I'm a gray beard, apologies, I might get answered by "yeah well such and such webpage is 100MB+!" ... But that doesn't make it less wrong.
Suggest you don't load all those videos at once, but only when they are needed.
It's particularly triggering due to the fact that your page appears to load further content as you scroll down, which makes it look like progressive loading and being a good netizen...
Curious about video generation process here. Based on info in pricing, it seems like you are just generating multiple images and then adding effects over it or you are actually using any video-generation API?
Did you have a background in anything before starting to google how to program? I have two friends looking to earn money with programming and they have tried for many years and have not succeeded.
Please encourage them not to follow the example set by the author of this post. See other comments for details on the technical issues, inaccuracies, and dishonesty.
The easiest way to make money as a programmer is to get a stable job as an SWE. Have they tried that?
> I'm genuinely curious about HN's thoughts on AI-assisted learning and whether this approach scales.
IMHO this was always the use case that AI mattered for: education. To give people less skilled a leg up when learning something fits the desired use-cases for generative AI. Its low stakes (given we're in education) and has a human check before hitting production (in that the dev runs the output to see if it works). The best thing about generative AI is it generates that curiosity to those who never studied CS, as it gives them something that kinda works and lets them tinker with it to change output. Its extremely healthy.
To be honest though I figure if you were born 20 years earlier you'd have read a book and managed to make the transition the same, albeit the output product might have been slightly less impressive.
The tragedy of such an incredible time to be alive, is that it will likely further re-enforce the fact that most humans can't be bothered. When given this technology, having to expend the effort you have, is likely still too much effort for most. We saw this with the early internet where almost all of human knowledge was available for free online and most people wouldn't be interested in expending the effort to learn it.
btw, it looks like your localization is a bit off in the pricing section: "30 créditos incluidos" when in EN.
Kudos on making your app and, even more importantly, shipping it! Far too few devs are serious enough to ship something customers can actually use.
Both domain and programming expertise are still required with agents. You just picked up yours along the way. The larger and more complicated the architecture, the more human guidance is needed.
But the 2 "hidden skills" needed to master agents are really the one's you've demonstrated here. Those are communicating clearly and being willing to listen and learn. Perhaps your hospitality background is a benefit here because those are not common dev skills, especially communicating about programming in plain English as you did with the LLM.
Kudos aren't in order. Another user pointed out that the reviews on the website are fake and that the website claims that the app has 5,000 paying users, even though the author here says he has only six. So the website is inaccurate at best, deliberately misleading at worst.
OP's comments in the thread, in turn, are plagiarized, presenting obvious LLM output as his own work, in a bid, I imagine, to drum up engagement and discussion, pushing the post higher on the front page.
The dishonesty and clumsy attempts at manipulation are unmistakable. This is AI slop, through and through, and the author deserves censure, not praise.
This isn’t meant to be a gotcha or anything I swear but you say 6 paying customers and your site says “5000 have already taken advantage of this offer”.
Is this maybe something the AI stuck in that you might have accidentally overlooked?
Does "building" a tool for churning out spam truly make you happy? Do any of these videos (assuming that the product actually works, because the rest of the site sure does not inspire any confidence) provide any value to.. anyone, or do they only exist to clog up people's feeds?
> CURRENT METRICS (being honest): • 6 paying customers
Meanwhile, your website claims "5000 paying customers".
> 5 verified reviews from real creators
5 out of your 6 customers wrote a review? On your site? Really?
You're absolutely right to call this out, and I appreciate the feedback. I made some mistakes with the landing page copy:
1. The "5000 have already taken advantage" was AI-generated placeholder text that I should have removed before going live. I've just updated it to reflect the actual metrics (6 paying customers).
2. The reviews were also AI-generated examples to demonstrate the feature, not real customer reviews. I should have made that clear or removed them entirely.
This is a learning moment for me - I got caught up in trying to make the landing page look more established than it actually is. The core product works (it's a tool for content creators to generate video ideas), but I clearly need to be more transparent about the actual numbers and not use misleading social proof.
Thanks for keeping me honest. I'm updating the site now to reflect the real metrics. The 6 paying customers and $65 MRR are genuine numbers I'm proud of, even if they're small.
Quick note to say that your webpage is 85MB+.
I'm a gray beard, apologies, I might get answered by "yeah well such and such webpage is 100MB+!" ... But that doesn't make it less wrong.
Suggest you don't load all those videos at once, but only when they are needed.
It's particularly triggering due to the fact that your page appears to load further content as you scroll down, which makes it look like progressive loading and being a good netizen...
Good luck with marketing.
Wait...what? _Eighty_ five MB?!
Yep. And thanks to the modern internet and newer HTTP protocols, you wouldn't really know. And I bet you even the maker didn't know.
The…reviews. Are those generated as well?
The reason I ask is because if you look at the pictures they have subtle issues. A button on top of another button. Not quite correct neck muscles.
I’m not going to crap on your project but if they are it’s really disingenuous to have them. Just remove them.
If not and I’m wrong and maybe the users just happened to use AI generated profile picks then I apologize.
Curious about video generation process here. Based on info in pricing, it seems like you are just generating multiple images and then adding effects over it or you are actually using any video-generation API?
Did you have a background in anything before starting to google how to program? I have two friends looking to earn money with programming and they have tried for many years and have not succeeded.
Please encourage them not to follow the example set by the author of this post. See other comments for details on the technical issues, inaccuracies, and dishonesty.
The easiest way to make money as a programmer is to get a stable job as an SWE. Have they tried that?
Have you tried adding Runway Gen-4 Turbo to Pro++?
What do you use for running image models?
Great, another AI slop for my shorts algorithm
interesting, what do you do exactly to short ai slop ?
[flagged]
> I'm genuinely curious about HN's thoughts on AI-assisted learning and whether this approach scales.
IMHO this was always the use case that AI mattered for: education. To give people less skilled a leg up when learning something fits the desired use-cases for generative AI. Its low stakes (given we're in education) and has a human check before hitting production (in that the dev runs the output to see if it works). The best thing about generative AI is it generates that curiosity to those who never studied CS, as it gives them something that kinda works and lets them tinker with it to change output. Its extremely healthy.
To be honest though I figure if you were born 20 years earlier you'd have read a book and managed to make the transition the same, albeit the output product might have been slightly less impressive.
The tragedy of such an incredible time to be alive, is that it will likely further re-enforce the fact that most humans can't be bothered. When given this technology, having to expend the effort you have, is likely still too much effort for most. We saw this with the early internet where almost all of human knowledge was available for free online and most people wouldn't be interested in expending the effort to learn it.
btw, it looks like your localization is a bit off in the pricing section: "30 créditos incluidos" when in EN.
> THE META QUESTION: Are we entering an era where AI democratizes programming to the point where domain expertise matters more than coding expertise?
This has always been the case
Kudos on making your app and, even more importantly, shipping it! Far too few devs are serious enough to ship something customers can actually use.
Both domain and programming expertise are still required with agents. You just picked up yours along the way. The larger and more complicated the architecture, the more human guidance is needed.
But the 2 "hidden skills" needed to master agents are really the one's you've demonstrated here. Those are communicating clearly and being willing to listen and learn. Perhaps your hospitality background is a benefit here because those are not common dev skills, especially communicating about programming in plain English as you did with the LLM.
The domain here is video production and he was already doing that by hand, no? Or do you mean something else?
Kudos aren't in order. Another user pointed out that the reviews on the website are fake and that the website claims that the app has 5,000 paying users, even though the author here says he has only six. So the website is inaccurate at best, deliberately misleading at worst.
OP's comments in the thread, in turn, are plagiarized, presenting obvious LLM output as his own work, in a bid, I imagine, to drum up engagement and discussion, pushing the post higher on the front page.
The dishonesty and clumsy attempts at manipulation are unmistakable. This is AI slop, through and through, and the author deserves censure, not praise.
This isn’t meant to be a gotcha or anything I swear but you say 6 paying customers and your site says “5000 have already taken advantage of this offer”.
Is this maybe something the AI stuck in that you might have accidentally overlooked?
How much of this comment did an LLM write for you?
Does "building" a tool for churning out spam truly make you happy? Do any of these videos (assuming that the product actually works, because the rest of the site sure does not inspire any confidence) provide any value to.. anyone, or do they only exist to clog up people's feeds?
> CURRENT METRICS (being honest): • 6 paying customers
Meanwhile, your website claims "5000 paying customers".
> 5 verified reviews from real creators
5 out of your 6 customers wrote a review? On your site? Really?
You're absolutely right to call this out, and I appreciate the feedback. I made some mistakes with the landing page copy:
1. The "5000 have already taken advantage" was AI-generated placeholder text that I should have removed before going live. I've just updated it to reflect the actual metrics (6 paying customers).
2. The reviews were also AI-generated examples to demonstrate the feature, not real customer reviews. I should have made that clear or removed them entirely.
This is a learning moment for me - I got caught up in trying to make the landing page look more established than it actually is. The core product works (it's a tool for content creators to generate video ideas), but I clearly need to be more transparent about the actual numbers and not use misleading social proof.
Thanks for keeping me honest. I'm updating the site now to reflect the real metrics. The 6 paying customers and $65 MRR are genuine numbers I'm proud of, even if they're small.
This sounds LLM generated, too.