News for hackers. Stuff that matters.

wsgeorge

Re:Anthropeum on Tuesday July 28, 2026 @03:30PM (#49088767)
Honestly surprised by this outcome. My geo estimates were way better than my period estimates.

Anthropeum.com · Jul 28 2026

64,056 · top 55% of players today!

Re:Firefox Containers Preview on Tuesday July 21, 2026 @02:20PM (#48996074)
But Firefox also has browser profiles now (something I wanted for so long I moved back to chrome to get it), so what’s the purpose of first-party containers now? Don’t browser profiles provide a more robust, very visible delineation between different accounts/user profiles?
Re:Internet traffic in Iran increasing on Thursday May 28, 2026 @04:39AM (#48306310)
> (Yes, the invader could try to get Mexico to agree to host the invading army, but the US would probably find out about that plan and respond by either blockading Mexico or invading it if it doesn't immediately abandon the plan.)

Curious, how viable would this strategy be when invading via Canada?

Re:Muse Spark: Scaling Towards Personal Superintelligence on Wednesday April 08, 2026 @08:03PM (#47697735)
Alexandr Wang suggesting this might be open-weights/source in the future gives me hope. Hopefully they stay on this path.
Re:TikTok's 'Addictive Design' Found to Be Illegal in Europe on Friday February 06, 2026 @09:59AM (#46913615)
> a positive reinforcement of not doing it because then the stress goes away and that is nice.

I may be similarly wired, and I've found abandoning Duolingo streaks on my own terms to be very rewarding.

Re:Deno Sandbox on Wednesday February 04, 2026 @05:18AM (#46883972)
What I gather from the announcement: it's part of Deno Deploy (their SaaS offering). I too would love a self-hosted version.
Re:Ask HN: How many screens do you usually work with? on Thursday November 27, 2025 @06:47AM (#46068341)
2 screens, laptop hooked to a monitor. But I usually have a second laptop next to me, so 3 screens.

Then there's my phone (when testing apps), so sometimes I do 4 screens. That's my limit.

I fell in love with elementaryOS when I first learned about it c. 2013, and followed it closely (used to read commits weekly) for perhaps a year. It was one of the first open source projects I really loved. But I eventually went back to Ubuntu because I was newish to Linux and I needed something that Just Worked™, without the funny business.

I eventually fell out of love with elementaryOS when the team seemed to double down hard on some unpopular design decisions wrt window control buttons/behaviour, for instance.

I always felt that they had taken on more than they could chew, and all the good will of their community wasn't going to change that fact. To this day I maintain that the project should have just been Pantheon, the Desktop Environment. The team seems to have strong opinions about UX, and that's where most of that matters.

I'm not the only one who thought of that, and the team's justifications for their decision to roll their own distro never came across as strong.

I've since moved on (to macOS and Ubuntu one the side), but once in a while I browse the official sub for the latest. I've never shaken off that feeling that the really talented founders could have spent their energy more wisely.

Re:Ollama and gguf on Monday August 11, 2025 @05:12PM (#44869519)
There's a reddit thread from a few months ago that sort of explains what people don't like about ollama, that "shadiness" parent references:

https://www.reddit.com/r/LocalLLaMA/comments/1jzocoo/finally...

> If the elite colleges are not comprised of the rich and well connected it beats the entire point of an elite college.

Depends on how you define "elite", and I assume you mean some sort of hereditary or economic-class-based definition. But elite colleges could (and should) still work if they run on competency-based merit. I believe elite talent in as many fields of endeavour should absolutely be catered to.

> The entire notion of "elite" universities is discriminatory.

Well, when you put it that way, many things are discriminatory, for better or worse.

Your experience mirrors mine. I have found LLMs kludgy to wield, but I deeply appreciate the technology, its pace of improvement and its promise.

And I feel myself getting better at not letting an LLM get in my way when doing anything: writing, coding, etc.

Ask HN: Have we reached the acceptance phase of generative AI?

In the immediate aftermath of GPT 3.5 and the LlaMA leak, Github Copilot, there was a lot of denial (negative sentiment, barely concealed rejection etc) about the world we were being ushered into.Then there was a lot of movement by governments and pressure groups to legislate.Post GPT-5, Grok4 et al, have we ridden the crest and gotten to the other side now?

( Read More... | 5 comments | 2 points | on HN )
Re: on Monday April 07, 2025 @07:54AM (#43610248)
I tend to see a lot of criticism of the Trump administration when such topics come up. One "standard" HN tends to uphold is not following every single turn of the news cycle. This is why political discussion isn't as active here as it is on, say, political communities on reddit.

Tell HN: GenAI companies stopped gatekeeping product access

Today OpenAI lets you try ChatGPT even before you logging in. Meta surfaced their models on WhatsApp. You can use a bit of Grok and Perplexity without giving up any personal information.Huge contrast from Q1 of 2023, when LLamA leaked; when you had to go through hoops just to test drive an open model in a chat setting.I'm just glad the market is more open now. :)

( Read More... | 2 comments | 3 points | on HN )
Re:Ask HN: What Are You Working On? (February 2025) on Monday February 24, 2025 @05:08AM (#43157797)
The name of your app reminds me of one I had started working on when I was out of employment: the app was going to send periodic reminders of stuff you've bookmarked/saved on various social media sites.

I was just about working on the Twitter api when Musk bought the company and restricted access. Real bummer. I got employment weeks later.

Re:Once You're Laid Off, You'll Never Be the Same Again on Monday January 27, 2025 @12:12PM (#42843289)
My guess, frontlines.
Re:DeepSeek-R1 on Tuesday January 21, 2025 @08:46AM (#42780018)
You should be able to run smaller models on an M1. I'm testing this in about 10mins

Tell HN: Needing a puzzle to prove you're human to log into ChatGPT is insulting

Short rant, but there's something awful about being blocked after authenticating by a puzzle - not one of those smart CAPTCHAs - that requires you to prove you're human just to log in to one of the most powerful general purpose AI services in existence today.People don't complain enough about these tests, but I expect them to come to the fore as AI becomes more of a thing and users begin to...

( Read More... | 2 comments | 2 points | on HN )
Just from a visual perspective, it's amazing what a colour palette can do
Re:How Couples Meet on Monday March 25, 2024 @07:41PM (#39822556)
I would have expected "Met in bar or restaurant" to be much higher in the 40s, and to have tracked with "Met in church" for longer.

Older »