I can paint my house and I can change the oil in my car, but I prefer not to do it. I might host something personally, but self-hosting comes with cost, uptime requirements and bug fixing (by the AI and how would I verify the fix and possible regressions?).
A lot of B2B services exist to provide complete solutions. The business pays so it won't spend time/management resource/risk to nurse that solution.
With AI the cost of software will go down, no more 15$ per seat plans, but it won't go to 0 or self-host everything.
My client's customers are not paying for the software. They are paying for a complete solution that results in compliant operations.
I don't pay for Spotify because I need a way to listen to music. If it was just about music playback I'd use foobar and get into torrents again. I pay for Spotify because they curate a massive collection and provide many ways to discover new experiences.
I’m sure, but for 1-2M annual revenue small businesses I work with the price points are not affordable. Enter LLM.
I thought about hosting my own photos. The time is just not there, cheaper to pay Google.
Until AGI or Super Intelligence saas is not dead.
Once AGI and Super Intelligence here, everything is dead.
It's also cheaper in terms of time.
Not everyone sees things the same!
False. People just see things wrongly. Self-hosting is the best option here.
A few days ago I wanted my obsidian notes on a linux machine. Well, the files are synced via iCloud Drive (no Linux), so I asked it to solve it across my devices. It used Syncthing which has been working great.
Then yesterday obsidian desktop stopped loading (probably a bug in a plugin?). I’ve always wanted to do my notes in the browser, so it forked https://files.md and set it up on my home server pointed at my sync folder.
I was of the mindset that I’d not want to self host my day to day tooling, happy to pay, but it seems having the ability to run open source tools where I can say “just fix it” is too convenient to ignore.
Why not? Notes are incredibly private data. One of the top priority items to NOT host in a cloud.
Trilium is incredible.
Consider it.
Also AGI does not exist and never will.
That's hilarious.
Some of the "alternatives" aren't even close. All but one of the Spotify/Apple Music replacements are self-hosted things where the user needs to provide the audio files. That's nothing like the near-infinite jukebox experience that Spotify and such offer. The only one that isn't self-hosted/BYOM is basically just an app-only (no web) front end for free YT music or you need to pay for the underlying streaming services anyway.
I think the SaaS is dead meme is really just a “SaaS can’t be lazy anymore”. The days of offering a product that technically does something, but is incredibly poorly designed, or not well integrated, or unsupported — are over.
If your SaaS doesn’t support both Google and Microsoft; macOS and Windows; iOS and Android; automated provisioning, complex/fine-grained RBAC, customisable onboarding.. You’re dead in the water.
The moat of features has evaporated, and your only defence is good UX, stability and TCO.
AI isn’t eating SaaS. It’s eating shitty businesses.
Password Manager SaaS that have been breached are dead.
Auth providers that scale their pricing model to their investors growth expectations are dead.
Platforms and technical products that did something a script could do, with little support for real business software, ecosystems and environments are dead.
AI isn't even eating most shitty SaaS businesses, because it's a weird HN fantasy that your hairdresser or car mechanic are itching to manage agent swarms and risk losing their business if that unnecessary adventure goes sideways. They would much rather pay several hundred bucks for a SaaS product that has a bunch of UI annoyances.
It's likely that there will be an explosion in vibecoded competitors to existing SaaS products, but I wouldn't underestimate the importance of familiarity and stability. I use a bunch of SaaS stuff for my business and it's not like I do a weekly comparative analysis to see if switching to some unfamiliar platform that popped up overnight could save me a bit of money or time. It's just not worth it.
Replacing something has cost involved and doing it too often will kill any savings. Which matters when you actually need to stay profitable.
Once a business grows where tool failure is not a option, they will buy salesforce or similar professional clones.
Now will salesforce be disrupted by a newer CRM called AIForce? Now your paying AIForce instead of paying Benioff.