Hacker News: ericpauley

New comment by ericpauley in "I run multiple $10K MRR companies on a $20/month tech stack"

ericpauley — Sun, 12 Apr 2026 18:14:16 +0000

If sshd is OOMing on 64GB something else is going on…

New comment by ericpauley in "Artemis II's toilet is a moon mission milestone"

ericpauley — Thu, 02 Apr 2026 02:47:32 +0000

Thoroughly enjoyed reading this, especially the author’s repeated obsession on the door vs. curtain innovation…

New comment by ericpauley in "The future of version control"

ericpauley — Sun, 22 Mar 2026 18:05:55 +0000

Yeah this seems silly. You can do the same thing in git (add and commit with the conflict still there)! Why you would want to is a real mystery.

New comment by ericpauley in "HP trialed mandatory 15-minute support call wait times (2025)"

ericpauley — Sat, 21 Mar 2026 11:53:05 +0000

Insisting on saying VoIP to the Mint rep instead of WiFi Calling (the term used by Apple, Google, Mint, and practically everyone else) is asking for a bad time.

New comment by ericpauley in "Waymo Safety Impact"

ericpauley — Fri, 20 Mar 2026 10:52:55 +0000

Yes, but Waymo also has to drive on the road with those drivers, and these stats include crashes that are their fault. Diligent drivers get hit by drunk/distracted drivers all the time.

New comment by ericpauley in "Autoresearch for SAT Solvers"

ericpauley — Thu, 19 Mar 2026 10:49:06 +0000

As GP noted the issue is that even better versions than competed in MaxSAT are likely in the training data or web resources.

New comment by ericpauley in "Ask HN: How do you deal with people who trust LLMs?"

ericpauley — Thu, 19 Mar 2026 04:02:23 +0000

Sure, but that doesn’t explain the volume of these complaints. I think the more likely answer is the pitiful sycophancy of some models as demonstrated in BSBench.

New comment by ericpauley in "What’s on HTTP?"

ericpauley — Thu, 19 Mar 2026 03:59:10 +0000

My thoughts exactly. By this logic both are fragile because they run over lossy wireless networks.

The composability of TLS/HTTP is really a beautiful thing.

New comment by ericpauley in "Autoresearch for SAT Solvers"

ericpauley — Thu, 19 Mar 2026 03:55:45 +0000

Funnily, this was precisely the question I had after posting this (and the topic of an LLM disagreement discussed in another thread). Turns out not, but sibling comment is another confounding factor.

New comment by ericpauley in "Ask HN: How do you deal with people who trust LLMs?"

ericpauley — Thu, 19 Mar 2026 02:29:49 +0000

Again a model issue. At the risk of coming off as a thread-wide apologist, here are my results on Opus:

Good:

> The research is generally positive but it’s not unconditionally “good for you” — the framing matters.

> What the evidence supports for moderate consumption (3-5 cups/day): lower risk of type 2 diabetes, Parkinson’s, certain liver diseases (including liver cancer), and all-cause mortality……

Bad:

> The premise is off. Moderate daily coffee consumption (3-5 cups) isn’t considered bad for you by current medical consensus. It’s actually associated with reduced risk of type 2 diabetes, Parkinson’s, and some liver diseases in large epidemiological studies.

> Where it can cause problems: Heavy consumption (6+ cups) can lead to anxiety, insomnia……

This isn’t just my own one-off examples. Claude dominates the BSBench: https://petergpt.github.io/bullshit-benchmark/viewer/index.v...

New comment by ericpauley in "Ask HN: How do you deal with people who trust LLMs?"

ericpauley — Thu, 19 Mar 2026 02:25:00 +0000

Sure LLMs make mistakes, but have you looked at the accuracy of the average top search results recently? The SERPs are packed with SEO-infested articles that are all written by LLMs anyway (and almost universally worse ones than you could use yourself). In many cases the stakes are low enough (and the cost of manually sifting through the junk high enough) that it’s worth going with the empirically higher quality answer than the SEO spam.

This of course doesn’t apply to high-stakes settings. In these cases I find LLMs are still a great information retrieval approach, but it’s a starting point to manual vetting.

New comment by ericpauley in "Ask HN: How do you deal with people who trust LLMs?"

ericpauley — Thu, 19 Mar 2026 02:18:56 +0000

This is an oft-repeated meme, but I’m convinced the people saying it are either blindly repeating it, using bad models/system prompts, or some other issue. Claude Opus will absolutely push back if you disagree. I routinely push back on Claude only to discover on further evaluation that the model was correct.

As a test I just did exactly what you said in a Claude Opus 4.6 session about another HN thread. Claude considered* the contradiction, evaluated additional sources, and responded backing up its original claim with more evidence.

I will add that I use a system prompt that explicitly discourages sycophancy, but this is a single sentence expression of preference and not an indication of fundamental model weakness.

* I’ll leave the anthropomorphism discussions to Searle; empirically this is the observed output.

New comment by ericpauley in "Autoresearch for SAT Solvers"

ericpauley — Thu, 19 Mar 2026 02:10:26 +0000

True. I’d be curious whether a combination of matching comp/training cutoff and censoring web searches could yield a more precise evaluation.

New comment by ericpauley in "Autoresearch for SAT Solvers"

ericpauley — Thu, 19 Mar 2026 02:05:25 +0000

It should be noted that MaxSAT 2024 did not include z3, as with many competitions. It’s possible (I’d argue likely) that the agent picked up on techniques from Z3 or some other non-competing solver, rather than actually discovering some novel approach.

New comment by ericpauley in "Aliens.gov ~ domain registered 17MAR2026"

ericpauley — Wed, 18 Mar 2026 14:17:22 +0000

All of .gov is contracted to Cloudflare: https://www.cloudflare.com/press/press-releases/2023/cloudfl...

New comment by ericpauley in "1M context is now generally available for Opus 4.6 and Sonnet 4.6"

ericpauley — Sat, 14 Mar 2026 14:43:59 +0000

Used Claude through copilot for so long before switching to CC. Even for the same model the difference is shocking. Copilot’s harness and the underlying Claude models are not well-matched compared to the vertically-integrated Claude Code harness.

New comment by ericpauley in "MCP server that reduces Claude Code context consumption by 98%"

ericpauley — Sun, 01 Mar 2026 13:04:27 +0000

> I had zero visibility into where my Claude Code tokens were actually going

/context?

New comment by ericpauley in "Man accidentally gains control of 7k robot vacuums"

ericpauley — Mon, 23 Feb 2026 12:44:27 +0000

This was the example that really drove home all the other points for me. Not only is Valetudo opinionated software, but you'll be accused of having "fictional budget concerns" for wanting a very reasonable feature.

I occasionally take my Roborock upstairs on weekends for a vacuum. Turns out it will also do a basic mop run with the water in the tank. Takes me 5 minutes of setup/tear down to get an extra floor for no extra cost. It would take me more time to babysit the extra base cleaning task of a second mop, so this saves me time and money.

To me, this demonstrates that Valetudo is intended to be hobby pursuit of maximal automation/freedom at all costs, resulting in a system that has worse features and takes more work than the base software. I applaud the creator for being so clear in this mission to the point of explicitly encouraging me not to use it.

New comment by ericpauley in "Man accidentally gains control of 7k robot vacuums"

ericpauley — Sun, 22 Feb 2026 16:25:57 +0000

I found the “Why Not Valetudo” page on that site extremely persuasive. I would consider myself technically inclined. I also own a robot vacuum so I can spend more time doing important things that leverage my skills. Valetudo does not serve this mission.

Very impressive, but I disagree that this is the clear best choice for anywhere close to anyone.

New comment by ericpauley in "I cannot curl https://example.com (on some distros)"

ericpauley — Sun, 15 Feb 2026 15:10:14 +0000

Really great analysis. Always cool to see the divide between PKI in theory and practice.

It does make me wonder if the zealous pursuit of shorter expirations has gone too far, especially up at the root. Is there good public discussion on root expiration? Seems to mostly come up when old devices get bricked because of it. Certainly 15 year expirations are not a substitute for extremely strict root key management or root key revocation.