issues than Anthropic amongst the online developer crowd. I use the word "better" not in the sense that the underlying GPT models are fundamentally smarter or more intelligent, but rather that as a product Codex is just simpler, cheaper, and abundantly reliable and low-drama.
nilkn, 10 hours ago | 601 |
| 20 | Re: GPT-5.6
The developer's guide (https://developers.openai.com/api/docs/guides/latest-model) has some interesting semantic tips for using the model:
> Intent understanding: GPT-5.6 can better infer the user’s underlying goal and intended level of work without you specifying every step. Continue to state important constraints, approval boundaries, and success criteria explicitly.
> Original image detail: GPT-5.6 preserves the original dimensions of images sent with original or auto detail instead of resizing them to a patch budget or pixel-dimension limit.
> Use shorter prompts: In internal evaluations, replacing long, explicit system prompts with minimal prompts improved scores by roughly 10–15%, while reducing total tokens by 41–66% and cost by 33–67%.
> Avoid generic brevity instructions: GPT-5.6 is more sensitive than GPT-5.5 to instructions such as “Be concise,” “Keep it short,” or “Use minimal text.”
> Control warmth: GPT-5.6 does not become meaningfully better when prompted to be broadly friendlier or more empathetic.
minimaxir, 10 hours ago | 560 |
