I'm a non-tech worker in a non-tech industry, let me state two things:
- Software today is written to cover as many use cases with as many features to target as many users a possible.
- End users very often only use a tiny slice of the program's capabilities, but still pay for the entire program.
This creates a situation where the people writing software see it as a monumental undertaking to get good functional programs (it is), and end users see programs as having annoying learning curves with lots of bloat and "unnecessary" features.
LLMs do an excellent job of fixing this for end users because it allows them to easily create a program that does the handful of tasks that they normally need to use MegaSoftware for. And it's tailor made exactly for the use case. And the LLM can tell you exactly how to use it.
I can give a brief example where I used gemini to create a CAD file transposition tool that utilized a simple GUI tailor made for the files my company works with. This allowed us to forgo a (very) expensive CAD software package to work through converting our archive of files. A probably 2M LOC program could be skipped because we only needed 3k LOC functionality.
I really cannot stress enough how often this is the case, and why SWEs see LLMs as weak tools while end users see them as gods.
There will still be a need for huge software packages in the future, but I know I never again have to pay for a huge class of "here is a large solution space that covers your small scope problem" software.
To bring it home, loveable understands this, an sees that the futures has lots of non-tech people "writing" software. Standard IDEs are not the tools your mom will use to make a "Friends and family birthday reminder" app.
End users rarely pay for the program. Someone in their management chain OKs the purchase, or there's a larger purchase with a cross-charge to the department for the license cost. the problem comes when software needs to meet every whim of the decision maker, when really the users only will ever use 20% at best.
I think you may be assuming a certain enterprise size and accompanying workflow, probably would need stats to actually know how much of software is bought in this way however, and if the other way described would open up possible purchases by smaller companies, as was claimed.
So true. Some of my favorite enterprise software I use often I could never afford or would never purchase for my own use. I've taken jobs because of this.
Certainly you must have enough detailed knowledge of CAD files to validate the output of the transposition tool you had AI create for you. This might not be enough for you to think of yourself as ”technical” but I’d argue that it’s far above the level of “entry level employee using CAD”.
This does also seem to fit the paradigm of “AI is a productivity booster for people who already know how to do x”
I agree with you, and I think the Jevons paradox will eventually manifest itself once again. How many smaller companies are stuck with outdated workflows and tools because they can't afford to pay ten engineers for months on end for something better?
Now those companies may very well be able to afford one engineer and some AI subscription to do the equivalent work.
I don't deny that there's utility in what you are describing: if you can make it work for you that's fantastic. However:
- if you can ship software like that given the current state of the technology, you are probably not the average non-tech worker in a non-tech industry. There are people paying exorbitant consulting rates for dashboards in PowerBI. LLMs in mid 2025 are orders of magnitude more operationally complex than anything most people have seen.
- "citizen developers" doing something to scratch their own itches sounds very much like how a professional software project starts. Suddenly the scope grows and you need a nerd to handle it. Then two. Then four. You get the idea. Maybe that won't be the case for your specific needs, but that's how it generally goes.
Weak or strong is a matter of framing, but that's why I see them as tools and not gods.
I don't disagree with the thrust, but I've recently cleaned some of those up.
One example: LLMs aren't smart enough to do things like properly manage zip codes with leading zeros. It was round tripping strings through an integer representation and corrupting them. The users did notice, but did not have the vocabulary/concepts to explain. To them, sometimes zipcodes get corrupted because inscrutable reasons (tm).
chatgpt also authored a bash script that would have blown away a chunk of my drive if any paths had a space in them. :shrug:
Most spreadsheet software does that because the zips are parsed as integers, not strings . Prefix numerical fields that may contain leading zeros with a non-numerical character/word and save yourself the pain of late discovery! You can just find and replace after it goes in successfully.
Spoken as someone that experienced the same pain years ago in local SEO data that made it way too far before discovery...
I think there was a consensus over the past decade, but we are now having to adjust our priors. The answer is changing month by month. It also means people are delaying their decision because they are afraid of the wrong solution right now except in the most obvious cases.
I don't get how this company makes money. Everyone who uses these tools is just building prototypes. They get to 60-70% functionality, but when it comes to actually productionize and launch, 99% of these projects will get abandoned. The churn has to be absurdly high, maybe people are just forgetting to cancel their subscriptions. Or they're just marketing very hard and getting new sign ups/subscriptions, but will crash and burn soon enough.
I noticed in the example product the person was bragging they made "$1.6k in 5 weeks" but looking at the screenshot they were selling "lifetime access" built on this product they pay $25 a month for. I think there is going to be a reckoning for all these fly-by-night SaaSlings and the poor people who start to rely on them.
> Everyone who uses these tools is just building prototypes.
But also I think that's kinda the point. Like for example we have to do a UI redesign of an app my company is building and our "wireframes" are just v0 projects my PO created in one afternoon.
I think wireframing is where these tools really shine. It's gives you roughly the same ideas as building a wireframe in Figma but it's way less work and you end up with higher fidelity wireframes.Like sure if I were to peek at the code under the hood I'm certain it's close to dogshit but the code doesn't really matter at that stage.
I haven't tried with lovable, but gemini studio is fully capable of following a styleguide based on a screenshot, which is probably the crudest and least refined protocol I could think of.
I'm sure there's already integrations somewhere that could allow a designer to specify a bunch of brand colours, to generate a styleguide out of it in plain english, and get a working prototype out of it.
A design system is not "following a styleguide based on a screenshot". A design system is an ontology of UI elements that your software has to adhere to, with several rules about how those elements are or aren't allowed to interact.
Following a styleguide will make your software break as soon as the design system gets updated.
I think you have an idealistic view of what design systems are in practice once filtered down to the developer (or AI) who is tasked with implementing them. You consider yourself lucky if the devs use the right fonts, icons, and colors.
Behaviors will be whatever component library they're using does.
I guess you could argue its just marketing, but if that's the case they're not building "the last software" as they claim. In fact they become entirely reliant on people building more software.
I don't think the idea sticks with me. The final products of these services are never reliable.
Right now, except for some hyperscalers, any similar service has integrated their deployments to some hyperscaler which means any day 2 operations will happen in someone else's computer (supabase, azure/aws, etc). On top of that, you have third party services that you need to integrate and manage their pricing plus auth. That alone is another challenge. Let's not even start with stateful data and migrations which is almost non existent.
The main problem is these tools don't tackle day 2 operations so it will be handled to some developer to make it happen which means exporting your code to some VCS service which I think it's only github. Right there, it's a threat for lovable and others. On top of that, there's not a real feedback loop between manual integration (external dev making it prod ready) and keeping the MVP workflow. Also, there's no real way these services can say "you are free to touch these components without breaking incompatibility with our system, anything else here be dragons".
In other words, You need to own the ecosystem to make more money. Funnel the capabilities to your own ecosystem.
> manual integration (external dev making it prod ready) and keeping the MVP workflow.
Right, the codebase generated by these can get huge, but maintenance can also be aided by AI tools like GitIngest[1], GitPodcast[1] etc. all helping you understand any new codebase easily or someone can query their doubts.
So, I wouldn't strike these kinds of tools yet. AFAIK, v0 already lets you connect github.
I’m using Lovable heavily for PM prototyping and it’s great. I think, if anything, current subscription may be too cheap! They’re probably want to get a huge mass of users now, eg they recently had a free usage weekend.
The comments on the Add Ons are spot on, I think:
“Lovable is creating lots of new software founders who will eventually spend lots of money on vendors. That money will flow, but Lovable currently captures zero of it.”
Having a Lovable App Store sounds like an excellent tool.
I think you must be the killer use case. As a programmer this is not good enough yet to warrant my time using it for production code (nor are its competitors)
However if I need to prototype for throwaway it would be ok.
These things right now compete with Figma and wire frames. Hopefully they lead ultimately to better UX in software.
> These things right now compete with Figma and wire frames
I think that is exactly correct. And beyond Figma or wireframes, they can actually be launched to see if they get traction and have product market fit.
Of course, I've seen tons of "throwaway" code that somehow never gets thrown away, and then, somewhat paradoxically, iteration velocity craters as the dev team tries to get a "prototype" to handle real load.
So what I'm saying is that I think things like Lovable are fantastic tools, but I'm quite confident they will be horribly misused and some poor sap will have the job of getting this stuff actually working with edge cases, security issues, scale, etc.
My prediction: this will look basically exactly like Visual Basic in the late 90s. VB was also heralded as "non-expert programmers can make apps just by drag and drop!" I actually think VB was a great product, the problem was most VB programmers were not, so VB apps took on a very negative connotation: like you could tell it was coded by a "VB coder", so you expected it to suck.
As a product designer, I'm seriously looking at using Lovable for quick ideation and prototyping. Showing users a series of Figma screens can be nice, but virtually everybody responds better to an interactive prototype.
The marketing is good. Especially on LinkedIn, for non technical people. But I am not convinced. Lovable is a nice Website or App builder. Nothing for professionals, unless they want to build an prototype. Even building MVPs is nearly impossible. It's like v0 from Vercel or anything you can do with other Agents like Claude or Co-Pilot. Lovable is promising a bit to much for my taste.
These apps will look dated in a few years, don’t waste your time. You’re just having fun playing around making old shit that could have made you a lot of money 10 years ago but is now just a weekend project. That’s the way things go in tech. Starry-eyed dreamers will let their imagination run wild, but they’re the laggards, the industry is already thinking ahead.
The next generation of apps isn’t going to look like the previous gen. No beautiful UIs and fancy CSS. No UI at all.
Instead, everyone will have some kind of platform like Cursor, but instead of just coding, it’s for everything.
Subscribing to new services for your AI to use will be the equivalent of downloading apps from an AppStore to your phone.
Then you can just say things like “fuck this person! AI, give me an OSINT profile of this Redditor!” and since your AI has the osint app it compiles the info instantly and says “here, damn”. No need to open an app, just straight info into your brain as quickly as possible.
AI has clearly made us tired of googling endlessly for info on random websites, so why are we still opening up apps to do various tasks? Because we want to see pretty interfaces? Get real. It’s time for the UNIX philosophy to go mainstream. Start thinking of how your product can minimize time to satisfaction, graphical interfaces get in the way of satisfaction.
The only problem is we currently don’t have a single unifying platform like an iPhone or something to consolidate a user base, but it will come. Start planning for that day so you can launch new services on day 1. It will be a gold rush.
And in the end, a lot of people will find they will struggle with coming up with good AI app ideas, because 80% of their idea was just putting a pretty interface in front of something complex. That’s how you know it was mostly a bad idea.
With that mindset you’d have skipped getting a PC in the 90s because the Internet in your pocket is gonna happen any day now.
For example let’s say the agent you describe can relieve people of manual data extraction from websites. Such an agent would be sheer utter magic for a friend of mine. Part of their role involves logging in to get data on employees from 4-5 different systems, compiling a spreadsheet of the data, turning the spreadsheet into a report, and sending it to internal folks. It’d be great to have an agentic tool that can drive the data extraction piece and automate their entire process with just a prompt - but today’s tech works just fine! A Playwright script, some carefully stored credentials, and a workflow automation are enough to get the job done - even if they require constant tweaking and monitoring.
Today’s tech is never the ideal version you can imagine, but it can still be used to solve important problems. Better to enjoy what is happening now even if you know it’s temporary.
I think you are right if we scale to a type of super intelligence in all domains.
Otherwise, I can't know everything so a human will be able to build a better tool with AI that I will use for domains I have no clue about. Just like it has always been without AI.
You probably should spend less time on Reddit though if that is the example that comes to mind.
Love this concept - LLMs as the new "API layer" to everything. Don't design a frontend because the user can generate one on the fly or use their own styling or pick a common "theme".
This analysis is very much on point. I'm building a product in this space (https://getmocha.com), and can share a few more insider insights:
- The churn from companies like lovable is indeed very high, and user frustration is high also.
- There are sub-niches available. Building internal tools is not the same as landing pages is not the same as saas. In the previous website builder market, different players (webflow, squarespace, wix) found and dominated a sub-niche
- The market is way bigger than anyone realizes. Today hundreds of millions for basically early adopters and highly tech-savvy users. This tech can and will go mainstream
- A huge issue with lovable, solved by others like mocha and replit, is the app backend. Lovable took a shortcut and partnered with supabase but that deal will not last. Supabase is losing big on their free tier (had to raise 200M to support it) and both want to capture the margin from the customer. There will be a reckoning.
What this article touches on is why I focused only on the serverless part with my project https://saasufy.com/ and why didn't implement any frontend UI for it (besides the control panel which is only concerned with backend concepts like schema definitions, auth, access control...). Now Claude is allowing generation and hosting of UI apps directly from their chat interface so it would be reinventing the wheel to build a UI.
AI companies are going to be working really hard to ensure that users do not bypass their chat interface because AI APIs don't provide much lock-in factor for them.
Oh great, another article on this company. I swear I keep being bombarded left and right with content desperately telling me hyperbolic things like "Lovable is the last piece of software humanity will ever build".
just have a question, for a young company like this, isn't mrr better than arr? or do they have that many yearly subscription? cuz u always subscribe and cancel different providers
I recently used lovable to create the scaffolding for an app. It did a great job, far better than I expected.
However, it also squishes everything into one JS file, making it unwieldy for anything moderately complex. After I fell off the happy path (integrating onnx was basically impossible), I had to spend a fair amount of time reworking it.
I’m probably not the target audience though. And I got the end result faster than I would’ve, and it’s better looking.
Perhaps me just stupid and do prompts bad, but I can't make Lovable come up with anything really differentiated as a design. I'm curious if folks have tips on how to use it well as I really love the idea behind it.
The kind of things I see mostly produced by these tools are landing pages, example dashboards and simple utilities like calculators and other simple stuff. There is certainly a space for these but nothing ground breaking that was not possible before with other beginner-friendly tools. The only difference is now that there is more freedom because LLMs can do more with less constraints.
I wonder what will happen after the honeymoon is over - after people realise that software development is not about writing new exciting code but maintaining things that should have been replaced years ago that are now too important because other people depend on it.
> I remember we were building a micro-app like that and it took at least a month, now it takes 3 minutes.
...please, statements like this have the power to invalidate the rest of the commentary. in 3 minutes you can complete a few prompts not nearly enough to write a piece of software the allegedly takes 3 months to do.
I work for a very small startup. Our CTO heavily uses lovable for our internal tools. Think dashboards, user management ... A somewhat standard UI over CRUD operations.
In previous roles, we either used some SaaS platform and would end up designing features based on the limitations of the tools, or we'd spin up a team to roll our own tailored solution. (or, God forbid, use Oracle and pay consultants a small fortune to do it for us).
With lovable, our CTO can validate assumptions and iterate on their own. Even if the code was absolutely garbage and we deleted all of it, this alone saves us a ton of man hours per feature. There's an entire lossy communication loop that's been removed.
Now, the code is only ok. We generate API clients from openapi definitions but it will struggle to properly orchestrate them. It will absolutely not have sane defaults anywhere, and will abuse the "any" type.
You still need a human in the loop, but my back of the napkin calculation is we'd have to hire two full time developers to do the work it's been doing for us.
Lovable operates at a higher abstraction level. You describe what you want, the UI magically updates, you never have to see the code.
Changes made by lovable aren't anything special. They live in their own branch, and get periodically merged to master, so you can carry on working on it as if it was any other project.
I'm pretty sure the person who wrote this has never run pricing research for a brand. Short answer, they can ignore Gabor-Granger because their cost base is so low compared to their revenue, so they'd be looking at Van Westendorp's Price Sensitivity Meter to set a benchmark for where the pricing probably lands, and a conjoint study to understand the value of different elements for segmenting different versions of the product at different price levels.
Obviously positioning, who they're positioning against, how they communicate that, the level to which they're known amongst the market etc all feed in to this, but that'd be a decent starter for ten.
This is an overly simplistic version of where to go with pricing for a brand like this, but that's where I'd begin with creating pricing for them.
How do you know what their cost base is? I don't quite understand how you'd be evaluating that. Their cost base is largely driven by the LLM and cloud companies. The $25/month pricing feels chosen to be similar to other unrelated SaaS businesses more than something based on a serious pricing analysis.
And the article confirms that, saying they made up a starting price and then immediately lost money due to (doh) selling a product where your costs scale by usage for an unlimited flat rate, which is surely one of the most basic pricing issues out there. And not just for LLMs: they host the apps too, putting them on the hook for hosting costs. They're using a ton of very expensive PaaS services like Supabase to do that too.
Then they have the free tier. Such services often have massive free tier costs; if their userbase is made up of a lot of people just trying stuff out quickly before exporting it to GitHub to continue, that problem will be worse. According to their blog, they have 30,000+ paying customers but there are 25,000 new apps created per day with 1.2M apps overall. So, clearly, almost all apps are being created by free users.
Don't get me wrong, maybe they're printing money like there's no tomorrow, or maybe they will soon. But it feels like the sort of business that's probably burning VC money to buy market share. If their cost base is good then they feel very vulnerable to an OpenAI or AWS releasing something tomorrow that takes away a big chunk of the business, seeing as that's where the bulk of the value lies. Oracle already has something quite similar launched in APEX.
This is interesting to pull apart if anyone wants to add more id love to hear.
Right now Lovable has competition in the vibe coding arena. Like Replit for example. I found Replit to be better in my testing.
I think there is an interesting curve where software is generally worthless (see Github!) but software plus marketing/sales etc. is valuable. But if you have any kind of scale that software needs to be robust and AI can't do that yet.
So there is a weird evolving Venn diagram where the final app industry fits in. If one player can take it yeah they'll be the next Google but that's a big IF and a big WHO.
i see one crazy real leverage: every prototype built here is a frozen snapshot of someone's product thinking in motion. like u can literally watch how ppl prioritize flows, kill features mid-wireframe, choose friction over flexibility. it’s raw cognitive output
if lovable ever starts versioning those moves, storing reasoning behind edits, even lightly, u got a time-series of product intuition across thousands of users. that’s applied decision memory.
there's a window here to become the place where product sense gets archived and replayed.
...and then PM's can have the same wonderful experiences as engineers: finding the exact commit where a major change was made 6+ months ago by a former employee, with a comment like 'updated behavior' that gives zero insight into what led to the change or why it was made.
Since working a one company that required it, ever since I've adopted the policy of including a link to the bug / feature ticket for every commit.
Even for my personal projects, if I'm fixing something, I'll manage them with JIRA and will create a 1 line bug ticket just before committing so that I have an ID to reference. I also have a long-lasting tickets for general refactoring / code cleanup without a bug fix.
I'm a non-tech worker in a non-tech industry, let me state two things:
- Software today is written to cover as many use cases with as many features to target as many users a possible.
- End users very often only use a tiny slice of the program's capabilities, but still pay for the entire program.
This creates a situation where the people writing software see it as a monumental undertaking to get good functional programs (it is), and end users see programs as having annoying learning curves with lots of bloat and "unnecessary" features.
LLMs do an excellent job of fixing this for end users because it allows them to easily create a program that does the handful of tasks that they normally need to use MegaSoftware for. And it's tailor made exactly for the use case. And the LLM can tell you exactly how to use it.
I can give a brief example where I used gemini to create a CAD file transposition tool that utilized a simple GUI tailor made for the files my company works with. This allowed us to forgo a (very) expensive CAD software package to work through converting our archive of files. A probably 2M LOC program could be skipped because we only needed 3k LOC functionality.
I really cannot stress enough how often this is the case, and why SWEs see LLMs as weak tools while end users see them as gods.
There will still be a need for huge software packages in the future, but I know I never again have to pay for a huge class of "here is a large solution space that covers your small scope problem" software.
To bring it home, loveable understands this, an sees that the futures has lots of non-tech people "writing" software. Standard IDEs are not the tools your mom will use to make a "Friends and family birthday reminder" app.
End users rarely pay for the program. Someone in their management chain OKs the purchase, or there's a larger purchase with a cross-charge to the department for the license cost. the problem comes when software needs to meet every whim of the decision maker, when really the users only will ever use 20% at best.
I think you may be assuming a certain enterprise size and accompanying workflow, probably would need stats to actually know how much of software is bought in this way however, and if the other way described would open up possible purchases by smaller companies, as was claimed.
So true. Some of my favorite enterprise software I use often I could never afford or would never purchase for my own use. I've taken jobs because of this.
IDA?
> I'm a non-tech worker in a non-tech industry…
Certainly you must have enough detailed knowledge of CAD files to validate the output of the transposition tool you had AI create for you. This might not be enough for you to think of yourself as ”technical” but I’d argue that it’s far above the level of “entry level employee using CAD”.
This does also seem to fit the paradigm of “AI is a productivity booster for people who already know how to do x”
I think users generally have the expertise to evaluate the quality of output for their own problems
I agree with you, and I think the Jevons paradox will eventually manifest itself once again. How many smaller companies are stuck with outdated workflows and tools because they can't afford to pay ten engineers for months on end for something better?
Now those companies may very well be able to afford one engineer and some AI subscription to do the equivalent work.
I don't deny that there's utility in what you are describing: if you can make it work for you that's fantastic. However:
- if you can ship software like that given the current state of the technology, you are probably not the average non-tech worker in a non-tech industry. There are people paying exorbitant consulting rates for dashboards in PowerBI. LLMs in mid 2025 are orders of magnitude more operationally complex than anything most people have seen.
- "citizen developers" doing something to scratch their own itches sounds very much like how a professional software project starts. Suddenly the scope grows and you need a nerd to handle it. Then two. Then four. You get the idea. Maybe that won't be the case for your specific needs, but that's how it generally goes.
Weak or strong is a matter of framing, but that's why I see them as tools and not gods.
I don't disagree with the thrust, but I've recently cleaned some of those up.
One example: LLMs aren't smart enough to do things like properly manage zip codes with leading zeros. It was round tripping strings through an integer representation and corrupting them. The users did notice, but did not have the vocabulary/concepts to explain. To them, sometimes zipcodes get corrupted because inscrutable reasons (tm).
chatgpt also authored a bash script that would have blown away a chunk of my drive if any paths had a space in them. :shrug:
Fun thing I noticed is converting a CSV to XLSX in Excel also drops leading 0s from zip codes...
Most spreadsheet software does that because the zips are parsed as integers, not strings . Prefix numerical fields that may contain leading zeros with a non-numerical character/word and save yourself the pain of late discovery! You can just find and replace after it goes in successfully.
Spoken as someone that experienced the same pain years ago in local SEO data that made it way too far before discovery...
You need to select the correct column data format in the Text Import Wizard: https://filestore.community.support.microsoft.com/api/images...
Good point. "Build vs buy" is a perennial controversy
https://www.thoughtworks.com/content/dam/thoughtworks/docume...
I think there was a consensus over the past decade, but we are now having to adjust our priors. The answer is changing month by month. It also means people are delaying their decision because they are afraid of the wrong solution right now except in the most obvious cases.
Build vs buy vs vibe?
You're absolutely right and one has to judge the HN consensus on AI tools through a lens of bias of the very demographic being challenged by AI tools.
I don't get how this company makes money. Everyone who uses these tools is just building prototypes. They get to 60-70% functionality, but when it comes to actually productionize and launch, 99% of these projects will get abandoned. The churn has to be absurdly high, maybe people are just forgetting to cancel their subscriptions. Or they're just marketing very hard and getting new sign ups/subscriptions, but will crash and burn soon enough.
I noticed in the example product the person was bragging they made "$1.6k in 5 weeks" but looking at the screenshot they were selling "lifetime access" built on this product they pay $25 a month for. I think there is going to be a reckoning for all these fly-by-night SaaSlings and the poor people who start to rely on them.
> Everyone who uses these tools is just building prototypes.
But also I think that's kinda the point. Like for example we have to do a UI redesign of an app my company is building and our "wireframes" are just v0 projects my PO created in one afternoon.
I think wireframing is where these tools really shine. It's gives you roughly the same ideas as building a wireframe in Figma but it's way less work and you end up with higher fidelity wireframes.Like sure if I were to peek at the code under the hood I'm certain it's close to dogshit but the code doesn't really matter at that stage.
Any serious company has a brand/design system which Lovable can't leverage well, thus needs designers to use in Figma.
Either way, my point is the customer LTV will be super low for use cases like this.
I haven't tried with lovable, but gemini studio is fully capable of following a styleguide based on a screenshot, which is probably the crudest and least refined protocol I could think of.
I'm sure there's already integrations somewhere that could allow a designer to specify a bunch of brand colours, to generate a styleguide out of it in plain english, and get a working prototype out of it.
A design system is not "following a styleguide based on a screenshot". A design system is an ontology of UI elements that your software has to adhere to, with several rules about how those elements are or aren't allowed to interact.
Following a styleguide will make your software break as soon as the design system gets updated.
I think you have an idealistic view of what design systems are in practice once filtered down to the developer (or AI) who is tasked with implementing them. You consider yourself lucky if the devs use the right fonts, icons, and colors.
Behaviors will be whatever component library they're using does.
I guess you could argue its just marketing, but if that's the case they're not building "the last software" as they claim. In fact they become entirely reliant on people building more software.
> building the software that can build all other software
All other software?
I'm afraid I stopped believing the author at that point...
It’s a marketing piece written by copywriters. Probably with AI suggesting the main arguments.
I don't think the idea sticks with me. The final products of these services are never reliable.
Right now, except for some hyperscalers, any similar service has integrated their deployments to some hyperscaler which means any day 2 operations will happen in someone else's computer (supabase, azure/aws, etc). On top of that, you have third party services that you need to integrate and manage their pricing plus auth. That alone is another challenge. Let's not even start with stateful data and migrations which is almost non existent.
The main problem is these tools don't tackle day 2 operations so it will be handled to some developer to make it happen which means exporting your code to some VCS service which I think it's only github. Right there, it's a threat for lovable and others. On top of that, there's not a real feedback loop between manual integration (external dev making it prod ready) and keeping the MVP workflow. Also, there's no real way these services can say "you are free to touch these components without breaking incompatibility with our system, anything else here be dragons".
In other words, You need to own the ecosystem to make more money. Funnel the capabilities to your own ecosystem.
> manual integration (external dev making it prod ready) and keeping the MVP workflow.
Right, the codebase generated by these can get huge, but maintenance can also be aided by AI tools like GitIngest[1], GitPodcast[1] etc. all helping you understand any new codebase easily or someone can query their doubts.
So, I wouldn't strike these kinds of tools yet. AFAIK, v0 already lets you connect github.
1. https://gitingest.com
2. https://gitpodcast.com
I’m using Lovable heavily for PM prototyping and it’s great. I think, if anything, current subscription may be too cheap! They’re probably want to get a huge mass of users now, eg they recently had a free usage weekend.
The comments on the Add Ons are spot on, I think:
“Lovable is creating lots of new software founders who will eventually spend lots of money on vendors. That money will flow, but Lovable currently captures zero of it.”
Having a Lovable App Store sounds like an excellent tool.
I think you must be the killer use case. As a programmer this is not good enough yet to warrant my time using it for production code (nor are its competitors)
However if I need to prototype for throwaway it would be ok.
These things right now compete with Figma and wire frames. Hopefully they lead ultimately to better UX in software.
> These things right now compete with Figma and wire frames
I think that is exactly correct. And beyond Figma or wireframes, they can actually be launched to see if they get traction and have product market fit.
Of course, I've seen tons of "throwaway" code that somehow never gets thrown away, and then, somewhat paradoxically, iteration velocity craters as the dev team tries to get a "prototype" to handle real load.
So what I'm saying is that I think things like Lovable are fantastic tools, but I'm quite confident they will be horribly misused and some poor sap will have the job of getting this stuff actually working with edge cases, security issues, scale, etc.
My prediction: this will look basically exactly like Visual Basic in the late 90s. VB was also heralded as "non-expert programmers can make apps just by drag and drop!" I actually think VB was a great product, the problem was most VB programmers were not, so VB apps took on a very negative connotation: like you could tell it was coded by a "VB coder", so you expected it to suck.
As a product designer, I'm seriously looking at using Lovable for quick ideation and prototyping. Showing users a series of Figma screens can be nice, but virtually everybody responds better to an interactive prototype.
The marketing is good. Especially on LinkedIn, for non technical people. But I am not convinced. Lovable is a nice Website or App builder. Nothing for professionals, unless they want to build an prototype. Even building MVPs is nearly impossible. It's like v0 from Vercel or anything you can do with other Agents like Claude or Co-Pilot. Lovable is promising a bit to much for my taste.
These apps will look dated in a few years, don’t waste your time. You’re just having fun playing around making old shit that could have made you a lot of money 10 years ago but is now just a weekend project. That’s the way things go in tech. Starry-eyed dreamers will let their imagination run wild, but they’re the laggards, the industry is already thinking ahead.
The next generation of apps isn’t going to look like the previous gen. No beautiful UIs and fancy CSS. No UI at all.
Instead, everyone will have some kind of platform like Cursor, but instead of just coding, it’s for everything.
Subscribing to new services for your AI to use will be the equivalent of downloading apps from an AppStore to your phone.
Then you can just say things like “fuck this person! AI, give me an OSINT profile of this Redditor!” and since your AI has the osint app it compiles the info instantly and says “here, damn”. No need to open an app, just straight info into your brain as quickly as possible.
AI has clearly made us tired of googling endlessly for info on random websites, so why are we still opening up apps to do various tasks? Because we want to see pretty interfaces? Get real. It’s time for the UNIX philosophy to go mainstream. Start thinking of how your product can minimize time to satisfaction, graphical interfaces get in the way of satisfaction.
The only problem is we currently don’t have a single unifying platform like an iPhone or something to consolidate a user base, but it will come. Start planning for that day so you can launch new services on day 1. It will be a gold rush.
And in the end, a lot of people will find they will struggle with coming up with good AI app ideas, because 80% of their idea was just putting a pretty interface in front of something complex. That’s how you know it was mostly a bad idea.
With that mindset you’d have skipped getting a PC in the 90s because the Internet in your pocket is gonna happen any day now.
For example let’s say the agent you describe can relieve people of manual data extraction from websites. Such an agent would be sheer utter magic for a friend of mine. Part of their role involves logging in to get data on employees from 4-5 different systems, compiling a spreadsheet of the data, turning the spreadsheet into a report, and sending it to internal folks. It’d be great to have an agentic tool that can drive the data extraction piece and automate their entire process with just a prompt - but today’s tech works just fine! A Playwright script, some carefully stored credentials, and a workflow automation are enough to get the job done - even if they require constant tweaking and monitoring.
Today’s tech is never the ideal version you can imagine, but it can still be used to solve important problems. Better to enjoy what is happening now even if you know it’s temporary.
> but today’s tech works just fine!
Absolutely fine. For software engineers. Agents are giving normal users access to the same power
I think it is a matter of scale.
I think you are right if we scale to a type of super intelligence in all domains.
Otherwise, I can't know everything so a human will be able to build a better tool with AI that I will use for domains I have no clue about. Just like it has always been without AI.
You probably should spend less time on Reddit though if that is the example that comes to mind.
Love this concept - LLMs as the new "API layer" to everything. Don't design a frontend because the user can generate one on the fly or use their own styling or pick a common "theme".
Exactly, bring your own front end. No more worrying about “user experience”
Would you need a new platform, or just a screen and a microphone?
This analysis is very much on point. I'm building a product in this space (https://getmocha.com), and can share a few more insider insights:
- The churn from companies like lovable is indeed very high, and user frustration is high also.
- There are sub-niches available. Building internal tools is not the same as landing pages is not the same as saas. In the previous website builder market, different players (webflow, squarespace, wix) found and dominated a sub-niche
- The market is way bigger than anyone realizes. Today hundreds of millions for basically early adopters and highly tech-savvy users. This tech can and will go mainstream
- A huge issue with lovable, solved by others like mocha and replit, is the app backend. Lovable took a shortcut and partnered with supabase but that deal will not last. Supabase is losing big on their free tier (had to raise 200M to support it) and both want to capture the margin from the customer. There will be a reckoning.
What this article touches on is why I focused only on the serverless part with my project https://saasufy.com/ and why didn't implement any frontend UI for it (besides the control panel which is only concerned with backend concepts like schema definitions, auth, access control...). Now Claude is allowing generation and hosting of UI apps directly from their chat interface so it would be reinventing the wheel to build a UI.
AI companies are going to be working really hard to ensure that users do not bypass their chat interface because AI APIs don't provide much lock-in factor for them.
Oh great, another article on this company. I swear I keep being bombarded left and right with content desperately telling me hyperbolic things like "Lovable is the last piece of software humanity will ever build".
just have a question, for a young company like this, isn't mrr better than arr? or do they have that many yearly subscription? cuz u always subscribe and cancel different providers
I recently used lovable to create the scaffolding for an app. It did a great job, far better than I expected.
However, it also squishes everything into one JS file, making it unwieldy for anything moderately complex. After I fell off the happy path (integrating onnx was basically impossible), I had to spend a fair amount of time reworking it.
I’m probably not the target audience though. And I got the end result faster than I would’ve, and it’s better looking.
Cybersecurity will be booming in 2026, trust me bro :)
I fully support vibe-coding in corporate env., plese bring more :D
Bug bounty programs will blossom like hell!!
We have AI tools working on those as well. ;)
Perhaps me just stupid and do prompts bad, but I can't make Lovable come up with anything really differentiated as a design. I'm curious if folks have tips on how to use it well as I really love the idea behind it.
The kind of things I see mostly produced by these tools are landing pages, example dashboards and simple utilities like calculators and other simple stuff. There is certainly a space for these but nothing ground breaking that was not possible before with other beginner-friendly tools. The only difference is now that there is more freedom because LLMs can do more with less constraints.
I wonder what will happen after the honeymoon is over - after people realise that software development is not about writing new exciting code but maintaining things that should have been replaced years ago that are now too important because other people depend on it.
> I remember we were building a micro-app like that and it took at least a month, now it takes 3 minutes.
...please, statements like this have the power to invalidate the rest of the commentary. in 3 minutes you can complete a few prompts not nearly enough to write a piece of software the allegedly takes 3 months to do.
What are the most popular use cases for these vibe coding products?
What are people trying to build where they are getting stuck at 75-90%?
what most people assume:
myth (a) - software engineers mostly work on software for consumers reality (a) - software engineers mostly work on software for enterprises
myth (a) - it's just crud reality (a) - it's crud but with a bunch of business rules behind it that ai can't grasp
every other shiny 'saas' that's not the reality of the majority of software out there.
Is there really a reason why should I pay for vibe coding apps like this when I can vibe code for free and unlimited with Google's Gemini?
Lovable should allow max flexibility to its users, allow them to monetize
Lovable should be in the money flow, allow their users to monetize
I work for a very small startup. Our CTO heavily uses lovable for our internal tools. Think dashboards, user management ... A somewhat standard UI over CRUD operations.
In previous roles, we either used some SaaS platform and would end up designing features based on the limitations of the tools, or we'd spin up a team to roll our own tailored solution. (or, God forbid, use Oracle and pay consultants a small fortune to do it for us).
With lovable, our CTO can validate assumptions and iterate on their own. Even if the code was absolutely garbage and we deleted all of it, this alone saves us a ton of man hours per feature. There's an entire lossy communication loop that's been removed.
Now, the code is only ok. We generate API clients from openapi definitions but it will struggle to properly orchestrate them. It will absolutely not have sane defaults anywhere, and will abuse the "any" type.
You still need a human in the loop, but my back of the napkin calculation is we'd have to hire two full time developers to do the work it's been doing for us.
Why does your CTO not use cursor or whatever other code IDE tools to build on your own standard internal stack?
At least that would allow them to handoff to other members of the team.
Lovable operates at a higher abstraction level. You describe what you want, the UI magically updates, you never have to see the code.
Changes made by lovable aren't anything special. They live in their own branch, and get periodically merged to master, so you can carry on working on it as if it was any other project.
I assume it’s because they don’t actually know how to code at all or how to use developer tools for that matter.
I'm pretty sure the person who wrote this has never run pricing research for a brand. Short answer, they can ignore Gabor-Granger because their cost base is so low compared to their revenue, so they'd be looking at Van Westendorp's Price Sensitivity Meter to set a benchmark for where the pricing probably lands, and a conjoint study to understand the value of different elements for segmenting different versions of the product at different price levels.
Obviously positioning, who they're positioning against, how they communicate that, the level to which they're known amongst the market etc all feed in to this, but that'd be a decent starter for ten.
This is an overly simplistic version of where to go with pricing for a brand like this, but that's where I'd begin with creating pricing for them.
How do you know what their cost base is? I don't quite understand how you'd be evaluating that. Their cost base is largely driven by the LLM and cloud companies. The $25/month pricing feels chosen to be similar to other unrelated SaaS businesses more than something based on a serious pricing analysis.
And the article confirms that, saying they made up a starting price and then immediately lost money due to (doh) selling a product where your costs scale by usage for an unlimited flat rate, which is surely one of the most basic pricing issues out there. And not just for LLMs: they host the apps too, putting them on the hook for hosting costs. They're using a ton of very expensive PaaS services like Supabase to do that too.
Then they have the free tier. Such services often have massive free tier costs; if their userbase is made up of a lot of people just trying stuff out quickly before exporting it to GitHub to continue, that problem will be worse. According to their blog, they have 30,000+ paying customers but there are 25,000 new apps created per day with 1.2M apps overall. So, clearly, almost all apps are being created by free users.
Don't get me wrong, maybe they're printing money like there's no tomorrow, or maybe they will soon. But it feels like the sort of business that's probably burning VC money to buy market share. If their cost base is good then they feel very vulnerable to an OpenAI or AWS releasing something tomorrow that takes away a big chunk of the business, seeing as that's where the bulk of the value lies. Oracle already has something quite similar launched in APEX.
I promise you they're negative on unit economics.
In addition to what you're saying: - hosting costs - live sandboxes costs
They're betting on LLM costs going way down and VC funded until then
I'm an engineer. Where do I begin learning these things without taking an MBA?
https://m.youtube.com/watch?v=iuRbDBAqyFY&t=9s&pp=2AEJkAIB
https://www.qualtrics.com/marketplace/vanwesterndorp-pricing...
https://www.amazon.com/Confessions-Pricing-Man-Affects-Every...
https://www.amazon.com/Strategy-Tactics-Pricing-Growing-Prof...
Thank you so much!
Yup. Like comparing pricing of cars to pricing of horses. Lovable is competing with future platforms, not present ones.
This is interesting to pull apart if anyone wants to add more id love to hear.
Right now Lovable has competition in the vibe coding arena. Like Replit for example. I found Replit to be better in my testing.
I think there is an interesting curve where software is generally worthless (see Github!) but software plus marketing/sales etc. is valuable. But if you have any kind of scale that software needs to be robust and AI can't do that yet.
So there is a weird evolving Venn diagram where the final app industry fits in. If one player can take it yeah they'll be the next Google but that's a big IF and a big WHO.
what's funny is they're a bill metering company.
Good product placement of Lago :)
i see one crazy real leverage: every prototype built here is a frozen snapshot of someone's product thinking in motion. like u can literally watch how ppl prioritize flows, kill features mid-wireframe, choose friction over flexibility. it’s raw cognitive output
if lovable ever starts versioning those moves, storing reasoning behind edits, even lightly, u got a time-series of product intuition across thousands of users. that’s applied decision memory.
there's a window here to become the place where product sense gets archived and replayed.
I think what current agent's are missing is taste or the ability to tune taste. So capturing taste across many users might be really valuable.
...and then PM's can have the same wonderful experiences as engineers: finding the exact commit where a major change was made 6+ months ago by a former employee, with a comment like 'updated behavior' that gives zero insight into what led to the change or why it was made.
Since working a one company that required it, ever since I've adopted the policy of including a link to the bug / feature ticket for every commit.
Even for my personal projects, if I'm fixing something, I'll manage them with JIRA and will create a 1 line bug ticket just before committing so that I have an ID to reference. I also have a long-lasting tickets for general refactoring / code cleanup without a bug fix.
[dead]