Hacker News: pllbnk

New comment by pllbnk in "Improving GPT‑5.6 Sol in ChatGPT, expanding GPT‑5.6 Luna access for free users"

pllbnk — Thu, 06 Aug 2026 19:50:16 +0000

The _Intelligence_ part of AGI should be able to guide the user through that without all the knobs.

New comment by pllbnk in "Humans missed 1 in 3 threats approving AI agent commands across 40k game runs"

pllbnk — Thu, 06 Aug 2026 14:30:19 +0000

I think minority do. Imagine, you have been vibe-coding this project for a while and it works kind of fine but you just have to fix a few more bugs and you get something like `node /tmp/claude-1000/-home-user-source-github-user-hn/27b740b1-9a45-47f3-ab99-61e5e3cf779a/scratchpad/hidden-smoke.mjs; echo "exit=$?"`. (I took it from my own agent right now and I don't have any idea what it's doing. Thankfully, it's sandboxed so I don't care _that much_ right now). Is it bad? You can probably go into that mjs file and see what's in there, but so far it's been fine every time, why would it be different this time? Approve!

We will see many disastrous bugs and hacks in the coming years with the way most developers are coding right now.

If you take time to understand _everything_ that an agent is asking of you, then nearly all those advertised productivity gains would be wiped out.

New comment by pllbnk in "BMW Is Showing Commercials on Their Car's Dash Screens as a Treat"

pllbnk — Sat, 01 Aug 2026 18:46:35 +0000

They have to show this full-screen banner about driving safely with the buttons to turn it off disabled for a few seconds. It's annoying. I want to turn it off but I can't so I start driving and I am forced to look at the screen until the button to turn it off gets enabled. That's very unsafe. Have I read the text that's written there during 4 years I drive the car? Never.

New comment by pllbnk in "OpenAI's work on Git for large repositories"

pllbnk — Sat, 01 Aug 2026 10:04:13 +0000

Just give your favorite LLM to read it, maybe it will enjoy. It's not meant for humans anyway.

New comment by pllbnk in "LearnVector – Andrew Ng's AI company building one‑to‑one learning experiences"

pllbnk — Wed, 29 Jul 2026 08:20:34 +0000

Qwen 3.6 27B will do that for you, or any other model with a few billion parameters. I don't see any value there other than nice UI on top of an LLM.

New comment by pllbnk in "Passkeys were invented by engineers with zero understanding of consumer brain"

pllbnk — Wed, 22 Jul 2026 17:21:25 +0000

Also, where do I store the backup Yubikey, or any other pass key owning device for that matter? It has to be easily accessible to set up the pass keys, but also safe from accidents like house fire.

There's a strange tension where I want to use pass keys because they are easy to use but also they are easy to lose, so I choose a KeePass synced over cloud and deal with a bit of a hassle by having to copy/paste my passwords.

New comment by pllbnk in "Who's afraid of Chinese models?"

pllbnk — Tue, 21 Jul 2026 09:53:33 +0000

2 parties isn't much more than 1. And their definitions of truth are almost identical when it comes to capitalism anyway.

New comment by pllbnk in "The US grocery slowdown is real"

pllbnk — Sat, 18 Jul 2026 09:18:32 +0000

I'm in Europe. I have been tracking my all expenses to the cent religiously since 2010 in Excel sheets. I also haven't changed my grocery habits much since around 2020, except very recently when I actually started looking at the deals (discounts) and buying more generic stuff. We haven't been eating out much (~5 times per month on average) Obviously, it varies month-by-month but my family's (2 adults, not counting the child) expenses on food have risen by around 100% since 2022.

New comment by pllbnk in "42% of adults rely on their parents for financial support"

pllbnk — Fri, 17 Jul 2026 19:08:57 +0000

Thanks for the perspective. I have intuitively thought adults live at their parents' mostly out of need and much less often out of choice. The dynamics change in the latter case as two independent adults can make an agreement on how they see their relationship.

New comment by pllbnk in "42% of adults rely on their parents for financial support"

pllbnk — Thu, 16 Jul 2026 19:25:35 +0000

I understand that such a game of life might be beneficial for a child. But I would like to understand the reasoning to do that with an adult.

> I know other fathers that charge their adult children rent, keep the money, and then gift it to them at the wedding or to help them buy a house.

This looks like a game which should have ended by the time a child is an adult.

I am raising a very young child so I am very concerned about the future and constantly thinking what's the right way to raise him.

New comment by pllbnk in "42% of adults rely on their parents for financial support"

pllbnk — Thu, 16 Jul 2026 19:11:12 +0000

I can’t understand morally or financially how can parents charge anything of their children. Not judging it, just honestly can’t understand it. Isn’t it the parents’ responsibility to build the foundation for their children and keep building it until it’s there? I am viewing it from a moral angle where it’s the parents’ selfish choice to have a child so they must take full responsibility until a child is or feels that they can be independent.

New comment by pllbnk in "Leak of San Francisco Police Drone Footage Exposes Reality of Urban Surveillance"

pllbnk — Mon, 13 Jul 2026 18:12:53 +0000

I think it's an important distinction whether crime stops happening after all criminals are locked up or after preventative measures are taken to prevent criminal activity happening in the first place.

When criminals are locked up, they will get released and many of them don't have anything else to do other than returning to do crimes, with gained experience from the previous failures.

But it's possible to reduce crime by preventing it. Do not let dangerous districts form, build communities, reduce access to weapons and so on.

New comment by pllbnk in "Precursor"

pllbnk — Mon, 13 Jul 2026 15:31:46 +0000

I have been noticing a lot of Cloudflare false positives where it keeps spinning on my sessions never actually redirecting me to the underlying page. If they keep just vibe coding and releasing a new solution every day, I am afraid it will be reflected in their services quality.

New comment by pllbnk in "The console wars have been lost"

pllbnk — Mon, 13 Jul 2026 11:21:14 +0000

Exactly. AAA games are not needed to have fun. I and my wife are having a lot of fun playing Dr. Mario or Street Fighter (the original one) on our 4K TV.

New comment by pllbnk in "We charge $10k a week to delete AI-generated code"

pllbnk — Wed, 08 Jul 2026 07:12:29 +0000

They can request an agent to reduce the amount of code, right? Why do it manually?

New comment by pllbnk in "Mark Zuckerberg's biggest legal nightmare yet could cost Meta $1.4T"

pllbnk — Tue, 07 Jul 2026 15:58:08 +0000

Meta is a private organization subject to the laws of the countries it operates in. Concentration camps are government entities, legalized by the country. Even then it's not always the choice of a person whether to comply the orders. But let's not go into extremes, they don't help the discussion.

New comment by pllbnk in "Mark Zuckerberg's biggest legal nightmare yet could cost Meta $1.4T"

pllbnk — Tue, 07 Jul 2026 15:39:55 +0000

We shouldn't blame people who work in companies making harmful decisions. Employees are not in a position of strength and should not feel responsible for choosing employment that helps them maximize their wealth. It's the government's responsibility to control the employers.

New comment by pllbnk in "As downtown Seattle offices empty, city facing years of 'zombie' towers"

pllbnk — Sat, 04 Jul 2026 19:45:10 +0000

What's the point of the "recovery" in terms of stuffing people back to the offices when they can successfully perform their work from home?

New comment by pllbnk in "The short leash AI coding method for beating Fable"

pllbnk — Thu, 02 Jul 2026 21:48:21 +0000

I think that Anthropic is gaslighting us with their new model releases. Specifically, I think they have some good base model and are just fine-tuning it until they achieve desired outcome, or the desired outcome is achieved accidentally as part of fine-tuning. My theory is based on the fact that as a long-term (if you can call it that way) Claude user I keep noticing the same patterns it outputs. It's not trivial but certainly possible to see when something has been written by Claude because it has a different style than GPT.

However they have quite good harness in their backend which is the actual model.

New comment by pllbnk in "Mullvad founder gave millions to extremist far right party"

pllbnk — Fri, 26 Jun 2026 21:10:28 +0000

Legal immigration to the US for an every day person is nearly impossible. You can’t just decide one day that you will move to the US and try to make a life there. European countries are orders of magnitude more open than the US.