Hacker News: pell

New comment by pell in "Prompts Aren't Real"

pell — Sun, 20 Sep 2026 17:38:10 +0000

From the HN guidelines:

> Be kind. Don't be snarky. Converse curiously; don't cross-examine. Edit out swipes.

>Comments should get more thoughtful and substantive, not less, as a topic gets more divisive.

>When disagreeing, please reply to the argument instead of calling names. "That is idiotic; 1 + 1 is 2, not 3" can be shortened to "1 + 1 is 2, not 3."

>Don't be curmudgeonly. Thoughtful criticism is fine, but please don't be rigidly or generically negative.

New comment by pell in "Key symbols we lost to time, pt. 2: The Mac side"

pell — Sun, 20 Sep 2026 17:17:44 +0000

Europe is also a little off here. Different countries use different layouts across Europe.

New comment by pell in "Pirate Face Rescues LLM Models from Deletion"

pell — Sun, 20 Sep 2026 17:04:14 +0000

> Is there a fully in-browser torrent option that has the same UX as a regular file download in Firefox?

Opera did back in the day.

New comment by pell in "Show HN: Liniora – Ever thought about replacing your project manager?"

pell — Sat, 12 Sep 2026 12:38:59 +0000

I was surprised to see that you don’t integrate with Linear which I assumed the name was related to.

New comment by pell in "TiVo to charge money for skipping commercials in your own recordings"

pell — Mon, 07 Sep 2026 13:09:08 +0000

I don’t see why China would have an incentive to keep investing in open models long-term. They aren’t doing that for the betterment of society.

New comment by pell in "TiVo to charge money for skipping commercials in your own recordings"

pell — Mon, 07 Sep 2026 13:08:03 +0000

Astra flagged my request to look into a security vulnerability into my own(!) code. So I’m not sure.

New comment by pell in "Anthropic's best AI model struggles to attract users as cheaper tools thrive"

pell — Mon, 24 Aug 2026 20:05:54 +0000

What are you doing that uses the entire Fable credits so quickly? Genuine question.

New comment by pell in "Anthropic's best AI model struggles to attract users as cheaper tools thrive"

pell — Mon, 24 Aug 2026 08:32:29 +0000

>I would way rather have Sol all day every data, consistently, than a slightly better Fable for like, 1 prompt every 5 hours, and only when Anthropic decides to not treat me like a cyber criminal.

I have used Fable heavily on the lower Max plan and you are really exaggerating the limits here. I've had many multi-hour sessions with Fable on Max.

New comment by pell in "Grok 4.6"

pell — Wed, 12 Aug 2026 22:41:40 +0000

It took many many many turns for me to have the model even acknowledge that the fake elector scheme was actually a thing. It's very much primed to answer vaguely when it goes against the current political ideals of its owner.

New comment by pell in "How to Make a Nintendo 64 Game in 2026"

pell — Thu, 06 Aug 2026 20:54:23 +0000

Giving people context is gatekeeping now?

New comment by pell in "Our position on open-weights models"

pell — Tue, 28 Jul 2026 10:23:24 +0000

One, as bad as it is, has its biggest critics live and speak out freely. The other silences, imprisons or kills them.

New comment by pell in "Decathlon Germany adds Wero payment option to decathlon.de website"

pell — Mon, 27 Jul 2026 20:22:11 +0000

Most EU countries have had some sort of payment standard and use different kinds of setups so it’s not as easy as doing it in one jurisdiction. Also the EU generally is quite conservative and slow.

New comment by pell in "LLM-Integrated Multivariable Calculus Course"

pell — Sun, 19 Jul 2026 04:09:33 +0000

How are you guaranteeing accuracy?

New comment by pell in "Founding a company in Germany: €9600, 152 days and I still can't send an invoice"

pell — Wed, 24 Jun 2026 14:01:06 +0000

I really didn’t have the feeling that Germans are shocked about some of these inefficiencies at all. Typically they’re just a result of long processes thus difficult to change. I do agree that Germany and the EU as a whole should not make this process so cumbersome. It shouldn’t be easier to found a US foreign LLC from abroad than a company in one’s country or residence. The proposed EU LLC might solve some of those issues if it comes to fruition.

Regarding the economy though, Germany is still the third richest country on earth. I think this talking point about their huge regression is mostly FUD.

New comment by pell in "There is minimal downside to switching to open models"

pell — Mon, 22 Jun 2026 11:25:37 +0000

Whenever I ask people to explain their issues with NPR it’s some cherry-picked news articles here and there that were somewhat biased. In my experience NPR often tries to be incredibly neutral, almost comically so, when criticizing any administration.

New comment by pell in "Dutch Railways offers unlimited off-peak train travel nationwide for €49/month"

pell — Thu, 18 Jun 2026 21:18:37 +0000

The Netherlands runs around 3000 trains a day vs. 50k in Germany. That doesn't excuse Germany's problems which were also predicted years in advance when they stopped investing in maintenance and infrastructure but also shows that the comparison is not entirely fair.

New comment by pell in "Your ePub Is fine"

pell — Mon, 15 Jun 2026 06:26:45 +0000

Flash had many issues for sure, first and foremost security. But I can’t help but feel sad of what was lost since then. The Flash era produced some really unique experiences on the web.

New comment by pell in "xAI is looking more like a datacentre REIT than a frontier lab"

pell — Mon, 08 Jun 2026 19:42:59 +0000

I don’t want to open up that whole can of worms but Grok on any vaguely philosophical or political topic is a scaredy cat and has a very hard time staying factual if it could make Musk or the conservative movement appear negatively.

New comment by pell in "CS336: Language Modeling from Scratch"

pell — Mon, 01 Jun 2026 19:39:33 +0000

The first section can be done on a M1 chip, I think the second one needs Triton support, so your 5080 should be fine.

New comment by pell in "How long until AI automates all cognitive labor?"

pell — Thu, 28 May 2026 14:51:27 +0000

I don’t think the definition is that clear cut at all. A human can remember the smell of something and invoke it right then and there even if only for half a second. Are we expecting an AGI to do the same?

Then again, transformers seem super-human in some ways already. Who do you know who can more or less recite and make associations from (even if not always intelligently) hundreds of billions of text fragments? Transformers already are better at math than your average human.

My bet is we’ll land in a weird place in between where these systems clearly have some superhuman intelligent capabilities but still are far from “do everything better than humans”.