In LLM era, can software startup survive?

I wonder if the flood of vibecoding startups may end up helping long established enterprises and OSS through eroding consumer trust towards new software, and highlighting the value of the industrial software development process. Bills of materials, 4-eyes code reviews, release gates, QA testers, planning involving domain experts, external audits etc. Any time I see a new large piece of software nowadays that seems to have popped out of nowhere and has generic sounding docs and testimonials with em dashes I tend to wonder if it was generated with a one shot prompt and may include vulnerabilities, malicious code, or business domain hallucinations.

It's a bit like with books: 10+ years ago I could have trusted a book from a new author I never heard about just because the large effort of getting published was hinting that the author is likely qualified. Now, hell no.

When the barrier to publish drops to zero, the barrier to gain trust skyrockets.

Creating new software is far more than just a matter of tokens. LLMs are like a small programming team: you still need a project manager (you!).

And tax return software is a perfect example of where "vibe coding" will go horribly wrong: a superficial tax return is trivial; handling tax rules for every different situation is vastly more complex. (1) You need to codify all the tax rules; (2) You need to do it reliably (which means really understanding how to work with LLMs: harnesses, gates, verification, reviews, etc.); (3) You need a corpus of real-world examples (taken from where?); (4) You need user feedback.

This would be an intensive multi-month project. And that's before even looking at marketing and sales.

Can a competitor race with you? Sure, but only if they're also determined to put in the work.

I had the same thought - that its end of software startups since most people can develop software on their own now using LLMs. But seeing how people use LLMs to develop the app made me realize that there is a difference in knowing how software is built vs just vibe coding it. For example - only a good s/w engineer would know to include CI/CD pipeline which will help her deploy changes faster. A person who is not familiar with s/w is not gonna do that unless they dig a little more deeply OR follows an advise from a good model. In 2 months time, the non s/w engineer is gonna drown in keeping up with all the pieces of the product without regression. So startups who provide value in short cycles would still be in demand. The cycle time of requirement -> delivery is gonna be very short.

The code was never the hard part. Understanding the market, developing product market fit, and evolving is the hard part.

Large companies aren't going to chase every opportunity.

I think it’s completely inverted actually. The big companies have a much smaller moat now. This is almost certainly the best time to do a startup. You can build a decent product very fast, GTM in a month or so, and pivot rapidly. But it requires a lot of skill to do it with AI tooling in a way that sustainable and the techniques and tools are very immature and shifting around rapidly. Large companies are struggling very hard to integrate these into their flows and workforce while a small company, with the right small set of people, can achieve as much or more than a large one with thousands of developers.

Some products are harder to build than others imo, even if making software has gotten easier.

Making consistent and reliable software is even harder.

To me, it seems that the modern software startup is geared up to provide highly customized solutions to customers. Software that easily bends to the customers needs, rather than customer changing their processes to make use of software.

The environment of computation and it's interactions in the 7 day week of a modern classed society is open to debate and great thought will come from it to solve the problems of mere boredom , as well as the vital life sustaining things people need.

If you are settled with distribution channels, can productize the solution or convert the startup feom product one to a service model then you are probably safe. The development itself is cheap and has no moat.

It's still usually cheaper to pay for software than it's to write it, maintain it and run it.

- getting it promoted like nord vpn is the hard part, you do that, someone or the other ll buy something as long as it serves a utility

[flagged]

[flagged]

[flagged]

[dead]