Ha ha, well that's a relief. I thought the article was going to say that enabling TCP_NODELAY is causing problems in distributed systems. I am one of those people who just turn on TCP_NODELAY and never look back because it solves problems instantly and the downsides seem minimal. Fortunately, the article is on my side. Just enable TCP_NODELAY if you think it's a good idea. It apparently doesn't break anything in general.
Erasure coding is an interesting topic for me. I've run some calculations on the theoretical longevity of digital storage. If you assume that today's technology is close to what we'll be using for a long time, then cross-device erasure coding wins, statistically. However, if you factor in the current exponential rate of technological development, simply making lots of copies and hoping for price reductions over the next few years turns out to be a winning strategy, as long as you don't have vendor lock-in. In other words, I think you're making great choices.
I question that math. Erasure coding needs less than half as much space as replication, and imposes pretty small costs itself. Maybe we can say the difference is irrelevant if storage prices will drop 4x over the next five years? But looking at pricing trends right now... that's not likely. Hard drives and SSDs are about the same price they were 5 years ago. The 5 years before that SSDs were seeing good advancements, but hard drive prices only advanced 2x.
Sure, but LLMs are trying to build the algorithms of the human mind backwards, converge on similar functionality based on just some of the inputs and outputs. This isn't an efficient or a lossless process.
The fact that they can pull it off to this extent was a very surprising finding.
Letters with rounded terminals are especially popular for public signage in a few Asian countries, e.g. Japan and Korea.
That is why Microsoft Windows has included such a rounded font for the Korean script: Gulim. On Windows, if you want to render a text with Latin letters with rounded ends, you can use Gulim for the normal text, coupled with Arial Round for the bold text.
On MacOS, there was a Hiragino Maru Gothic rounded font for Japanese (where also the Latin letters are rounded). I no longer use Apple computers, so I do not know whether the Hiragino fonts have remained the fonts provided for Japanese.
I'll take a stab at this: LLMs currently seem to be rather good at details, but they seem to struggle greatly with the overall picture, in every subject.
- If I want Claude Code to write some specific code, it often handles the task admirably, but if I'm not sure what should be written, consulting Claude takes a lot of time and doesn't yield much insight, where as 2 minutes with a human is 100x more valuable.
- I asked ChatGPT about some political event. It mirrored the mainstream press. After I reminded it of some obvious facts that revealed a mainstream bias, it agreed with me that its initial answer was wrong.
These experiences and others serve to remind me that current LLMs are mostly just advanced search engines. They work especially well on code because there is a lot of reasonably good code (and tutorials) out there to train on. LLMs are a lot less effective on intellectual tasks that humans haven't already written and published about.
Are you sure? The third section of each review lists the “Most prescient” and “Most wrong” comments. That sounds exactly like what you're looking for. For example, on the "Kickstarter is Debt" article, here is the LLM's analysis of the most prescient comment. The analysis seems accurate and helpful to me.
phire
> “Oculus might end up being the most successful product/company to be kickstarted… > Product wise, Pebble is the most successful so far… Right now they are up to major version 4 of their product. Long term, I don't think they will be more successful than Oculus.”
With hindsight:
Oculus became the backbone of Meta’s VR push, spawning the Rift/Quest series and a multi‑billion‑dollar strategic bet.
Pebble, despite early success, was shut down and absorbed by Fitbit barely a year after this thread.
That’s an excellent call on the relative trajectories of the two flagship Kickstarter hardware companies.
Until someone publishes a systematic quality assessment, we're grasping at anecdotes.
It is unfortunate that the questions of "how well did the LLM do?" and "how does 'grading' work in this app?" seem to have gone out the window when HN readers see something shiny.
Yes. And the article is a perfect example of the dangerous sort of automation bias that people will increasingly slide into when it comes to LLMs. I realize Karpathy is sort of incentivized toward this bias given his career, but he doesn't even spend a single sentence even so much as suggesting that the results would need further inspection, or that they might be inaccurate.
The LLM is consulted like a perfect oracle, flawless in its ability to perform a task, and it's left at that. Its results are presented totally uncritically.
For this project, of course, the stakes are nil. But how long until this unfounded trust in LLMs works its way into high stakes problems? The reign of deterministic machines for the past few centuries has ingrained a trust in the reliability of machines in us that should be suspended when dealing with an inherently stochastic device like an LLM.
I get what you're saying, but looking at some examples, they look kinda of right, but there are a lot of misleading facts sprinkled, making his grading wrong. It is useful, but I'd suggest to be careful to use this to make decisions.
Some of the issues could be resolved with better prompting (it was biased to always interpret every comment through the lens of predictions) and LLM-as-a-judge, but still. For example, Anthropic's Deep Research prompts sub-agents to pass original quotes instead of paraphrasing, because it can deteriorate the original message.
Some examples:
Swift is Open Source (2015)
===========================
sebastiank123 got a C-, and was quoted by the LLM as saying:
> “It could become a serious Javascript competitor due to its elegant syntax, the type safety and speed.”
Now, let's read his full comment:
> Great news! Coding in Swift is fantastic and I would love to see it coming to more platforms, maybe even on servers. It could become a serious Javascript competitor due to its elegant syntax, the type safety and speed.
I don't interpret it as a prediction, but a desire. The user is praising Swift. If it went the server way, perhaps it could replace JS, to the user's wishes. To make it even clearer, if someone asked the commenter right after: "Is that a prediction? Are you saying Swift is going to become a serious Javascript competitor?" I don't think its answer would be 'yes' in this context.
How to be like Steve Ballmer (2015)
===================================
Most wrong
----------
> corford (grade: D) (defending Ballmer’s iPhone prediction):
> Cited an IDC snapshot (Android 79%, iOS 14%) and suggested Ballmer was “kind of right” that the iPhone wouldn’t gain significant share.
> In 2025, iOS is one half of a global duopoly, dominates profits and premium segments, and is often majority share in key markets. Any reasonable definition of “significant” is satisfied, so Ballmer’s original claim—and this defense of it—did not age well.
Full quote:
> And in a funny sort of way he was kind of right :) http://www.forbes.com/sites/dougolenick/2015/05/27/apple-ios...
> Android: 79% versus iOS: 14%
"Any reasonable definition of 'significant' is satisfied"? That's not how I would interpret this. We see it clearly as a duopoly in North America. It's not wrong per se, but I'd say misleading. I know we could take this argument and see other slices of the data (premium phones worldwide, for instance), I'm just saying it's not as clear cut as it made it out to be.
> volandovengo (grade: C+) (ill-equipped to deal with Apple/Google):
>
> Wrote that Ballmer’s fast-follower strategy “worked great” when competitors were weak but left Microsoft ill-equipped for “good ones like Apple and Google.”
> This is half-true: in smartphones, yes. But in cloud, office suites, collaboration, and enterprise SaaS, Microsoft became a primary, often leading competitor to both Apple and Google. The blanket claim underestimates Microsoft’s ability to adapt outside of mobile OS.
That's not what the user was saying:
> Despite his public perception, he's incredibly intelligent. He has an IQ of 150.
>
> His strategy of being a fast follower worked great for Microsoft when it had crappy competitors - it was ill equipped to deal with good ones like Apple and Google.
He was praising him and he did miss opportunities at first. OC did not make predictions of his later days.
[Let's Encrypt] Entering Public Beta (2015)
===========================================
- niutech: F "(endorsed StartSSL and WoSign as free options; both were later distrusted and effectively removed from the trusted ecosystem)"
Full quote:
> There are also StartSSL and WoSign, which provide the A+ certificates for free (see example WoSign domain audit: https://www.ssllabs.com/ssltest/analyze.html?d=checkmyping.c...)
>
> pjbrunet: F (dismissed HTTPS-by-default arguments as paranoid, incorrectly asserted ISPs had stopped injection, and underestimated exactly the use cases that later moved to HTTPS)
Full quote:
> "We want to see HTTPS become the default."
>
> Sounds fine for shopping, online banking, user authorizations. But for every website? If I'm a blogger/publisher or have a brochure type of website, I don't see point of the extra overhead.
>
> Update: Thanks to those who answered my question. You pointed out some things I hadn't considered. Blocking the injection of invisible trackers and javascripts and ads, if that's what this is about for websites without user logins, then it would help to explicitly spell that out in marketing communications to promote adoption of this technology. The free speech angle argument is not as compelling to me though, but that's just my opinion.
I thought the debate was useful and so did pjbrunet, per his update.
I mean, we could go on, there are many others like these.
I think your intuition matches mine. When I try to apply Claude Code to a large code base, it spends a long time looking through the code and then it suggests something incorrect or unhelpful. It's rarely worth the trouble.
When I give AI a smaller or more focused project, it's magical. I've been using Claude Code to write code for ESP32 projects and it's really impressive. OTOH, it failed to tell me about a standard device driver I could be using instead of a community device driver I found. I think any human who works on ESP-IDF projects would have pointed that out.
Have you tried having AI build up documentation on the code first and then correct it where it’s understanding is wrong, then running code changes with the docs in the context, you can even separate it out for each module if you are daring. Ai still takes alot of hand holding to be productive with, which means our jobs are safe for now until they start learning about SWe principles somehow.
In large projects you need to actually point it to the interesting files, because it has no way of knowing what it doesn’t know. Tell it to read this and that, creating summary documents, then clear the context and point it at those summaries. A few of those passes and you‘ll get useful results. A gap in its knowledge of relevant code will lead to broken functionality. Cursor and others have been trying to solve this with semantic search (embeddings) but IMO this just can’t work because relevance of a code piece for a task is not determinable by any of its traits.
I‘ve had a similar feeling before Opus 4.5. Now it suddenly clicks with me, and it has passed the shittiness threshold, into the „often useful“ area. I suspect that’s because Apple is partnering with Anthropic and they will have improved Swift support.
Eg it‘s great for refactoring now, it’s often updating the README along with renames without me asking. It’s also really good at rebasing quickly, but only by cherry-picking inside a worktree. Churning out small components I don’t want to add a new dependency for, those are usually good on first try.
For implementing whole features, the space of possible solutions is way too big to always hit something that I‘ll be satisfied with. Once I have an idea on how to implement something in broad strokes, I can give a very error ridden first draft to it as a stream of thoughts, let it read all required files, and make an implementation plan. Usually that’s not too far off, and doesn’t take that long. Once that’s done, Opus 4.5 is pretty good at implementing that plan. Still I read every line, if this will go to production.
I start new projects "AI-first" – start with docs, and refining them on the go, with multiple CLAUDE.md in different folders (to give a right context where it's needed). This alone increases the chances of it getting tasks right tenfold. Plus I almost always verify myself all the code produced.
Ironically, this would be the best workflow with humans too.
The article suggests that because the power and cooling are customized, it would take a ton of effort to run the new AI servers in a home environment, but I'm skeptical of that. Home-level power and cooling are not difficult these days. I think when the next generation of AI hardware comes out (in 3-5 years), there will be a large supply of used AI hardware that we'll probably be able to repurpose. Maybe we'll sell them as parts. It won't be plug-and-play at first, but companies will spring up to figure it out.
If not, what would these AI companies do with the huge supply of hardware they're going to want to get rid of? I think a secondary market is sure to appear.
At minimum, you'd need to wire in new 240V circuits, and you could only run one or two of these servers before you'd need a service upgrade.
Then you'd have to deal with noise from a literal wall of fans, or build a separate high capacity water cooling system (and still deal with dumping that heat somewhere).
A utility is most likely only going to offer a 240V 400A single-phase service at best for a residence in the US, 320A can be used continuously. If you need more they’ll usually offer multiple 400A services.
I’ve heard stories about people convincing their utility to install three-phase service drops in their homes, but usually it’s not an option.
Anyways, 320A continuous load at 240V single-phase is 76.8kW, if you assume 25kW per server (20 kW for server, 5kW for cooling), you can run (3) servers and 15 tons of cooling and still have just enough left for one 120V 15A circuit to charge your phone and power a light.
While I agree with you in principle (people should do their own thinking if they want gifts to be genuine), I thought I would go ahead and see how well Gemini can advise someone on choosing a gift for you.
The advice seems very good. What do you think? A donation to EFF or an open source project, a rare book, or handcrafted headphones seem like a good start for someone who can't afford anything extravagant.
Yeah, that was a little joke I slipped in, but OTOH, I've seen headphone kits that used 3D printed parts and it's not difficult to imagine someone replacing the 3D printed parts with handcrafted wood.
And at the same time it's a not uncommon activity especially around university physics labs to wind various purpose-built magnets and transformers with exotic shapes.
I usually want to cut pies into 14 pieces. Some might want 11 or 13. (17 is just too many.) I demand that we implement a system where a circle is 2 * 3 * 4 * 5 * 7 * 3 * 11 * 13 = 360360 degrees, so that we can cut pies evenly at anywhere from 2 to 15 slices. If my baker cuts a slice at 25739 degrees, I want a refund! (I'll keep the pie, because the pie is obviously useless.)
(720720 might be OK too so we can cut 16 pieces, but honestly, if you're cutting 16 pieces, you're not going to measure. You're just going to divide pieces in half until you have 16. 360360 is the future.)
reply