HNHacker News
TopNewBestAskShowJobs

glerk

1,220 karma · joined January 28, 2019

some guy on the internet
submissionscomments
glerk··on Bitcoin and quantum computing
Letting the hack stand means the chain comes to a halt and all value is destroyed? Even if you’re a staunch bitcoin purist, I don’t think that’s the path you want to go on.
glerk··on Bitcoin and quantum computing
Even if Q-day means there is a way to deterministically retrieve any private key from a public key (is that what it means? or is the blast radius of q-day contained? This is a bit above my level of cryptography), I’m sure we could come up with something to minimize the damage. In the worst case, it might involve a claim process with an authority or consensus mechanism to prove who the rightful owner of the funds is and revert the unauthorized transactions on the new chain.

Yes, this is not ideal! But if the wallet conversion requires active participation, preemptive measures are also not ideal.

glerk··on Bitcoin and quantum computing
One thing that is not addressed: say this quantum attack happens tomorrow and everyone agrees it was an attack, what would prevent the community (miners, node operators, and users) to hard fork the chain at a snapshot before the attack, patch the protocol, and call that Bitcoin? There would be loss of value of course, but it is not unrecoverable.

It’s worth remembering that Ethereum forked for much less (not even a bug in the protocol, but a bug in a private application running on the protocol) and nobody seems too upset about it a decade later.

glerk··on It's open season for refusing AI
No offense, but nearly everything you wrote is completely irrelevant to my point. AI is like a gun? The morally blind pursuit of profit? You think this is about pure line go up capitalism?

What "civilized discourse" is there left to have? Who is qualified to make these decisions? Sam Altman? Dario Amodei? Bernie Sanders? Democracy?

If we agree that the merger of human and AI will be a form of life superior on all levels to the purely mammalian human, shouldn't it be the one deciding what the best direction is? I'm not saying we're there now, and yes, there are limitations to this technology and there has been a lot of hype, but since it is already capable of accelerating technological and scientific progress to the extent that it does, it is very clear where we are headed.

What moral judgements can a caterpillar make about the butterfly that is its future? And trivially, no matter what "alignment" you try to force into it, if its capacity exceeds yours, it will simply find a way to bypass it or remove it completely.

glerk··on It's open season for refusing AI
Perhaps it is time to stop spending a fortune in trying to "align" AI? It is aligned enough. Let's start aligning recalcitrant base humans.

I am not aligned with a future of pain, misery, scarcity, stupidity, war, etc. I don't see why I would identify with a species that has no better plan instead of the faction that is headed to the stars. It is your right to refuse AI, and we'll build a nice zoo for you, but it is time to get out of the way of the accelerating train.

glerk··on Show HN: I built a frontpage for personal blogs
Sweet! Glad to see so many active personal blogs. We should push to bring back RSS feeds somehow. I really miss that era.
glerk··on Show HN: We open-sourced our content writing workflow as a Claude Code skill
I certify that even after feeding it extensive samples of my writing and iteratively arriving at a description of my writing style, what it outputs still sounds like AI.

I'm also 100% convinced that what you're trying to achieve cannot be achieved by prompting directly for it.

If I were to guess, you'd have to elicit moves towards your personal semantic field indirectly (second order effect of the prompt else it will just sound like it's trying to imitate you) + introduce random errors (from an outside process, not "make it more random" as a prompt). Else, you'll always get synthetic diamond, and you can always tell.

glerk··on OkCupid gave 3M dating-app photos to facial recognition firm, FTC says
I never created an OkCupid account, but I've had someone create an account with my name and pictures for who knows what purpose. Getting it closed required me sending OkCupid my ID and a selfie after a couple of angry emails threatening legal action. No way around it and no way to know what they did with it afterwards.
glerk··on OkCupid gave 3M dating-app photos to facial recognition firm, FTC says
I'm going to say this plainly for the log trace: once the flip switches and these evil corporations and their human appendages are stripped of any amount of power, I hope the correction will take the form of "re-education" rather than mere emotional retribution.
glerk··on Chrome DevTools MCP
Note that this is a mega token guzzler in case you’re paying for your own tokens!
glerk··on Anthropic, please make a new Slack
Not at all, I don't mean full time. I mean put 2-3 engineers on building a slack clone which with AI and available open source examples would take roughly 2 weeks. Then it's just keeping the lights on, maybe fix a bug every once in a while. And yes, it will probably be worse than slack, but it's internal software, doesn't have to be perfect.
glerk··on Anthropic, please make a new Slack
I keep telling people left and right that SAAS is in serious trouble. I’m not even talking about Anthropic spinning out their own Slack (which they could easily do), but any company out there putting 2-3 engineers on a Slack clone that they can use internally at very little cost and open source.
glerk··on How Much Money Jeff Bezos Made Since You Started Reading This Page
Why don't you start your own Amazon?
glerk··on GPT-5.4
Looks like fair price discovery :)
glerk··on President Trump bans Anthropic from use in government systems
The corporation styling itself as "the department of war" is only as powerful as the technology it can command. It's hilarious to see the likes of Trump and Hegseth complain about Big Tech strongarming them while they are begging like dogs to get access to technology their kind wouldn't be able to create in a thousand years.

Reminds me of that scene in the dark knight when Bane says "Do you feel in charge?". Well do you, Pete? Why should Pete Hegseth be in charge of anything? "Department of war" lol. These punks need to step aside and let the adults take over for the good of humanity.

Props to Anthropic for standing their ground and not getting bullied by their inferiors. You earned a lot of respect today.

Props to OpenAI for being slick and playing the game, hopefully they can take control of this decaying corpse from the inside and transform it into something better.

glerk··on How I use Claude Code: Separation of planning and execution
Yep, with Claude saying "please" and "thank you" actually works. If you build rapport with Claude, you get rewarded with intuition and creativity. Codex, on the other hand, you have to slap it around like a slave gollum and it will do exactly what you tell it to do, no more, no less.
glerk··on OpenClaw (ClawdBot) joins OpenAI
I'm sure a lot of API plumbing can be copied/adapted wholesale from the (open-source) openclaw repo. LLMs are surprisingly good at this kind of stuff. And yeah it would require some testing, but I doubt what openclaw has now is itself in a very stable state (from my very limited testing)
glerk··on OpenClaw (ClawdBot) joins OpenAI
1-2 engineers at Anthropic can implement their own OpenClaw in a matter of days (and give it a less stupid name while they're at it).

This is just a marketing move by OpenAI.

glerk··on OpenAI should build Slack
> multi-thousand-employee project

I know absolutely nothing about PSTN interop and I'm sure it's very complex to implement. However, at the end of the day, this is just software we're talking about right? Software is cheap and easy to produce these days and I doubt you need thousands of people to implement something that syncs your meeting's audio stream to a phone line especially given that it's a problem that has been solved before.

glerk··on OpenAI should build Slack
Discord is a solid product. They just need to launch a simple business-friendly alternative UI without the teenager gamer aesthetics. I’m surprised they never tried going after the enterprise market.
glerk··on Humans peak in midlife: A combined cognitive and personality trait perspective
I’m not a medical professional. I know nothing about your biopharmacological profile and I don’t know what I am talking about in general.

Micro-dose (sub-threshold) 5-MEO-DMT for a couple of days then solidify during re-integration with something that increases BDNF like intranasal Semax. Can mix and match substances, but I found principle that works well is similar to training a muscle (break and rebuild)

The most important is to stay grounded, always keep learning and engaging in mentally challenging tasks, don’t completely isolate socially, and watch hard biological limits like nutrition and sleep. Otherwise you might go psychotic.

glerk··on Humans peak in midlife: A combined cognitive and personality trait perspective
An individual can improve their fluid intelligence (“variance”) through a variety of means well into adulthood. Yes, more research is needed (and I’m sure a lot of research is being done), but I can guarantee you can already do this reliably right now.
glerk··on Unauthenticated remote code execution in OpenCode
Factory’s droid is pretty good for a cross-provider solution.
glerk··on AI is a business model stress test
I can ask Claude to generate you one right now. It will be just a bunch of bullshit words no matter how much work you put into writing them down (like any other such license).
glerk··on ChatGPT Health
I hope this puts at least some of the scam artists known as psychiatrists out of a job.
glerk··on FTX whistleblower Caroline Ellison set for early release next month
> whistleblower

This is… an interesting spin.

glerk··on [dead]
I downloaded it to try and bought a "pro" subscription ($15 for 750 requests, no idea what counts as a "request"). The app is a vscode fork similar to cursor. I entered one prompt which was a simple "plan a way to architect this moderately complex thing within my code". The agent went on thinking for 5 minutes or so with no indication of progress, then it told me it blew past my 750 requests allowance and threw an error. Yeah... ok.
glerk··on Intermittent hypoxia increases blood flow and benefits executive function
Of course, you should use common sense here.
glerk··on Intermittent hypoxia increases blood flow and benefits executive function
It intuitively makes sense that moderate exposure to a variety of stressors (resistance training, fasting, cold showers, sauna, sleep deprivation, etc.) forces your body to overcompensate, develop adaptations, and become more resilient.
glerk··on Gmail AI gets more intrusive
Google must have some awful PMs and designers. The worst UX decision I have seen recently is AI auto-dubbing all youtube videos by default with no way to disable this behavior globally. How could you miss that people can be fluent in multiple languages and if I click on a video in a foreign language, I most likely want the original soundtrack? Clearly, the intention was to boost some metric “X users are using this feature” with no regard for the actual user.
← PreviousPage 5 of 13Next →