Hacker News: fwipsy

New comment by fwipsy in "Backpacks got worse on purpose"

fwipsy — Wed, 15 Apr 2026 14:45:19 +0000

I missed this but now that you point it out, seems plausible.

Pretty ironic on an article about quality products being replaced by cheaper ones.

New comment by fwipsy in "Backpacks got worse on purpose"

fwipsy — Wed, 15 Apr 2026 14:42:06 +0000

I don't see which of those he violated?

Also from the guidelines: "HN is for conversation between humans." Granted, it's under comment guidelines, but I think it's acting in the spirit of the rule to point out AI output.

New comment by fwipsy in "Backpacks got worse on purpose"

fwipsy — Wed, 15 Apr 2026 14:37:25 +0000

Perhaps it's gotten harder to determine by eye, but Google will still point you towards trustworthy brands in 2 minutes. The problem is people don't care or can't be bothered to Google.

New comment by fwipsy in "I wrote to Flock's privacy contact to opt out of their domestic spying program"

fwipsy — Wed, 15 Apr 2026 01:37:22 +0000

Police prioritizing responses to violent crimes where lives may in danger seems reasonable to me.

New comment by fwipsy in "The Closing of the Frontier"

fwipsy — Sun, 12 Apr 2026 19:37:48 +0000

Sorry, I removed that line from the comment because I realized it wasn't really in line with the rest of the point I was making.

New comment by fwipsy in "The Closing of the Frontier"

fwipsy — Sun, 12 Apr 2026 19:34:10 +0000

Security researchers always having a model one generation newer than the general public would still achieve the stated goals.

New comment by fwipsy in "The Closing of the Frontier"

fwipsy — Sun, 12 Apr 2026 19:32:40 +0000

The opportunity from the early days of the American frontier is not typical. Instead, it's the brief burst of unrestrained growth as a better-adapted organization (the US, software companies) expands into, and expands, a niche--cannibalizing the previous occupant (Native Americans, older stagnant companies.) At times growth is so rapid that individuals are able to advance the frontier, but if the field stagnates, individuals will be outcompeted by corporations.

So, opportunity for individuals comes from disruption. Creative destruction is good up to a point, but it results from advancing capabilities. Technological advances compound and accelerate exponentially. Eventually we reach the point where any malcontent can destroy the world by snapping their fingers. At some point we need to place restrictions on the capabilities accessible to individuals. We have reached that point with nuclear weapons, and I think it is sensible to believe that AI is reaching that point as well.

New comment by fwipsy in "AI Will Be Met with Violence, and Nothing Good Will Come of It"

fwipsy — Sun, 12 Apr 2026 18:07:44 +0000

Turns out I misremembered. Incas never fully united, and even though Spaniards had a huge technological advantage in some battles, the war as a whole was more evenly matched. Technology, disease, and infighting ALL played a part in their victory.

New comment by fwipsy in "AI Will Be Met with Violence, and Nothing Good Will Come of It"

fwipsy — Sun, 12 Apr 2026 17:53:08 +0000

Don't forget horses, armor, and steel weapons. It seems like Incan weapons had a lot of trouble penetrating Spanish armor, while the reverse was not true. Also, the Incas didn't just lack cavalry; they lacked the weapons and tactics to counter cavalry (such as pike formations.)

That said, I was thinking of the Battle of Cajamarca, which was actually a Spanish ambush. 100x was probably overstating it; under other circumstances (e.g. rough terrain) Spanish technology had less of an edge.

New comment by fwipsy in "AI Will Be Met with Violence, and Nothing Good Will Come of It"

fwipsy — Sun, 12 Apr 2026 15:22:06 +0000

Actually this is another good counterexample! As I recall, Incas lost battles against the Spaniards where they had something like 100x the numbers. It's true that they were initially divided, but they quickly united against the Spanish--and it didn't really help. The technological advantage was insurmountable.

New comment by fwipsy in "AI Will Be Met with Violence, and Nothing Good Will Come of It"

fwipsy — Sun, 12 Apr 2026 15:12:11 +0000

Right. I am not saying diseases were a factor in every conquest. Just refuting parent saying that conquest is "only possible" through infighting. It's not - overwhelming technological advantage or disease are also sufficient even against a united culture.

New comment by fwipsy in "AI Will Be Met with Violence, and Nothing Good Will Come of It"

fwipsy — Sun, 12 Apr 2026 14:34:20 +0000

> internal competition and in-fighting of the natives.

What about diseases which killed up to 95% of the population? I think you are basically correct, except for the historical analogy.

New comment by fwipsy in "Show HN: Pardonned.com – A searchable database of US Pardons"

fwipsy — Sun, 12 Apr 2026 14:14:21 +0000

I'm sorry that you experienced that. I understand the importance of listening to abuse victims. However, if child sacrifice did happen, it seems unlikely to ever be proven. I do not think that the case against Epstein or his associates is strengthened by assuming every accusation is true. For people who are not thinking about the subject carefully, learning that some accusations are inflated may cast others into doubt. What is provable is heinous enough.

New comment by fwipsy in "Show HN: Pardonned.com – A searchable database of US Pardons"

fwipsy — Sat, 11 Apr 2026 16:02:08 +0000

From the article you linked, child sacrifice allegations came from an anonymous FBI interview in 2019 and are not confirmed by any credible evidence. There are no cannibalism allegations; the word "cannibal" only appears in innocuous contexts.

So child sacrifice and cannibalism are only technically "in the Epstein files;" there's very little evidence that anyone did those things. For other readers, if you hadn't heard about this, that's probably why.

New comment by fwipsy in "Mystery over 8 missing or dead scientific experts"

fwipsy — Fri, 10 Apr 2026 02:49:31 +0000

Seems like only two of the individuals were related in any way. I'm not convinced that 8 sudden deaths/disappearances is above the base rate for "people involved with aerospace?" That's a pretty big group, I'd imagine.

New comment by fwipsy in "Cambodia unveils statue to honour famous landmine-sniffing rat"

fwipsy — Wed, 08 Apr 2026 03:53:14 +0000

Agree if the animals are treated well; but I have a very high bar for that. I would also accept eating animals which died of old age if it could be done safely. It's easier to just round this to "vegan."

New comment by fwipsy in "GLM-5.1: Towards Long-Horizon Tasks"

fwipsy — Wed, 08 Apr 2026 03:42:16 +0000

First answer: If you haven't, give it a shot on whatever you already have. MoE models like Qwen3 and GPT-OSS are good on low-end hardware. My RTX 4060 can run qwen3:30b at a comfortable reading pace even though 2/3 of it spills over into system RAM. Even on an 8-year-old tiny PC with 32gb it's still usable.

Second answer: ask an AI, but prices have risen dramatically since their training cutoff, so be sure to get them to check current prices.

Third answer: I'm not an expert by a long shot, but I like building my own PCs. If I were to upgrade, I would buy one of these:

Framework desktop with 128gb for $3k or mainboard-only for $2700 (could just swap it into my gaming PC.) Or any other Strix Halo (ryzen AI 385 and above) mini PC with 64/96/128gb; more is better of course. Most integrated GPUs are constrained by memory bandwidth. Strix Halo has a wider memory bus and so it's a good way to get lots of high-bandwidth shared system/video RAM for relatively cheap. 380=40%; 385=80%; 395=100% GPU power.

I was also considering doing a much hackier build with 2x Tesla P100s (16gb HBM2 each for about $90 each) in a precision 5820 (cheap with lots of space and power for GPUs.) Total about $500 for 32gb HBM2+32gb system RAM but it's all 10-year-old used parts, need to DIY fan setup for the GPUs, and software support is very spotty. Definitely a tinker project; here there be dragons.

New comment by fwipsy in "GLM-5.1: Towards Long-Horizon Tasks"

fwipsy — Wed, 08 Apr 2026 03:27:04 +0000

qwen3:0.6b is 523mb, what model are you talking about? You seem to have a specific one in mind but the parent comment doesn't mention any.

For a hobby/enthusiast product, and even for some useful local tasks, MoE models run fine on gaming PCs or even older midrange PCs. For dedicated AI hardware I was thinking of Strix Halo - with 128gb is currently $2-3k. None of this will replace a Claude subscription.

New comment by fwipsy in "GLM-5.1: Towards Long-Horizon Tasks"

fwipsy — Wed, 08 Apr 2026 01:48:54 +0000

Agree directionally but you don't need $50k. $5k is plenty, $2-3k arguably the sweet spot.

New comment by fwipsy in "GLM-5.1: Towards Long-Horizon Tasks"

fwipsy — Wed, 08 Apr 2026 01:46:14 +0000

No moat: yes. Cooked: no. It's a race. Why assume they're going to lose? It relies on (2) which is only true if AI usefulness plateaus at some level of compute. That's a huge claim to be making at this stage. (3) AI has lots of killer products already. The big one is filling in moats. Unrealized potential though for sure.