The infrastructure side is what makes this really interesting, once AI agents can actually run parts of a business, security, permissions, and human oversight become just as important as the AI itself, i think that’s where the real challenge will be.

AI today is becoming a bigger part of everyday life and work, it’s not just about chatbots anymore, it’s helping people automate tasks, make decisions, and get things done faster, the pace of change is honestly pretty interesting to watch.

One thing that really stands out is how much harder choosing SaaS is becoming, it’s easy to get distracted by AI features, but things like reliability, integrations, pricing, and what happens when the AI gets something wrong matter just as much, i think the best reviews will be the ones that show how a product actually performs in real workflows, not just what it claims on the pricing page.

The ‘extra team member’ idea really makes sense here, when software takes care of all those small tasks you’d otherwise have to keep remembering, it doesn’t just save time it removes a lot of the mental clutter too, that can make a pretty big difference in how smoothly an agency runs.

This is where I think AI gets really interesting, a flashy demo might get someone interested, but if the tool consistently saves time or prevents costly mistakes, that’s what makes people keep using it, the boring, reliable AI that solves a real problem may end up being much more valuable than the impressive demos.

The distinction between adding AI and actually becoming AI native is really important, simply putting an AI feature on top of an existing product doesn’t necessarily change how the product works, the bigger shift is making AI part of the workflow so it can actually take action and deliver results, rather than just assist the user.

The part about coding being the easy part is so relatable, it’s usually all the little things around the product payments, auth, emails, edge cases that end up eating the time, i’d be really interested to hear what experienced founders found themselves rebuilding or fixing after launch.

I think the shift toward evaluating software by what it actually gets done makes a lot of sense, a long feature list can look impressive, but if the team still has to do most of the work manually, those features don’t mean much, i’d much rather look at how much time and effort a tool actually saves.

The shift from selling software to delivering actual results is probably the biggest change here, i can see a lot of traditional roles being reduced or completely changed, but I think the companies that use AI to support people rather than just replace them will have a better chance of making the transition work.

The part about software becoming the worker itself is what really stands out to me, once that happens, pricing based on the number of people using the software starts to feel less connected to the value being delivered, i can see hybrid pricing becoming the sweet spot, especially for companies that still need humans involved in the workflow.

I think the point about not automatically replacing existing SaaS is easy to overlook, there’s a lot of excitement around AI native tools, but if the new platform creates migration headaches or breaks workflows that already work, the trade off may not be worth it, sometimes adding AI to the existing stack is actually the more practical move.

One thing I’d add is that integrations can make or break an AI tool, it’s easy to look at how much a platform can automate on its own, but if getting it connected to the rest of your stack takes a lot of extra work, some of those time savings disappear, i think the best alternatives will be the ones that fit into existing workflows without creating another layer of complexity.

I think that cognitive load point is really important, when something breaks, the last thing you want is to spend more time figuring out the monitoring tool itself, if Better Stack can make it easier to go from an alert to actually understanding the problem, that’s probably where the long-term value shows up.

The comparison between enterprise focused platforms and simpler tools is really interesting, i think the biggest thing is that the “best” alternative depends on how much automation a business actually needs, a smaller company might get more value from something simple and easy to deploy, while a larger support team may need deeper integrations and more control, it’s less about having the most features and more about finding the platform that fits the workflow without adding unnecessary complexity.

One thing I’d add is that knowing when not to act is just as important as being able to take action, an AI that can handle simple issues on its own but quickly recognizes when a human is needed can probably be much more useful than one that tries to automate everything, that balance is what would make me trust it in a real support environment.