Is AI at this level the end of many older SaaS products?

I've flaired this with Coding, because it's directly related.

Over the past year or so I've been accelerating development on the projects I run myself, and recently rebuilt them from the ground up now that the models have got so powerful.

I generally look for holes in the market before I take a project on. Recently I built an automated site that hunts for and only posts good news about the UK, running on a cheap VPS. Claude wrote the site, and free tiers on various AI platforms evaluate the news feeds to extract the content, then rewrite the clickbait headlines, because I can't stand those. It links back to the original article so you can read it in full, so I'm not busting the journalists.

Then I looked at the video hosting site I run internally for my company, where we keep playable versions of our Masters library along with cuts and exports for editors, production staff and broadcasters to watch. Think a cheap Vimeo. It wasn't good enough for 2026 and we wanted the features the big boys have (Vimeo, Frame.io for review, Trint for transcription). On modest hardware, I had Fable and Opus rewrite a decade of my php/ajax/nginx work into a lovely Python setup: Django, React and Vite, MariaDB, Redis and Gunicorn. Honestly, it's fast and slick. Then I added Ollama with Qwen locally for diarization and speaker identification. It's brilliant. Time taken: two weeks. That's replaced Vimeo, Frame.io and Trint completely, saving us thousands a year.

Next up, the edit suite booking system. Most companies use Farmer's Wife for this. It's expensive. You can tell from their website, they only offer to book a demo and there's no pricing anywhere. Full booking system written with Claude, finalised with every feature we needed, a lot of it bespoke, about 80% of Farmer's Wife minus the billing we didn't need. Even more thousands saved.

Yes, it helps that I'm an IT junky who's been everything from web dev to sysadmin and is now in a senior role. I know how to write software and I know how to spec it. But if I can do it, so can plenty of other people in businesses who can replace their SaaS products quickly and cheaply. Surely that's going to eat massively into these companies' profit margins and eventually kill them off, or at least drag them down to being a lot cheaper?

I'd love to get people's thoughts on how this is going to shape future SaaS offerings from various businesses. Are we going to see a death knell to many? So much competition we don't know what to do with it? How does the next couple of years look?

Massive players like M365 aren't being replaced any time soon, but the smaller systems... I cannot see them lasting.