I don't really have a strong opinion on the letter because I agree with a few of the comments that the cat is kind of out of the bag. Even if we don't reach AGI, however you want to define it, I think we are headed for something existential, so I'm both skeptical of letters like this ability to work and sympathetic to the need to focus on the issue it's attempting to address.
So with that said, I find the "what about existing AI problems" and especially the "I don't like Musk comments" annoying. The current issues aren't as severe as what we could be facing in the not too distant future, and Musk is pretty tangential to that concern.
If the price difference is justified, I can't argue with a refurb. Most of my purchases from Dell (mainly laptops, a couple monitors) in the past 10+ years have been refurbs from the outlet. On a sale, about 20-30% discount over new, and they are essentially like-new with full warranties, etc. Never really had a problem except one laptop that was a little wonky at hte start, but got fixed and went on to last 7 years.
If someone is offering me $50 off a $800 device that has been "seller refurbished" , yeah, I'm just gonna buy new.
OpenAI is training on copyrighted data without a licence. I would argue copyright law has much stronger legal standing than some ToS.
Now OpenAI is arguing their training is fair use, but that has certainly not been legally established so far and could just as much be used as a defence against ToS violation.
So in short yes OpenAI is pretty much doing the same thing.
Scratch gets a lot of hate with semi-experienced programmers (the kind to browse this sub) because of its simplicity, but it's actually really good. It teaches you basic programming concepts without all of the complexity of other programming languages.
I remember back in school, some of the students in my computer science classes didn't understand the basic concepts of programming: variables, loops, functions, etc. and programming is 90% logic and problem-solving.
You can't teach programming by teaching the syntax of a language, you have to teach logic. If you know one language, figuring out another is going to be easy because most of the things are the same, the only difference is the syntax.
All: if you ever notice anything like this and have a minute to let us know at [email protected], we'd really appreciate it. I only saw this by change.
What a bummer. Open, anonymous, and free forums are some of the most interesting places on the internet. this profit-motivated antisocial shit makes me so sad.
edit: oh. most if not all of the comments are from reddit.
https://lobste.rs/ has an interesting mechanism to curb bots on the platform.
To join to website, you have to get a referral from another current user. And this referral is public information. So if someone you referred starts acting like an ass, then you can expect some concern being directed your way. People are rightly a bit cautious with giving out referrals as a consequence.
The topics on lobste.rs are more focused on programming and computers (stuff you'd likely also see on HN), and there's not really any political discussion. Traffic is light, expect maybe a dozen new links per day.
This comment, by the way, does not constitute an invitation to ask me for a referral. I don't really know too well anyone on HN, other than the most famous users (none of who know who I am), so if you ask me for a referral, the answer is very likely "no".
I'm curious, how did you figure this out? Do you search texts of random HN comments? Did you remember that comment? I see these posts on here sometimes, and just don't understand how the commenters so consistently find these types of things.
I think it's great- most of the hate probably comes from people who came into their first programming class already knowing how to code to some degree and had to "downgrade" from a traditional language to Scratch to learn along with the class.
I've been there, but there's still some cool stuff you can do when you bring in your outside programming knowledge.
Scratch gets a lot of hate with semi-experienced programmers (the kind to browse this sub) because of its simplicity...
"This sub"? Semi-experienced? Nice way to start a conversation. OK, I'll bite.
My first contact with Scratch was ten years ago when my son wanted to learn it in a workshop organized by Medialab Prado, a group funded by the city council. The wait list was already very long so, in order to cut it, I volunteered as an assistant teacher for another course.
I reviewed my son's assignment and helped him make some modifications after the classes.
I don't hate Scratch. I have a good opinion in general. But it had its shortcomings, that made easy to end up with some sort of visual spaghetti code, as soon as the project grew a little over the size of the examples. IIRC all variables were global.
My son chose a different tool for the next workshop, I don't remember the name (appstudio?), Python for the next and then Python again, but as a teacher. So good for initiation, but my impression was that not so good for bigger programs.
That might have changed, it's been a long time, but if you're curious about where criticism comes from, maybe it's not hate from semi-experienced wannabes :)
Oh and BTW, the guy that was the main teacher in my course defected in a couple of classes, so I had to take over. The children were bored with HTML and I tried introducing JavaScript. Surprisingly they understood it very quickly and liked it. Of course the group had a selection bias, people interested enough in programming to know about the course, etc. but my guess is that with some syntactic sugar and graphic libraries, it could reach a wider audience.
My two cents: every language should make super easy to draw shapes in a canvas and move them. If you need more than ten lines of boilerplate to do that, you shouldn't be designing languages.
My first CS course used Scratch to teach it and I have to say I enjoyed it a lot more than the C++ which followed and I certainly remember more from the experience. C++ was my introduction to pain.
How do you define "consistent"? Consistent with what? Because at the moment it looks like it's using completely other terminology than the VCS I am used to.
> To do so, syntax is based on subcommands. For example, to delete a branch, run gut branch rm rather than git branch -d, same to delete a remote (gut remote rm) and so on.
I looked at a couple comments. They are in fact copying Reddit comments to HN, and it seems to have been a mildly successful strategy for... what exactly I'm not sure.
Sometimes HN moderator dang will notice when his name is mentioned.
Otherwise you might click the “contact” link on the bottom of the front page and send an email to the HN mods. Include a link to the comment in question and and a brief description, give it a subject like “Possible karma farming bot”.
Some software is not worth paying for to use. But other services might.
Look at nextcloud: they wouldn’t have been anywhere near as big if they weren’t open source. Very few people would pay for the product but instead users now provide pull requests and improve the product. Meanwhile they make money now through enterprise support and specific plugins they provide for business.
Probably an unpopular opinion, and I realize I'm kind of ranting on a relatively unrelated subject, but I have become really dissuaded with the Node ecosystems dependence on seemingly boundless dependency trees. The fact that Window's file system can't handle moving project directories (without deleting the node_modules), and relatively simple projects using megabytes of raw text to work... anyways.
While I understand that you don't want to re-invent the wheel, it seems like the this is an important enough part of your project that your own implementation would be the only one without compromises.
> Probably an unpopular opinion... but I have become really dissuaded with the Node ecosystems dependence on seemingly boundless dependency trees.
I wouldn't be quite so dramatic about that; HN as a collective loves complaining about NPM and dependency trees. (At the same time, it loves complaining about NIH syndrome. Although I suppose existent but limited dependency trees are far from an impossibility.)
Yup for sure, 100%. Pulling in a library every time you don't know how to do something is a choice. Only pulling in dependencies that have 10,000 Github stars or are in every react Youtube video without evaluating alternatives is also a choice. I learned to be way more discriminating about npm libraries from a tech lead a few years ago, and to be honest it's one of the best lessons I've learned in a while.
But it is not a viable choice anymore to “not include this useful dependency, because its dependency tree is huge, so I will just rewrite it from scratch”, which is what practically happens in most cases. No one deliberately imports bullshit like leftpad on the root level. If you use react alone it will probably already make enough of a mess that windows’s file operations will take considerable time on your node_modules folder, which is ridiculous in and of itself.
We're saying "think about each dependency you're considering pulling in. Maybe have a quick browse through the code. Is it a gigantic hot mess? Is it tiny and elegant? Does it only have 3 downloads/week on npm? There are lots of things you can do before deciding to rewrite it yourself, but yes, I argue there are definitely some dependencies where that is the right call. But also, YMMV - it depends on your team and resources too.
there are room between huge dep tree and rewriting everything, that's where we should aim
for leftpad, even if I know it's just an example, there's a native String#padStart, and else lodash is pretty small, most mainstream libs have few deps actually
Tolerating an iceberg of bad habits under a surface of abstractions is a way to get up to speed on something fast, but you eventually have to invest time learning better ways to do things. Except in web development where it's normal to send multi-megabyte blobs to the browser.
If you always in include 'vanilla' as a verbatim search term when looking for Node.js tutorials you'll get better results that tend to avoid that problem.
I don’t necessarily disagree but I have to say that in 10 years of working almost daily with sizeable node applications, this hasn’t been a problem for the past 7 or 8 years.
Maybe I shot myself in the foot enough times to have learned what not to do.
Imagine complaining about the idea of living in a society where pretty much every single picture we see reflects an AI's view of what reality "should" be rather than actual reality, and where all pictures of everyone's friends and family is automatically altered to suit the AI's view of what a human "should" look like. The only images of actual reality are those taken by the select few who both care enough and are technically proficient enough to get their cameras to not do automatic AI retouching.
Doesn't sound like such a ridiculous thing to complain about anymore, does it? Do you at least see why some people may be opposed to further steps in that direction, even if you yourself is fine with it?
The whole metaverse thing has felt like a buzzword-without-meaning that was never going to live up to the hype (at least in any meaningful good way), but still sucks for the people losing the jobs.
The metaverse is the ultimate, final instance of the "... but on a computer!" fallacy [1], even moreso than the "virtual worlds" that prompted me to originally write that article. It's reality... but on a computer!
We already have a reality. It's called reality. The entire value proposition of computers is to do things reality couldn't already do. Replicating reality, but poorly, is a complete and utter waste of time.
Or, to put it another way... no, the metaverse isn't happening. Or to put it yet another way, it already happened and the silly science-fiction descended ideas about what it would look like are as silly as the idea that in the future everyone will constantly wear form-fitting jumpsuits.
The metaverse is an actively stupid idea. When the useful bits and pieces are reified over the next couple of decades, I'll still be right, because those things won't be "the metaverse", they'll still be extensions of the real things that are not only happening now, but have already been happening for decades, including yea verily this very site we're communicating on right now, which would not even remotely be improved in any sense whatsoever by being "in the metaverse".
[1]: https://jerf.org/iri/post/2916/ (Rereading that ten years later, it seems education has hardly gotten anywhere. Still BOAC-ville.)
It may finally take off if/when VR is just another feature of portable, broadly popular AR glasses.
Until then, it loses for anything but niche applications, because it's doomed by the popularity of the smartphone. Nobody wants a social network (or whatever) that can't be reached from their smartphone, or that sucks to use on there, which means nobody wants VR as the primary interface to anything, making it rather pointless.
The whole market is spinning its wheels until or unless someone finally manages to get normal people excited about AR glasses, which means some serious hardware advancements. Whatever's the "next big thing" has to be as portable and usable-everywhere as smartphones, or it's doomed. The Web boom among normal folks wasn't because they started sitting at desktop computers way, way more, but because computers got ultra-portable and cellular Internet got cheap enough to actually use. They're not going to clamber to join anything that they can't comfortably and non-dorkily use at a coffee shop or in line at the grocery store or while sitting on the couch watching Netflix or what have you.
I remember on my first day teleporting into a random location around some mansion, and an expensive-looking female avatar yelled at me how dare I spawn at a private property. Later I ran through some bars and shops and never visited this swamp again.
I agree, and one can say “but videogames” however that doesn’t seem to be what any of this is reaching for. People play “realistic” videogames, which has an element of reality in creating a 3D world, but it tries to tell a story that one can escape into away from our everyday lives. Instead the point for the Zuckerberg metaverse is in mirroring reality but with even more control given to the corporate machine to feed you ads and convince you to buy more things. Who wants a hyperreality built by advertisers? That sounds like hell.
It reminds me of the brief fad when some folks really thought VRML or 3D Java or Flash interfaces to websites was going to be The Future not just for games or art project, but general web navigation. Turned out to suck for nearly everything, total dead-end.
Flash did end up influencing the animations of CSS. I'm not sure about the other technologies you mentioned, but I think the delineation here may be that the idea was the future, not the tech.
No, I don't mean 3D animated elements, I mean 3D interfaces.
Think like if you went to Amazon's website and had to navigate product categories by moving around in an FPS-like interface. This was a thing for a while—it never went big, but there was real excitement around it and some effort was put into it, only to find that (obviously) it sucks for anything but games (duh) and maybe art projects of some kinds.
So with that said, I find the "what about existing AI problems" and especially the "I don't like Musk comments" annoying. The current issues aren't as severe as what we could be facing in the not too distant future, and Musk is pretty tangential to that concern.