I think you're right, and we've been training ourselves for the shortest path of least resistance for quite a while while kidding ourselves that we're able to somehow preserve the result as worthwhile. Before I ever tried using LLMs for search I had already been training myself for years to offload my thinking to a quick web search and a skim of some plausible result served up.
Arguably things like Cliff's Notes and "for dummies" books were sating this impulse before that, I wouldn't be surprised if the line can be traced back further.
Since dropping LLMs altogether I've made the connection to the impulse for a superficial jump to the conclusion just by noticing how I use web search, or feverishly scan some text to get to the point.
Theoretically I studied critical thinking and have learned decades ago how to recognize when this is OK and when it's a superficial placation of some addiction to closing the book on a subject or micro-subject... but boy whatever good instincts I might have developed back then have significantly eroded over time. Trying to reclaim that again is pretty humbling.
I guess this is a boring and predictable thing to say, but I miss seeing a project like this and being excited about it. Instead I looked at the codebase, see that it seems LLM-generated and assume it hasn't had much thought or care put into it, so I'm not really motivated to even try it out. Sorry. :-/
Does it do what it says on the tin? If so, does it matter that it's made with an LLM?
In my own experience, LLMs+TUIs are surprisingly effective. They deal with the tedium of ncurses development much more than I can. It took an LLM about an hour to put a TUI over recutils so that I can monitor task progress. I doubt I could have done the same that shortly. My point is, 'that it works' is much more important to me than how it works.
I'll come out and say it, heavily AI-assisted. First rust project, and wanted to dive on in with something that I'd been wanting to make. Lots of iteration, almost entirely free opencode tier models and agentic edits. I know what behavior I'm looking for though, and rust is pretty good at finding obvious errors (clippy too). It's not a production app, and it's just for fun, so ya. :)
Perhaps one needs to go a level up: if making the program is now semi-trivial, it remains to be proven if using the program to actually perform is. If I have a Star Trek replicator that can reproduce a perfect Stradivarius for me, is it any less impressive when I play a beautiful song on it?
If we want to get excited about a mixing board... we probably need to see someone using it in ways that traditional mixers don't facilitate.
exactly this. Id think the main thing for this would be to get midi controllers working. its a digital tool for digital djs. looks a bit niche ofc if you look at 'the market' etc. but as OP said for newer thigs like algorave and just for fun it looks pretty cool :D.
id love this workin on a phone with a little usb-c midi controller attached!
> assume it hasn't had much thought or care put into it
> I'm not really motivated to even try it out
That's a little uncharitable. I get where it's coming from, because I also still feel a tinge of (hypocritical) disappointment when I realize things are built with LLMs, but I truly think we need to start moving past this. It's not a giant corporation trying to scam you out of your hard earned money by passing slop off as a lovingly crafted product: it's somebody's pet project, somebody who clearly had a fun "what if" idea and was able to materialize it over time, with a TBD amount of care/iteration/refinement. I think it's something worth engaging with. Worst case scenario it sucks, like many cheapo projects do -- best case scenario it's awesome and/or it sparks a new interesting idea in somebody else.
I won't harp on you too much for submitting yet another vibe-coded project. What I like about this is that the source code is fairly readable and small. Have the LLM walk you through it piece-by-piece if you haven't already.
> HERTZ is a web-based Digital Audio Workstation (DAW) utility designed for instant vocal polishing.
I appreciate you have aspirations for the project beyond its current form, but this isn't an accurate description of what it is now. It:
- Uses wavesurfer + the wavesurfer regions plugin to display a waveform and allow a subsection of the uploaded audio to be selected
- Loads the uploaded audio into a buffer with the web audio API's decodeAudioData and then cuts the selected region out of it
- Applies a compression to the selection with the web audio API using hard-coded values (based on a screenshot you gave the LLM apparently according to the comment in the source)
- Normalizes the audio with the web audio API
- Encodes the audio as a WAV and then feeds that to ffmpeg to convert it to an mp3 for download
It seems complete for your use case, to me. You might consider asking the LLM to reword the readme and present it as a tool that applies these compression and normalization settings to a section of an audio file in the browser:
It doesn't make a lot of sense as the basis for a web-based DAW right now. I like that you tried to keep it simple and small, though. Even so, you could improve readability by making this all a single HTML file since it's so small. As a learning exercise, you could try doing that yourself without the LLM. It would be a good way to get to know your project.
The irony was that Sony wanted all those CDs on streaming, so they sent them to me to rip because they didn't have any other proper digital copy for ingestion at the time. (It wasn't just the rootkit ones, either, they tried a bunch of other really ugly methods across their catalogue to stop the CDs being useable in a CD-ROM drive)
"To begin, it is useful to remember that all iPods play music that is free of any DRM and encoded in “open” licensable formats such as MP3 and AAC. iPod users can and do acquire their music from many sources, including CDs they own. Music on CDs can be easily imported into the freely-downloadable iTunes jukebox software which runs on both Macs and Windows PCs, and is automatically encoded into the open AAC or MP3 formats without any DRM. This music can be played on iPods or any other music players that play these open formats."
And this part might be interesting in the context of the article:
"The third alternative is to abolish DRMs entirely. Imagine a world where every online store sells DRM-free music encoded in open licensable formats. In such a world, any player can play music purchased from any store, and any store can sell music which is playable on all players. This is clearly the best alternative for consumers, and Apple would embrace it in a heartbeat. If the big four music companies would license Apple their music without the requirement that it be protected with a DRM, we would switch to selling only DRM-free music on our iTunes store. Every iPod ever made will play this DRM-free music.
Why would the big four music companies agree to let Apple and others distribute their music without using DRM systems to protect it? The simplest answer is because DRMs haven’t worked, and may never work, to halt music piracy."
I'd say Steve was actually wrong here. The move to streaming (which Apple had a big part in pushing, directly or indirectly), is actually what halted music privacy. Giving access to all music (I remember the big deal Apple made when they were finally able to sell The Beatles on iTunes), and making it easier than privacy is what stopped it.
He's right that DRM alone didn't stop it, but at the same time you can't really say it didn't help. If we could just download Spotify playlists and share DRM free files with out friends, there would be no reason to subscribe to Spotify on an ongoing basis.
Hollywood still hasn't figured that out, but given how little artists get paid, maybe that's what they haven't push for the same.
Still why I like DVDs. Smaller in size, easy to rip, and 480p is good enough that I don't mind the quality loss. Blu-ray is great, but if I'm buying something I'm not sure about or just want to have because it's worth having, DVD all the way
While that's true, it glosses over the battle to make DVDs convenient. Hollywood did not want them to be convenient, they wanted them region-locked and unrippable, and spent a fortune prosecuting anyone who thought otherwise:
On a small display that I usually want to watch digital movies on, it's fine. 720p is the minimum for anything I actually want to watch/enjoy watching. Like I said, a lot of stuff I have on DVD is stuff that is good to have that I'll probably never watch regularly
I like 4K UHD HDR Blu Rays very much and have a big TV to take advantage of them but I agree with gp that 480p is good enough in the sense that a good movie will still be enjoyable in 480p. And if you are engrossed in what you are watching you won't even notice the reduced detail. There are some DVDs with atrocious encode quality with a much lower effective resolution due to low bitrate (i.e. multiple full length flicks squeezed onto one DVD) or unfortunate processing (NTSC master -> PAL DVD release or the inverse is to be avoided) but that's thankfully rare.
Now, 480i is something I'd rather leave behind but even that is a lesser concern than the content of the film.
I recently started collecting Blu-ray because of thrift and second hand stores. In a few cases, I was able to purchase some of my all time favorite movies unopened in the original retail packaging for a dollar. Not to mention my local library has a larger Blu-ray selection than my local video rental place did before they closed.
Exactly. I mean sure, people were definitely pirating music. But lots of people are own huge collections of CDs, and you could also just borrow other people's CDs to rip them. We were kids without money, but older folks at the time did spend money on CDs.
> and you could also just borrow other people's CDs to rip them
Which is "piracy" - not that that makes it ethically wrong. It's actually the main kind of copying that is targeted by DRM since users of the LimeWire kind never see that.
Which is why we should never use the word "piracy." Don't let the industry dictate language (for their own benefit). Equating sharing music with a friend to robbery and murder on the high seas is a wildly out of touch exaggeration. If we let booksellers dictate language in the same way, they'd call libraries and book clubs organized crime.
Agreed, I didn't put the word in quotes on accident. I'm just quite happy to point out that copyright infringement is something that every day people do without thinking and not restricted to the realm of hardened cyber criminals aka. nerds with an internet connection.
I still borrow CDs to rip, lol. Half my digital music library comes from my library having a way better library of music than books (at least for my taste)
Yeah, GP is rose-tinting piracy and Apple’s stance a bit…
When I was a teenager we had _dial-up_. My first 2 iPods were strictly playing ripped CDs, which I, friends, or family had bought. Buying the iPod itself was probably cheaper than 2 months worth of internet traffic back then.
It took about 30 minutes to get 3.5 MB mp3 song over dial up. I would let downloads run overnight and could get a dozen songs by next day. I would do my bulk mp3 downloading in the morning before class at schools library to immediately transfer over USB to my 4 GB Archos Jukebox hard drive mp3 player. That was around 2001 - 2003 either before schools blocked p2p or I must have been using public file downloading sites.
Hmm, with unmetered traffic? I played DarkGalaxy one month at home and we got a ~95€ bill, my dad wasn’t amused. Literally just DG’s website once a day.
I don't remember worrying about data usage. I got into wardriving and was torrenting from neighbors unsecured linksys ssids. Will never forget that warm feeling of hearing netstumbler pings picking up new APs.
I mean, sure, but at some point with 3 or 4 thousand cds crated up, it became a lot easier to steal than go crate digging in my own basement. And then when what.cd happened and you could literally grab a torrent of perfectly curated files of an artist’s whole catalog, the laziness really spiked.
Unfortunately, that kind of hyper-atomism isn't really an effective argument if you don't already agree with the premise being presented. To make matters worse, the law is a collection of postulates. It can give itself the predetermination of any high-level conclusion it wants, and no amount of reasoning or appeal to lower principles will ever matter.
In otherwords, it's theft if the law says it is. Simple as that.
Only if you operate under the misguided assumption that the law cannot be wrong. I care less for legalese than morals. Its plainly not theft, regardless of the hand wringing of bureaucrats and gatekeepers.
What? The context of the conversation was about piracy which is a distinct legal fiction, it's inherently grounded in the law's opinion, so yes necessarily the law cannot be wrong in this context, that's incoherent. An idiosyncratic disagreement with the law isn't saying anything relevant to the discussion.
It's the end of history. What could we possibly discover about a series of technologies that are already nearly 100 years old now?
I've heard people say that various things are "solved" now because of LLMs too -- programming included. This implies we've "solved" thinking. I'm worried about these sentiments.
The original recordings sound much better and more interesting to me. Way better. The AI generated versions sound slicker in some sense but... like the re-recorded versions of old hit songs from the 60s you hear at the grocery store sometimes. Technically the song is still there, but it blends in with the rest of the muzak.
I'm sorry to be so negative, it's great you're returning to the material after all these years, but the AI versions I've listened to all have the same smoothed-over quality that loses everything interesting and relatable to my ears in the original versions.
I think I can put the finger on how it differs and why the AI version seems more "generic." I think the difference is in the "post production" that became more common in pop music in the 2010's. I'm not at all a musician so I don't know the right terms for it, but the AI version has a bunch more of little "flairs" and auto-tuning and audio tweaks which I assume is put in during post-production that makes it sound "slick."
However, it seems to me that all pop music uses the same post-production tricks these days, and so it all sounds somewhat formulaic even if the individual songs themselves are very different. As such the original version may sound more interesting simply by being different.
So the 2001 version sounds more like an "MTV Unplugged" performance whereas the AI version sounds more like the professional and polished version that gets released commercially.
Each has their allure, however. I suspect the AI version will do better with younger crowds.
I'm glad to hear people share this sentiment. I think the human quality will be part of what people will start to enjoy more about music now that AI has improved so much in terms of it's ability to produce highly polished songs.
This is my hope as well but I'm doubtful. I'd expect it's more likely that the generations growing up with this stuff will enjoy being able to annoy their elders with it and fully embrace the human-less aesthetic.
I get it. Recorded music is a lot of fun to play and listen to. This still was a fun project for us to work on together after 25-years and with both of us living on opposite ends of the country. Hopefully our next release will be an in-person recording!
I gotta brag on Chuck a bit more though. I do love the depth to his lyrics and the way he puts them to melodies. I think his stuff, at the core, is excellent. Can't wait to work on some new stuff!
My girlfriend has been using an AI music generator and like, her music is the best generative music I've listened to. Her prompts are crazy, lots of actual like, music theory and specific requests.
That said, it still shares these qualities. It sounds like AI images look. Oversmoothed/overtuned.
Like as a creative myself, what AI has taught me more than anything is that a lot of what we call artistry happens in the friction between an artist and their tools, when they're pushing to the limit of what they know how to and what their tools are able to do, that's where the great shit is. And it's not even that AI is a bad tool necessarily, it just doesn't create any friction. It can be frustrating, for sure, when you just can't get it to do what you want, but it's not an interesting frustration like friction because you aren't the problem, nor your skill: the tool is the problem.
That probably sounds like I'm saying it's bad and again, no. I'm saying the friction between the artist and their tools is the interesting part, and because there's very little friction, meaningfully, between artists and AI tools, it just comes out... boring.
The push for generative AI is "you can make anything," and that's true. But if you can make anything, then by necessity, anything you make is unimpressive. It's only really impressive if you're not sure if you can make it.
The charm is that as humans it's physically impossible to do the same thing exactly the say way twice.
If I play 3 notes, without something like quantization it's always going to be a bit different from you playing the same 3.
Anything AI with vocals sounds completely non sensical. I guess you can feed it words, but at that point just record yourself. I'm not all that good , but that's the point.
Like your thoughts! And then mastery is being able to play things exactly the same. But with mastery comes another way for humans to recognize each other. Specific sounds to specific masters (or movements, techniques for say sports). Each master adds individual grace that we can recognize. That might be why we discount AI-creations vs human mastery. There is a grace (I’m stretching) but it is not unique and not acquired via a process of human mastery. (The word ‘bland’ fits very well!)
I bought this Barenboim on Beethoven DVD set 20 yrs ago only for the bonus DVD where he gives a masterclass. These were exceptional students (Lang Lang is rated as one of the worlds best now), and I can’t play piano, but in that video he shows me as a layman the difference between playing with competence and playing with grace. One can hear the difference in depth of the story behind each note. Found one: [1]
> Effective teaching isn't just about answering a child's question quickly, rather making the right move at the right moment. AI is also going to be an integral part shaping how this generation of kids learn to read and think, tackling this responsibly means getting the design right.
Can you elaborate on what the experience is like for the child? How does this system help them learn? The article focuses on optimizing for interactivity and engagement, but doesn't discuss how this system challenges or facilitates learning and why AI needs to be the solution.
Elizabeth here, co-founder (and clinical child psychologist). Fair question. Catalin's post was about the engineering (he is my co-founder and our CTO), so the learning side got short shrift :).
Here's what it actually looks like for a child. Say a 6-year-old is reading a story out loud (I will use a reading example here). The tutor is listening to every word. When she stumbles on "chick," it doesn't just tell her the word; it decides, based on her history, whether to break it into sounds, point back to a pattern she's seen before, or let her wrestle with it a moment longer because she's close. If she misses the same pattern twice, that digraph shows up woven into her next story. If she reads fluently but can't tell the character what happened in the comprhension conversation after the story, she gets another text to work on comprehension instead of just pushing harder words. The instructional approach isn't novel or new, it's what a good teacher does, grounded in the science of learning. We run evals on the interactions and real subject matter experts are grading and annotating the behavior. What's new is doing it responsively, for one specific child, on every turn.
On engagement: I'd push back a little on the framing that engagement and learning are separate things (anyone on our team will tell you this is a drum I have beaten for years). A disengaged child learns nothing, no matter how good the pedagogy is. But we're not optimizing for time-on-screen. The lessons and sessions are bounded. The engagement work exists so the child stays in the productive struggle zone long enough for the teaching to happen.
Why AI: it's not that AI "needs" to be the solution. In fact, a great human tutor is better, full stop, but it has never scaled. A classroom teacher with 25+ kids teaches to the middle. This is the first technology that can make real-time, child-specific teaching decisions, which is what tutoring actually is.
More on the pedagogy here if you're curious: https://www.ello.com/our-teaching-approach)
> I'd push back a little on the framing that engagement and learning are separate things (anyone on our team will tell you this is a drum I have beaten for years). A disengaged child learns nothing, no matter how good the pedagogy is.
Engagement and learning are definitely separate things; you're right that engagement is required, but that's only part of it. This is a classic case of 'necessary but not sufficient'.
A disengaged child is not learning; an engaged child might be.
Thank you -- your post was downvoted into being hidden for a while, sorry I didn't see it earlier. This context is helpful, the example of the system delivering reading challenges that match their struggles makes a lot of intuitive sense to me.
I really appreciate that (it seems to me) your goal is not to replace human tutors, but to raise the general baseline. You emphasize scaling, how does that work in practice if you're trying to target audiences who may not have access to devices that can run your program? What is your plan from the perspective of funding and resources to scale infrastructure as needed to support these audiences?
Edit: I also think of other learning systems like duolingo and the application of tablets and computers in schools which begin from good places, but I'm curious if you are studying these alternatives and what you have learned from them?
I really think your goals are great, and if you're starting your design of this system from research about effective learning methodologies and working backward from there rather than starting from AI and working backward from there that erodes a lot of my personal skepticism about a project like this. I hope you find a way to make this work.
taught mental math (abacus) to 3-15yr olds for 10+ yrs. A great teacher notices the gaps in the fundamentals and fixes those along with variants before piling on.
Thank you. The segment showing a child reading text on the screen which highlighted a word they had difficulty with seems like it could be a useful learning interaction. How does your system follow up in that case? Have you studied this type of interaction?
That's the only moment in the video that gave me a sense of what it might be like for a child using this system.
In the blog post you say:
> Imagine a custom story about dragons this week, ice princesses the next — woven with the letter blends your child needs to practice right now.
Have you considered using an automated orchestration system to deliver literature that already exists? This example seems like an opportunity to introduce children to really thoughtful literature like Astrid Lindgren's Pippi Longstocking stories but I'm deeply skeptical that generating the stories with an LLM would inspire a similar experience.
Are there other examples of your platform from the perspective of a child using it? I think those are both interesting cases: 1) interactive feedback on a subject they are making an effort toward mastering, and 2) trying to deliver information when it seems relevant. I'd like to know more about how you are approaching these things and other aspects of the learning process.
Totally agree that we should expose kids to the great classics and I love Pipi Langstrumpf as I grew up to know her. The challenge is often that these books aren’t the best to learn to read with, because they don’t have decodable words. So you’ll want a mix: windows into the works of great literature appropriately scaffolded and ways to explore your own curiosity. But see Elizabeth’s reply below for much better depth.
There's no way to formulate this as a set of rules or best practices, each decision has to be made in a universe of context, but I appreciated this discussion.
Something I like to do is funnel the desire to build for something I don't need right now into TODOs and even FIXMEs. Sometimes they just get deleted, but other times I'll come across an old TODO during refactoring and be able to pay down a bit of accumulated technical debt.
Arguably things like Cliff's Notes and "for dummies" books were sating this impulse before that, I wouldn't be surprised if the line can be traced back further.
Since dropping LLMs altogether I've made the connection to the impulse for a superficial jump to the conclusion just by noticing how I use web search, or feverishly scan some text to get to the point.
Theoretically I studied critical thinking and have learned decades ago how to recognize when this is OK and when it's a superficial placation of some addiction to closing the book on a subject or micro-subject... but boy whatever good instincts I might have developed back then have significantly eroded over time. Trying to reclaim that again is pretty humbling.