I guess the author hasn't done real software development. The cost isn't just for the code. It's for the whole process - especially the architecture. Which database to use for the use case, which framework and language to use, how the database should be structured,table naming standardization, best practices, security audits and everything else.
Can AI do all that? Sure, but you must know to ask for all that in the first place. Look what happened to Clawd/Molt.
> It's because building an app went from a $50K project to a weekend with Claude.
Sure, why don't you deploy your vibe coded app over the weekend and see if it falls apart after handling one request per second
This article was written by AI btw
Unless there is inherent complexity in the problem (and assuming subscriptions don’t get pricey soon) I can see nontechnical people getting into designing their own apps.
It makes me think of 3d printing. A lot of people got into 3d modeling because of it. And a lot of people publish cute baubles 3d models (analogous to vibe coded ai wrappers?) but there is genuinely useful stuff that people not in the fabrication or 3d design industry create and share, some even making money off of it.
I just can’t think of a way saas margins will stay as high as they are now.
If there was a text based file format for models, it could generate those and you could hand that to the slicer. Like I’ve never looked, but are stl files text or binary? Or those 3mf files?
If Gemini can generate a good looking pelican on a bicycle SVG, it can probably help design some fairly useful functional parts given a good design language it was trained on.
And honestly if the slicer itself could be driven via CLI, you could in theory do the entire workflow right to the printer.
It makes me wonder if we are going to really see a push to text-based file formats. Markdown is the lingua franca of output for LLMs. Same with json, csv, etc. Things that are easy to “git diff” are also easy for LLMs…
It's just gimped to the point that you can basically only use it for hobbyist projects, anything reasonably professional looking is using STEP compatible files and that is much more complex to try to emulate and get right. STEP is a bit different - it's more like a mesh in that it contains the final geometry, but in BRep which is pretty close to the machining grade, while OpenSCAD is more like what you're asking about - a textual recipe to generate curves that you pass into an engine that turns it into the actual geometry. It's just that OpenSCAD is so wholly insufficient to express what professional designs need it never gets used in the professional world.
Most of the problems you talk about are problems if you intend your software to be used at scale.
If you're building an app for yourself to track your own food habits; why does DB, framework, best practices matters?
People used to do this in an Excel sheet.
Now they can ask Claude to make them a nice UI similar to MFP or whatever.
Data can be stored in a single JSON file.
It's going to take years before they see actual performance issues.
And even though it becomes an issue, right now an AI Agent can already provide a fix and a script to migrate the data.
My only concern really is about security.
But a private VPS only reachable through Tailscale and they're ahead of 99% of the rest.
AI can fix bugs, sure. But every time you ask it to fix the same problem, it will come up with a new solution - usually unnecessarily complex. Will we reach a point where the AI can be its own architect? Maybe. But, I know for a fact that it's not what we have right now.
Right now, AI needs an architect to tell it how it should solve a problem. The real value of software is in the lived human experiences, not just the code. That's why we make certain decisions different than an AI would.
Ask an AI to vibe code an invoice app. It will make some really lovely looking UI - which is what unfortunately people judge an app by - but with a MongoDb backend which is totally not the right solution for the problem. That's what I mean.
Unless you had an AI write the article, you can't possibly know that. I'm sick of this being randomly thrown around: it's basically mentioned for every article posted. Sometimes the author chimes in to say that no, they wrote it themselves. Other times sure, the article was written by AI. I don't know, and you don't know either.
Actually this number was updated to 135,000 exposed instances recently
I think this is great for everyone to be a developer, the gatekeeping has now been removed and we will see a creative explosion of apps that everyone can build.
The security and maintenance aspect of apps is just a claude skill away to be a solved problem.
Who was preventing you from learning how to do it yourself and then doing it?
Nobody gatekept anything. The software, tools, knowledgebase (MIT, Coursera, etc) were always there. It was a choice. Some of us chose it, rest didn't for whatever reason.
To think that someone on Hacker News actually wrote this seriously in 2026, after a couple of decades of CVEs, security breaches, and data thefts being in the news every single week and after 50+ years of the industry experiencing how arduous software maintenance is. I doubt even Anthropic or OpenAI would be brave enough to say that.
Ironically, AI tend to be better at securing code, because unlike the squishy human, it is much more cable of creating tons of tests and figuring out weaknesses.
Let alone the issue when lots of meatbags with different skill levels are working on the same codebases.
I have barely seen any codebase that has been in production for a long time, that did not have glaring issues.
But if you tried to do a code audit, your spending somebody their time (assuming this is a pro), for a long time. Where as a AI with the correct hints on what too look for, can do insane levels of work, testing, etc...
Ironically, when you try to secure test a codebase, and you use multiple different LLMs, you get a very interesting list of issues they can find. Many that are probably in tons of production level software.
But its up to you, as the instructor of that LLM codebase, to actually tell it to do regular security audits of the codebase.
Sentences like this make me think AI is honestly the best thing that happened for my imposter syndrome. AI is great for simulating test case, and that's it. If you leave it, it write the most basic, useless tests (i mean, half of them might be usefull when you refactor, but that's about it). It can't design reusable test components and have trouble with test double, which i would think is the easiest test case for AI. Even average devs like me write test double faster than AI, and i'm shit at writing tests.
AI is also extremely bad at understanding versionning, and will use a deprecated API for no reason except increasing the surface of attack.
AI is great for writing CLI scripts, boilerplate and autocomplete. I use it for frontend because i'm shit at it (even though i have to clean its shit up behind), and to rewrite small functionalities of some libraries i want to avoid loading (which allowed us to remove legacy dependencies). It's good at writing prototypes (my main use nowadays), and a very good way to use it is to ask it a plan to improve/factorize your code (it's _very_ bad at factorizing, but as it recognize patterns, it is able to suggest interesting refactors. Half the time it's wrong, so use the "plan" mode)
I'm on a network security and cybersecurity tooling team, i guarantee you AI is shit at securing the code (and at understanding network).
Having the ideas necessary to know what to write is where practically all the value lies (caveat: there is value in doing the same as someone else but better, or cheaper.) AI can help with that, but only in so much as telling you the basics or filling in the blanks if you're really stuck. It can't tell you the 'clever bit' because that is by definition new and interesting and doesn't appear in the training data.
What this means is that at some point Anthropic will be able to prompt Opus to clone Jira and never pay an Atlassian bill again. Opus just needs to figure out what Jira is first. It's not there yet.
Bang on, and Jira is the perfect example! Because Jira isn't a bag of features: Jira is a list of features and the way they fit together (well or poorly, depending on your opinion).
That's the second-order product design that it's going to take next-gen coding AI workflows to automate. Mostly because that bit comes from user discovery, political arguments, sales prioritization, product vision, etc. It's a horrendous "art" of multi-variable zero-sum optimization.
When products get it right (early Slack) then it's invisible because "of course they made it do the thing I want to do."
When products get it wrong (MS Teams, Adobe Acrobat, Jira, HR platforms) then it's obvious features weren't composed well.
Expect there's more than one {user discovery} -> {product specification} AI startup out there, working on it in a hierarchical fashion with current AI now.
I bet it would burn a lot of money very fast and not just on tokens.
For the vast majority of companies they would (and should) rather let the SaaS figure that out and focus on their actual company
People always gripe about the poor quality of software staples like Microsoft Office or GitHub or Slack. Why hasn't OpenAI or Anthropic released a superior office suite or code hosting platform or enterprise chat? It would be both a huge cash cow and the best possible advertising that AI-facilitated software development is truly the real deal 10x force multiplier they claim.
If someone invents a special shovel that can magically identify ore deposits 100% of the time, they aren't going to sell it with the rest of the shovelmongers. They're going to keep it to themselves and actually mine gold.
For now though, building smarter models / general integration tooling is a better us of model companies' capital.
Once/if performance gains plateau, expect you'll see them pivot right quick to in-house software factories for easily cloneable, large TAM product spaces, then spin off those products if they're successful.
My guess is two-fold. One, they are specialized in AI. Two, building another anthropic is a big moat and they like to keep it big vs what you could build with it.
They have special knowledge to leverage AI to clone (and even improve) huge revenue businesses with high margin. If their claims about the abilities of LLMs are accurate it would be foolish to just leave that on the table.
It would also prove the power of their LLM product as truly disruptive. It would be amazing marketing!
Those apps aren't that bad, it's just internet people complaining about things like react.
Imo "higher quality" isn't a way to sell software
Cloning Slack and wasting ultra-expensive engineers on that might be more expensive, and it's not your core mission.
PS. If you're claiming that coding an application is ultra-expensive, you are already entering the argument on the side of the comment you're arguing against, which is making a counterpoint to the article, which claims in the first sentence:
> The math is simple: if it costs almost nothing to build an app, it costs almost nothing to clone an app. And if cloning is free, subscription pricing dies.
could be an unfortunate thing of the author growing up in an era of gated ecosystems.
however much of the software out there - is via web - and some desktop - some internal use - some external - some shit without ui - some billed yearly, some billed by subscriptions
but I guess tell us how AI is gonna kill subscriptions
At one point I had an idea I brought to AI, got ready to code it, then said "wait, someone has to have done this before me", sure enough, found it, written with warp!
So I can't say it'll kill all app subscriptions, but AI is definitely enabling people to finally make reality out of that idea they've had rattling around their heads but never took the time to realise.
That all said, I could see some killer features coming to AI companies if they really want to make a dent.
Think Vercel, Supabase, et al. Because most of the time agents prefer glueing together managed services than building from scratch, unless they're told otherwise.
And if I'm someone building a custom in-house solution to replace a SaaS subscription product, I'm going to pay lower managed costs without blinking.
The claim is that now every random person now will build their own app and have to make those hard decisions instead of paying $5 a month for someone else to do that work. Comparative advantage doesn’t just apply to the cost of writing code, but also the effort of making product decisions.
Edit: I don’t mean that a grocery app should cost $5/month, the grocery app was a toy example and the $5/month refers to an example of a separate app you’d pay for with much more value.
I in fact did just that. I used Claude to reverse engineer my grocery store's API and build a grocery list app that automatically pulls in the aisle information for each item and sorts it by how I typically walk through the store. It's the kind of thing that would be incredibly difficult to scale but works just fine when you only have one user. No SaaS grocery app can hope to compete with me being able to tailor my own shopping list app to my exact preferences.
I just don’t think most people will end up doing that just like how most people don’t 3D print their own desk drawer organizers even when Gridfinity does all the work for you. Automation doesn’t fully replace the volition to build a thing and make tricky decisions that are familiar to us software engineers but not others.
Given the number of apps put there, from dozens of OSS hobbyist apps to industrial resturant inventory management ones, I wan't alone in thinking this is a solved problem and someone should just have the perfect interface for it. Between auto-unit converting apps, natural language processing apps, @cooklang, a million ideas about tracking pantries and ingredients and their categories, frequency of use charts, etc..
Then one time I went on a trip with a friend to his home town where we stayed at his parents house. His 78 year old mother had a 2 notepads attached to the fridge with a pencil on a string. As she worked in the kitchen, between washing hands she would just jot down random notes, cross others, doddles some on one notepad, and the other she would just add meal plans as she went along. Then when we were going to market she just ripped the page off.
Sounds so fucking simple and easy and I felt so stupid for the amount of effort I put trying to figure out the right app, the right device to mount on my fridge, how to connect power to it. How to make it not always on to blind me at night, but also so I don't have to keep fiddling with it to unlock it. how to use it with wet fingers, how to keep translating units and "catch up" when I miss updating it for a couple of meals, how to hide ingredients I don't care about and highlight ones I do, how to rearrange the interface. It seriously gave me a pause at how dumb I was that the solution is much much simpler and I pigeon holed my thinking on a tech solution for some reason.
Can't sell people notepads though. There is no margin or lock-in in that stuff.
But for apps that have a lot of ceiling, people will still gravitate to apps that have had more care and attention than someone vibe coding it once and throwing it on the store, just like how people choose those well-built and maintained apps today over using their built-in Reminders app.
The reason we pay a subscription is because the company that built the software knows our business, knows how to get in touch with the decision maker, and knows how to market their product as something desirable. The actual software has little influence in that decision.
On the contrary, I think the price of SaaS subscriptions will go up as a result of AI. Because the only customers who will switch to a cheaper (or home made) alternative are the ones for whom the software is a commodity. These customers used to form the long tail of subscriptions, usually on the lower tier. When the entry pricer disappears, and the software editor has to generate a high return for their investors, the only way to keep profitability is to increase the price for the other tiers.
My guess is that copying is not enough, but adding value or saving costs is.
Your point is based on wrong assumptions.
Sucks for the todo-list-for-x and pretty basic game app creators to have even more competition, but they weren't making bank from subscriptions anyway
Did it ever make sense? I always scoffed at the idea of paying a subscription to use a text editor or paint tool.
Good riddance to software subscriptions.
I hope proprietary software goes the same way entirely. If it's trivial to build an open source competitor, why pay for software can't modify (also trivially).
Counter argument ... at what point is software still profitable to be sold?
I am running my Office 2007 still, and that thing is now almost 20 years old. That was a one time sale, with no other revenue for Microsoft.
I am not condoning subscriptions but one time selling software only works good, if your a small team with low overhead. The more you sell, the more support becomes a issue. And normal customers do not pay for support.
Making software now has become easier with LLMs but the same problem keeps existing in regards to support. Sure, you can outsource this to LLMs but lets just say that is problematic (being kind).
So unless you plan on making software that is not heavily supported/updated, and keep a low single/team cost...
If you sold a program for a one time fee of ... $39.
What if somebody now sells the same for $29 with LLMs. And the next guy in China does it even cheaper because his overhead is even smaller. Eventually you get into abandonware where software is made to just eat sales from the bigger guy and that is it.
Unless you focus on companies, and they have way less issue paying for subscriptions (if it includes support). You see the issue. People kind of overlook the cost of actually running a self employed job or a company (this is a MAJOR cost the moment you need to hire somebody).
So no, i do not see subscriptions going away because companies will pay for it. And on the normal consumer level, paid support as the solution?
No need to pay for someone else’s one.
Reasons why subscriptions may be a "better" than upfront licenses, even when the subscription cost more in the long run:
1. Cashflow management
2. Bypass budget approvals due to smaller amounts
App creators will be competing and copying each other. The software that can support change will probably win in the market. Probably…?
It’s inevitable at this point.
[0] https://mikelovesrobots.substack.com/p/wheres-the-shovelware...
Yes and no.
I have created 15 small apps that solves all kinds of things for me. However, at the department I work full of non-technical people, most of them don’t even know lovable exists.
And for the one that does know lovable exists, they tend to build stuff with some botched backend and you’ll get to scaling issues, security issues and who knows what else
AI can produce mediocre or outright bad code no problem.
It was trained on the average, not just on Carmack level code...
It needs to be checked by a professional.
This won't ever change with LLMs or gen AI like it is now
This also got me thinking about open source might be dying. For this tool, there is no reason for me to open source it, anyone can create the same thing in minutes. I didn't add anything, the only maybe interesting part would be to share the prompt, but then someone else can create their own prompt to have their tool do what they want.
Software world is really getting weird.
Automation of easy templated tasks will cause a huge disruption. Production of software used to be a skilled job, but now is automated to a large degree. This has huge impacts to the profession as a whole. Already, enrollment to the UC CS program is declining.
More people with more agents freely contributing more to even more concentrated and scaled-up projects.
Those agents will get more potent. The projects can get more ambitious.
One user with N Claude usage. 100 users with 100x the Claude usage. Who can build the superior product? If you put the right structure on it, the 100x wins by a drastic margin. Those 100x Claudes benefit in combination courtesy of the open source effect, their potential additive value is greatly enhanced.
The 1x outcome will end up being relegated to triviality (the one page homepage as website). The bar is about to be raised really, really, really high in software if you want to be relevant. This is merely a very short transition period.
Probably wouldn't be a bad thing.
What I think actually changes is the packaging. Per-seat subscriptions assumed humans in dashboards. That's the part on borrowed time. The workflow logic underneath still has value — it just gets consumed differently as agents start doing the clicking.
Point solutions are going to be free. Complex systems with support, integrations, switching costs, customer data, etc., are not going to be free.
IBM still sells mainframes but is no longer a growth darling.
> Markets are right to reassess multiples. But reassessing multiples is very different from pricing in extinction
What you are missing is that the SaaS companies were extremely overpriced. For instance, crm after all the carnage is still priced at 25 times earnings which is historically high for anything that is not a growth company. The perception was that these companies would print money year after year selling software trinkets on their platforms and as such were placed in the growth category. Now, it is plainly obvious that these software trinkets can be produced easily by anyone using AI. Their pricing-power has dramatically declined. Hence the re-rating. None of this contradicts the thesis in your ai-assisted article that these businesses have moats just like IBM and its mainframes. These businesses are now in a vicious reflexive narrative loop where the narrative will impact the real-world which will further fuel the narrative.
And even if you do build your own models, unless they run locally on the device, you still need to pay for hosting?
Never did tbh. These apps should be one time purchases at best.
- Open source - Outsourcing - Offshoring
It was driving the labour cost of an engineer to zero I felt as a young man.
Then time passed, and I learnt that engineers aren't paid to code. Engineers are paid to solve problems for a business.
If you recall, the dot.com bust and 9/11 crashed finances for a few years. When the money printing gun went whir because "Deficits don't matter" Washington, then engineers were in demand again.
Right now we are in a weird situation where money is being printed and it is also tight. Most of it is going to the hardware and infrastructure layer, like the fiber optic bubble in the dot.com. Software will have its time in the sun again.
Take a look at the history of the power loom which automated weaving in the 19th century. The number of handloom weavers dropped two orders of magnitude after the power loom.
"The math is simple: if it costs almost nothing to build an app, it costs almost nothing to clone an app. And if cloning is free, subscription pricing dies. We're already seeing this play out in the numbers. Apple's App Store got 557K new submissions in 2025, up 24% from 2024 (source: Appfigures). That's not because people suddenly got more creative. It's because building an app went from a $50K project to a weekend with Claude."
No. It's because people got more creative. There are tens of thousands of us who are absolutely on fire creating new products, better versions of old products, new product categories etc. Many of us are burnt out OG programmers who have rediscovered our love for programming. Now we can create without the drudgery.
You're about to see the most tech innovation our species has ever experienced. Hold on to your seat.
Nitpicking, but I would argue that people have always been creative, it's a function of our brains. With the ubiquity of camera, videos now show that even birds and animals have levels of creativity. Biological/physical/physics/societal restrictions prevent them from taking it to the next level. Look at what ancient peoples managed to achieve without the benefit of modern tools and techniques; hard to argue people haven't always been creative.
What has changed is the ability to implement our ideas and harness our creativity - that has become significantly simpler in the age of AI.
Perhaps the discrepancy between the OP's framing of what's happening (negative impact to developers because app cost has gone to zero) and your positive perspective (hey, look at all these creative ideas we are now implementing) is a matter of perspective: you're both describing the same phenomenon, just different angles.
1. Local models or token cost plummets
2. Subscriptions more prevalent, given token costs
3. A single subscription (tokens) to rule them all
4. Apps with use-based pricing
"AI", the new Napster.
But if people want to make more good creative games and the store helps me find them I have plenty more money to shovel at them
My theory as an old guy is that the standards will just go up.
There’s a current business model where you can make a basic but useful tool that solves a specific business problem and make money. That’s going to end.
We’ve seen this before. A good example would be when the mobile app stores launched and you could get traction with just about anything. And then you couldn’t.
I don't know... Because the tool that solves a specific business problem usually requires tons of business expertise. And when company buy this tool, they mainly do it for the expertise diluted in it.
If they didn't already made their own in-house implementation, it's because they don't want to invest in maintaining the tool that requires expertise outside of their actual business.
Meanwhile, the company building the tool can invest in keeping this expertise because it's financed by the multiple companies paying for the tool.
Why? VERY good design signals this is tasteful and quality. Not an AI-slop-vibe-mess.
I'm not too sure about this. Very good design, at least in terms of marketing materials such as screenshots, may be end up signalling the use of AI in the coding as well. I know if an app has obviously-AI-generated photos in the screenshots I would be disinclined even to try it.
Subscriptions arnt going away. Software will just be like cable and now streaming.
If there is a feature that cannot be done on-device, and that on-Internet feature can be effectively moated and duplication-resisted, then if it’s a feature that people think they need they will absolutely open their wallets to pay for.
The trick is finding that feature or attribute that cannot be done on-device, and then moating it against AI duplication. Do that, make it appear indispensable in the minds of people, and they will absolutely pay for it.
My best bet is that some NYC traders take on agents was to post the same bullish take a million times in order to drive tech stocks (which make up the most of the market cap), and buy them for low.
most apps i use are not AI clone-able yet with AI’s current faculties. i’m not going to switch to an ai vibe code of Google Photos, Tailscale/Mulvadd VPN, or YouTube. For those three apps, i pay for cloud infrastructure. sure, you can say with enough AI i could vibe code a Tailscale backend system, but it sounds like it would take more tokens than my $20/mo ChatGPT plan PLUS a mountain of cloud provider bills and such to host my backend.
i do pay for some premium apps that run entirely on device, like Halide Camera. But there again, is my $20/mo tokens enough to clone a high quality image processing app, to such a degree i will trust it to capture precious memories effectively? ehh.
Thanks to AI abundance, everyone will be better off.
For a company, paying $10K a year for a quality service, that's a no-brainer. Most companies spend that money on alcohol in company onsites. However, if you're charging really high prices (the Datadogs of the world), then you're going to face tougher competition from cheaper alternatives that might be as good as you, and when companies need to cut costs, which they often do, you'll be in trouble.
I think what it means to many software companies is that prices will significantly go down on average but the median might not see significant decrease. Companies will be smaller and more lean, hiring less people in general (not just engineers!). There will be more companies out there, so hopefully it will even out.
Last thing is that every product will have too many options to choose from. This has been the reality actually for a long time and going to get much worse. How you market and brand your product and acquire customers will become more difficult than ever.
It's identical to Craigslist hollowing out offline classified ads. Classified ads used to be a hyper lucrative market for newspapers (both local and national). That market imploded from ~$17 billion ($32b+ adjusted) in 2000 to $1-$2 billion last year. Once it could, it did.
AI should enable software to touch more things more cheaply (more efficiently in many cases). As it can, it will. Expect a lot more wipe outs.
I'm seriously wondering if this blog is just some rage bait or if that guy is really that dumb? I can't tell anymore.
It's like these companies are trying to get us hooked, then try to make us explode in frustration because it doesn't actually quite work. Not because the AI is bad, but because the interface (30-old-tech, a chat ui) is just broken.
Yes, the guy who owns a boat and wants to track his calories is going to fuck around in claude code and figure out deployment, and sign up to some free PaaS and pay $1.38 a month to self host their app.
Sure.
This is AI abundance for all and for free.
Also the end of the app store grifting.
I welcome this, having an app was never a competitive advantage at all.