The post reads like written by someone who read too much about AI rather than tried to build a startup with the help of AI that they advocate so much. I'm still bounded by system design, UX, pricing and feature decisions, if not by the speed of code output, by the review time for sure. Yes, iterating is faster, but we're nowhere near agentic AI loops spitting out working products. Technically it's possible, but then you just spent that time planning and writing the spec up front, which you'd interleave with dev time otherwise. If the product is a simple CRUD database skin, then yeah, chances of success are lower I think, but this is not the type of startups the post seems to write about.
I'm not, but this is not a great introduction. It's handwavy and makes the assumption that AI dev tools are much farther along than they are. I have seen this a lot lately; the farther up the management chain and farther away from putting hands on code, the more confident people seem to be in the power of AI tools.
For big complex real world problems, and big complex real worlde codebases, the AIs are helpful but not yet earth shattering. And that helpfulness seems to have plateaued as of late.
Made me wonder if there's a live-streaming equivalent for blogging... some platform that both ensures the reader knows the blogger is a person, and promotes a parasocial relationship.
There's live-coding, so it's not totally a crazy idea.
Your comment immediately made me think about the extreme opposite: a Davy Force-like, Infochammel-style livestream of a never-ending AI generated Ted Talk, offering delectable morsels of tech startup wisdom, but is ultimately zero calorie.
Well of course it is, most startups are dead on arrival.
The big pinch of salt I throw in with advice like this though is that startup failure rate hasn't dramatically shifted despite two decades of lean startup methodology, accelerators, and an entire cottage industry of startup advice. It's never the fault of the framework, mind you.
If anything the failure rate has probably increased with more capital and founders chasing after the same opportunities. Being a startup founder gained prestige and became the default thing to do after college for certain types of people who before the GFC would have ended up in finance.
isn't this a problem of being able to see whatever we want in the data? For example maybe more startups than ever are created in part due to access to this info. Volume increase but rate doesn't. Or maybe magnitude of success increases. Surely this is true but largely attributed to internet overall. But maybe methodologies are indeed a factor.
edit: Tobi from Shopify has an insight that relates. His north star metric is user churn. sounds crazy on its face. he's known for that. But increasing churn means you've increased top line exposure to more would-be entrepreneurs. Not all of them will succeed, but shopifys mission is to create more entrepreneurs. Grow the pie. A focus on increasing conversion tends to have a narrowing effect.
Sure, but if the advice works, you'd see failure rate drops over time. It would work like medicine - we have more people than ever before, but almost no one dies of polio anymore. That's not what we see though, failure rate is basically the same.
"You better be doing something in AI" applies to "startups" - businesses that are expected to spend every penny as quickly as possible, to meet the metrics needed to raise the next (bigger) round of funding.
It is also not the same as, "If you want to be a profitable company...". For that you need to somehow make more money than you are spending.
> Founders who started pre-2025 typically have built a technical stack optimized for a world where software development was bespoke and expensive.
Of all the things that AI has changed, tech stacks aren't one of them. The bots will gladly write Typescript, Java, Python, Rust, what have you. They could not give less of a shit.
This article resonated with me, especially since I've noticed the fund raising hoopla's in my circle has dramatically dropped. Either investors are tightening the belt so founder-investor fit has crossed into the realm of disillusionment
Building SaaS businesses has become a whole lot less capital intensive. Solo founders can go much further than they've been able to previously. New startups probably don't need funding anymore.
Seems like the headline should have been "is now dead on arrival". As currently written, it fails to convey the temporal aspect that is the focus of the blog post.
It also fails to convey that he's actually only talking about startups that were created 2+ years ago, rather than the many AI startups founded in the last 2 years.
It’s easier to view it in terms of DCF - the value of a cash flow generating asset = present value of expected cash flows discounted back at a risk discount adjusted rate. In other words what you’ve invested into your existing assets is irrelevant - the cash flows generated by them and the growth assets through future investment, is what matters.
I read it less as “every startup now needs an AI feature” and more as “your assumptions expire faster than they used to.” That part feels true even if the examples are a bit overstated imo...
> Now, before you build a physical prototype, you can simulate more design variants, create digital twins, and stress-test assumptions earlier and much cheaper than before.
Hahahahahahahaha no you can't. The rise of LLMs has done little to nothing in this area because it's very much compute-limited. Digital-twins and other ML-based strategies predate ChatGPT by a long shot. There are definitely places in hardware design where LLMs and agentic workflows will help, but that's largely because the existing tooling is utter garbage, and now the industry has a fire under its ass to make things automatable so they can build their own agents.
I find the story of a startup founder who entirely missed the developments of the last two years and did absolutely nothing with AI difficult to believe. If that actually happened it's the exception, not the rule. Most startup founders are way more in-tune with AI developments. This makes it sound like Chris (the mentioned founder) is behind marketing people who use LLM bots to post slop on LinkedIn.
In that case, yes their startup is most certainly DOA.
Look around, you're most definitely in a bubble. LLMs are bleeding edge by themselves. Using agentic anything is mega bleeding edge. Having something actually working reliably is a tiny sliver of the bleeding edge audience. We have barely entered early adoption phase. Most AI users out there are Q&A'ing it and they have no idea what agents, tool calling or context compaction are.
I don't follow. Tech startups are bleeding edge. You may be over-generalizing here. I talk a lot of my dev friends, they are all using AI for work. So if Joe Blow at some consulting company is using it, then a SV startup CEO should be too.
>Most AI users out there are Q&A'ing it and they have no idea what agents, tool calling or context compaction are.
Again, talking about a tech CEO not a random "AI user."
I know a 60-ish year old serial startup CEO who never coded. Now he's running teams of Claude Code agents which he used to build a healthcare platform by himself which he is now selling. He's already wealthy, just doing it for fun. This skill set is the bare minimum for a startup founder today.
Going to VCs with a 2 year outdated deck with no AI functionality or plans or tool use is unimaginable.
Yeah I get what you're saying. Sounds like this guy is driven, smart and savvy then? Definitely a bleeding edge minority who goes where the puck is.
As to my original point I'll go and say that 95/100 of business owners out there in the world don't have any idea about Claude Code, Codex or anything of that caliber. It's too early. By the time that group gets to it there'll be tools tailored to their needs, not a terminal-based coding agent messing up filesystem.
You keep returning to this which is a strawman of my argument. From the start I have been talking about technical startup founders, which is also the subject of the blog post.
You on the other hand keep assuming that startup founders, even in SV are technical enough to envision AI or can change the thesis of their business on a whim of technology change that has been introduced less than a year ago. It's not realistic. Businesses have systems, people, obligations, legacy, etc.
Did you read the blog post? It seems like you may be missing a lot of context here.
I'm not assuming anything, that is my argument. I think you could do a better job countering it than pointing at the status quo of every business in the world. lol.
Me: 'cutting edge startups should use cutting edge tools'
You: 'every business in the world isn't using cutting edge tools!'
It's not about the status quo. It's about the speed of change and focus. While I'm all in on building custom agentic loops and building extensions for my pi I have enough bandwidth and xp to see that it's quite easy for business people to be focused on the matters that are completely non-technical. That plus plain sunk cost fallacy and blind denial ofc.
A laundromat isn't a startup in that sense. There's no potential for exponential growth. A VC would never give you money to open a laundromat, you'd go to a bank and get a business loan for that.
The post reads like written by someone who read too much about AI rather than tried to build a startup with the help of AI that they advocate so much. I'm still bounded by system design, UX, pricing and feature decisions, if not by the speed of code output, by the review time for sure. Yes, iterating is faster, but we're nowhere near agentic AI loops spitting out working products. Technically it's possible, but then you just spent that time planning and writing the spec up front, which you'd interleave with dev time otherwise. If the product is a simple CRUD database skin, then yeah, chances of success are lower I think, but this is not the type of startups the post seems to write about.
Yeah… also, it’s just weird. Interfaces are important, they contain information and affordances, everything should not become a chatbot.
Are you familiar with Steve Blank? What you’re describing really isn’t his MO at all.
I'm not, but this is not a great introduction. It's handwavy and makes the assumption that AI dev tools are much farther along than they are. I have seen this a lot lately; the farther up the management chain and farther away from putting hands on code, the more confident people seem to be in the power of AI tools.
For big complex real world problems, and big complex real worlde codebases, the AIs are helpful but not yet earth shattering. And that helpfulness seems to have plateaued as of late.
I am extremely skeptical of posts like this.
> the bottleneck is no longer engineering, it’s ____
90% of blog articles created in the last two years are probably dead on arrival
Made me wonder if there's a live-streaming equivalent for blogging... some platform that both ensures the reader knows the blogger is a person, and promotes a parasocial relationship.
There's live-coding, so it's not totally a crazy idea.
Your comment immediately made me think about the extreme opposite: a Davy Force-like, Infochammel-style livestream of a never-ending AI generated Ted Talk, offering delectable morsels of tech startup wisdom, but is ultimately zero calorie.
Isn't it just podcast but text
Radio host
Well of course it is, most startups are dead on arrival.
The big pinch of salt I throw in with advice like this though is that startup failure rate hasn't dramatically shifted despite two decades of lean startup methodology, accelerators, and an entire cottage industry of startup advice. It's never the fault of the framework, mind you.
If anything the failure rate has probably increased with more capital and founders chasing after the same opportunities. Being a startup founder gained prestige and became the default thing to do after college for certain types of people who before the GFC would have ended up in finance.
Arguably that mostly says stuff about average VC skill to pick winning idea.
If you have good product idea, the methodology to get there mostly affect profit marigins, not whether it will be success or total failure
Execution is very important. Startups almost never start with the "right" idea.
isn't this a problem of being able to see whatever we want in the data? For example maybe more startups than ever are created in part due to access to this info. Volume increase but rate doesn't. Or maybe magnitude of success increases. Surely this is true but largely attributed to internet overall. But maybe methodologies are indeed a factor.
edit: Tobi from Shopify has an insight that relates. His north star metric is user churn. sounds crazy on its face. he's known for that. But increasing churn means you've increased top line exposure to more would-be entrepreneurs. Not all of them will succeed, but shopifys mission is to create more entrepreneurs. Grow the pie. A focus on increasing conversion tends to have a narrowing effect.
Sure, but if the advice works, you'd see failure rate drops over time. It would work like medicine - we have more people than ever before, but almost no one dies of polio anymore. That's not what we see though, failure rate is basically the same.
I almost feel like this is always true. Every startup is dead on arrival. Most of them fail.
You’ve always needed to constantly learn and innovate to launch a successful business.
"You better be doing something in AI" applies to "startups" - businesses that are expected to spend every penny as quickly as possible, to meet the metrics needed to raise the next (bigger) round of funding.
It is also not the same as, "If you want to be a profitable company...". For that you need to somehow make more money than you are spending.
> Founders who started pre-2025 typically have built a technical stack optimized for a world where software development was bespoke and expensive.
Of all the things that AI has changed, tech stacks aren't one of them. The bots will gladly write Typescript, Java, Python, Rust, what have you. They could not give less of a shit.
This article resonated with me, especially since I've noticed the fund raising hoopla's in my circle has dramatically dropped. Either investors are tightening the belt so founder-investor fit has crossed into the realm of disillusionment
Building SaaS businesses has become a whole lot less capital intensive. Solo founders can go much further than they've been able to previously. New startups probably don't need funding anymore.
VC for conventional SaaS is dead.
Seems like the headline should have been "is now dead on arrival". As currently written, it fails to convey the temporal aspect that is the focus of the blog post.
It also fails to convey that he's actually only talking about startups that were created 2+ years ago, rather than the many AI startups founded in the last 2 years.
> How can we throw away years of work?
This trap has killed many startups, well before AI.
Now that code is cheaper to write, hopefully it becomes less of a problem?
In either case, founders should never fall in love with their solutions.
It’s easier to view it in terms of DCF - the value of a cash flow generating asset = present value of expected cash flows discounted back at a risk discount adjusted rate. In other words what you’ve invested into your existing assets is irrelevant - the cash flows generated by them and the growth assets through future investment, is what matters.
I read it less as “every startup now needs an AI feature” and more as “your assumptions expire faster than they used to.” That part feels true even if the examples are a bit overstated imo...
True, it increases the trials and errors you can have to receive pmf
> Now, before you build a physical prototype, you can simulate more design variants, create digital twins, and stress-test assumptions earlier and much cheaper than before.
Hahahahahahahaha no you can't. The rise of LLMs has done little to nothing in this area because it's very much compute-limited. Digital-twins and other ML-based strategies predate ChatGPT by a long shot. There are definitely places in hardware design where LLMs and agentic workflows will help, but that's largely because the existing tooling is utter garbage, and now the industry has a fire under its ass to make things automatable so they can build their own agents.
Not being AI focused might mean fewer competitors.
I dig MPO as a term and that's exactly what I think of implementing agentic systems.
Chris has been so focused for 5 years that he had no clue about anything else going on in the industry?
It used to be 90% of startups would unfortunately fail.
Now with AI, it is likely going to be 98%.
Still better odds than winning the lottery right ? Right ?
I find the story of a startup founder who entirely missed the developments of the last two years and did absolutely nothing with AI difficult to believe. If that actually happened it's the exception, not the rule. Most startup founders are way more in-tune with AI developments. This makes it sound like Chris (the mentioned founder) is behind marketing people who use LLM bots to post slop on LinkedIn.
In that case, yes their startup is most certainly DOA.
Look around, you're most definitely in a bubble. LLMs are bleeding edge by themselves. Using agentic anything is mega bleeding edge. Having something actually working reliably is a tiny sliver of the bleeding edge audience. We have barely entered early adoption phase. Most AI users out there are Q&A'ing it and they have no idea what agents, tool calling or context compaction are.
I don't follow. Tech startups are bleeding edge. You may be over-generalizing here. I talk a lot of my dev friends, they are all using AI for work. So if Joe Blow at some consulting company is using it, then a SV startup CEO should be too.
>Most AI users out there are Q&A'ing it and they have no idea what agents, tool calling or context compaction are.
Again, talking about a tech CEO not a random "AI user."
- Fair enough but tech CEOs aren't necessarily technical, developers or keep up-to-date with the tech on the daily like HN crowd.
- The AI jump happened in Q3-Q4 2025 with Opus 4.5 so it's been six months or so? Not long enough.
- Most developers out there use AI for their coding work, not for re-envisioning business models.
I know a 60-ish year old serial startup CEO who never coded. Now he's running teams of Claude Code agents which he used to build a healthcare platform by himself which he is now selling. He's already wealthy, just doing it for fun. This skill set is the bare minimum for a startup founder today.
Going to VCs with a 2 year outdated deck with no AI functionality or plans or tool use is unimaginable.
[delayed]
Yeah I get what you're saying. Sounds like this guy is driven, smart and savvy then? Definitely a bleeding edge minority who goes where the puck is.
As to my original point I'll go and say that 95/100 of business owners out there in the world don't have any idea about Claude Code, Codex or anything of that caliber. It's too early. By the time that group gets to it there'll be tools tailored to their needs, not a terminal-based coding agent messing up filesystem.
>95/100 of business owners out there in the world
You keep returning to this which is a strawman of my argument. From the start I have been talking about technical startup founders, which is also the subject of the blog post.
You on the other hand keep assuming that startup founders, even in SV are technical enough to envision AI or can change the thesis of their business on a whim of technology change that has been introduced less than a year ago. It's not realistic. Businesses have systems, people, obligations, legacy, etc.
Did you read the blog post? It seems like you may be missing a lot of context here.
I'm not assuming anything, that is my argument. I think you could do a better job countering it than pointing at the status quo of every business in the world. lol.
Me: 'cutting edge startups should use cutting edge tools'
You: 'every business in the world isn't using cutting edge tools!'
It's not about the status quo. It's about the speed of change and focus. While I'm all in on building custom agentic loops and building extensions for my pi I have enough bandwidth and xp to see that it's quite easy for business people to be focused on the matters that are completely non-technical. That plus plain sunk cost fallacy and blind denial ofc.
It still does not appear that you read the blog post. Going to stop here.
Not if you’re launching a startup based in the real world. Tell me how AI will make a laundromat business DoA?
If your business is selling services at 40% margin that are entirely digitally based, then maybe you’ll need to cut some margin, sure.
A laundromat isn't a startup in that sense. There's no potential for exponential growth. A VC would never give you money to open a laundromat, you'd go to a bank and get a business loan for that.
Was uber a startup? Airbnb?
Doordash for laundry.
You jest but I searched "Uber for laundry" and found services partnering with both Uber and Doordash for transportation.