Bigpet

That's part of it, but if you lived through months of news about where milk is unsafe and how to wash your vegetables and how much iodine to take. You'd maybe think differently too.

All the "oh, but it's different now, it's really safe" implies that the scientist at the time of the Chernobyl disaster didn't give assurance that it's totally safe either.

I am in favor of nuclear power and think that closing the plants was a huge mistake, but it's not somehow fully irrational to opose nuclear power. Not everyone has the hubris of thinking they can evaluate the risks when being shown some data. Nor can they distinguish the difference between "the experts then said it's safe" and "the experts now say it's safe".

Top comment doesn't say it's better. Just says it's a "workhorse".

The article itself doesn't say "it's better", basically just says "in this one specific benchmark it beat Claude with Claude code". Mind you with multimodality it Opus still beat GLM 5.2 very handily in that same benchmark.

I can't find any contradiction and I don't see anyone lying directly. At most they lead you to imply false things, but they're not untrue at a literal reading.

If you did "web stuff" in the early 2000s (like 2005-2010). You'd probably know who he is. He did Ruby on Rails, a backend web framework.

But that was also very Start-up and America focussed. So if you did web dev in some other country and didn't have colleagues who were into that culture you still might've missed the name.

Well maybe we shouldn't then? Before LLMs some not just clueless but also malicious rando could've send a PR too. And the maintainers might've gotten burned out any and just said f-it and merged stuff randomly. I don't see how AI changed the calculation here much.

Disabling ssh alternate channels will make it uncomfortable enough for most people to be a deterrant. But if you require multiple ssh channels for different reasons that won't be a solution for you.

He asked for a steel-man, you were the one to provide the straw-man.

Whoa there, am I missing something, why so aggressive and immediately with the ad homs?

I think by that logic dangerously-skip-permissions and openclaw should've never been a thing. I agree that people use them too liberally, but I think at some point you have to find a balance between systemic safety risks and individual freedom.

Are you being sincere? This is one layer of irony too much for my brain to comprehend.

The person you're replying to is making a joke about OpenAI shutting down Sora their video generation "social media" app recently.

Well it seems like whatever's on that page is blocked by the corporate proxy malware scanner I'm currently behind.

Is that in addition to mainframes or for completely replacing them?

When initially opening the page it had broken colors in light mode. For anyone else encountering it: switch to dark mode and then back to light mode to fix it.

I think the dev that was responsible for the bootloader or some security chip having the option to be opened posted on twitter a while ago. Pretty sure he implied or mentioned that this was what he was hoping for.

Edit: this was what I was remembering: https://x.com/XenoKovah/status/1339914714055368704

Oh, thank for the correction. I must've muddled it up in my mind with the contact tracing integration that had during Covid.

Shouldn't most comodity GPS receivers also be GLONASS compatible (I get that Galileo is more niche and might not be included).

Does the Sensor Apple uses not use GLONASS in Russia? Or is it cheapo Android Phones picking up the tag and then sending GPS coords into cloud?

edit: Nvm, I might be dumb, I guess unless your jamming includes all commodity GNSS it's pretty useless.

I don't even care about that. My suggestion to him was earnest. I don't have a problem with LLMs. Just with how people use them. I just don't like "slop". I see the same user-interaction problems every time.

I just don't want people to litter their heavily polished immaculately styled products that have so clearly bad user-interaction design. E2e testing and closing the loop on LLMs does seem to help here.

Though I really would prefer people click around their own product for at least 5 minutes.

Was this done heavily LLM assisted? Especially the PDF Edit tools have user-interaction quirks and bugs that a human developer would catch immediately during the regular manual testing when developing.

I'd suggest you at least try and mitigate that by having the LLM do extensive e2e testing if you aren't interested in using your own product.

I would not have guessed that iOS allows enough access to APIs to implement anything vpp-based. Very cool to see. I also enjoyed working with vpp (for the brief 6 months that I had with it).

Python is heavily OOP, everything is an object in python for example.

I strongly disagree. How is everything being called an object in any way "heavily OOP"? OOP is not just "I organize my stuff into objects".

You can write OOP code with python but most python code I've seen is not organized around OOP principles.

Don't know if you need to compile, you might want to but I think interpreting might seem reasonable if size/complexity is a concern.

Is the runtime really that large? I know with wasm 2.0 with garbage collection and exceptions is a bit of a beast but wasm 1.0? What's needed (I'm speaking from a place of ignorance here, I haven't implemented a WASM runtime)? Some contiguous memory, a stack machine, IEEE float math and some utf-8 operations. I think you can add some reasonable limitations like only a single module and a handful of available imports relevant to the domain.

I know that feature creep would almost inevitably follow, but if someone cares about minimizing complexity it seems possible.

Could you try to address the obvious point being made instead of trying to obfuscate? There's plenty of bad outcomes besides death. Not all fentanyl users die. If some fent wholesaler were to provide unrefutable evidence that Noone died from his fent you wouldn't go "Oh alright then, nothing wrong with what you're doing".

Very appropriate as well because the machines are given a bunch of feed to digest multiple times and to spew it out the other end as a big steaming pile.

You make it sound like there isn't a very wide range of solutions between "just write html files" and "use a complete website framework". There's a space in there where a large percentage of web projects used to be located in.

You can math that out pretty well. If your code has a breakage chance of 50% and your dependencies all have a breakage chance of 1% then with 70 dependecies you get to 50.5% breakage chance from dependencies.

I don't think you're arguing against anything that was said in that post. There was never an "at all cost". The author was hedging even in the headlines ("indiscriminately", "before you know you need one" and "always, always").

You can already subscribe to projects or single issues/PRs on github and reply via email to post comments.

I can understand not wanting to use GitHub/GitLab/etc. for various reasons. But I don't understand how usability vs mailing lists is one.

How is a set of 9+ mailing lists any better? It has significantly worse discovery and search tools unless you download all the archives. So you're creating a hurdle for people there already.

Then you have people use all kinds of custom formatting in their email clients, so consistent readability is out the window.

People will keep top-posting (TOFU), transforming the inconsistent styles in the process. Creating an unnecessarily complicated problem for your email client to "detect quotes", or you have to keep reminding people.

Enforcing structure of any kind in email lists seems so tedious. I'm not advocating for bugzilla style "file out these 20 nonsensical fields before you can report anything" but some minimal structure enforced by some tooling as opposed to manual moderation seems very helpful to me.

I was looking for the current state of the art of component testing for angular. After reading a bunch of drama around playwright and component testing I stumbled across this unrelated blog post.

I liked the cute web page idea for the app under test. Was a neat diversion. I just liked the blog post and wanted to share.

I too think there is merit in exploring to what degree conciousness can be approximated by or observed in computational systems of any kind. Including neural networks. But I just can't get over how fake and manipulative the framing of "AI welfare" or concern over suffering feels.

It reeks of some navel-gazing self-aggrandizing. I bet not even half the people doing the hand-wringing over how some matrix multiplications might feel are vegan or regularly spare a thought about how their or their companies consumption behavior indirectly leads to real human suffering.

It's just so absurd how narrowly their focus on preventing suffering is. I almost can't imagine a world where their concern isn't coming from a disingenious place.

The solid axle turning without slippage from that video works for regular inter-city rail. I think it doesn't work for rails with the tighter turning radii often required of inner city trams and similar rail transport.

well looking at the list of projects and seeing the site `ZeroAcquire` there it looks like he asked himself the same question and sold the solution