FiveTech Support Forums

Claude Code to the Max – and what is our role here?

In a very short time, we will be faced with an enormous amount of programs and apps – and often find ourselves not really knowing what to do with them. The biggest challenge will soon no longer be programming itself, but finding real testers and paying customers in this flood of software.

The huge interest in the "Claude Code" webinar clearly shows where we stand in 2026: AI tools are taking over the tedious tasks and generating code at an incredible pace. At the same time, the classic Q&A threads ("How do I fix this bug?", "Why is this API throwing an error?") are slowly dying out. That makes perfect sense, because AIs can solve exactly these issues in a matter of seconds today.

This forces us as a community to face an exciting question: If we as a forum will no longer be a mere "Q&A archive" for coding problems – what will we be?

I believe we need to start thinking much more systemically. When AI writes the raw code, we developers automatically move one abstraction layer up. Our discussions here in the forum should reflect exactly this development. Instead of syntax, we should focus more on the big picture in the future:

• Infrastructure & Architecture: AI generates code building blocks at breakneck speed. But how do we deploy them into production safely, scalably, and maintainably? How do we orchestrate systems that are becoming increasingly complex?

• Business Solutions instead of Code Snippets: It is no longer primarily about how to program a function, but why. How do we translate an ambiguous business problem into a technical solution that truly provides value to the company?

• Customer Contact & Requirements: An AI cannot advise a client who suddenly changes their requirements in the middle of a meeting. How do we figure out what the customer really needs before we even launch the AI?

Exactly here, on this macro level and in the search for market relevance, we need human creativity. And that is the huge advantage this forum can offer. Precisely because we are a worldwide group, we have access to countless markets, networks, and different perspectives.

Let's use this combined human potential to help each other with marketing, to put use cases through a global reality check, and to find answers together on how to successfully bring our software to the market.

Hi Otto,

I think part of the problem is that a lot of us learned on the fly, empirically. We never had the rigor to build software the proper way, so we improvise. And AI makes improvising much faster than it makes discipline faster.

Our sin is that instead of first designing a folder structure, a file list, and giving the AI the rules to follow, we just start coding with no real direction, based only on whatever the model was trained on. Then we end up with a pile of files and we don't even know what half of them are for. Something like an MVC layout should be the normal way to build, even when the AI writes it, so we still know where everything lives. Same reason we use separate files and separate classes instead of one big blob.

So the question to ask first is: what is the best way to implement this? Ask that before the AI implements anything.

The way I see it, AI can make an app for anyone in minutes, nice and dandy. But a business app is another thing. It needs specifics, and those come from knowing what is needed (owner), how it must behave (user), how it must be implemented (developer), and which technology to use (developer and AI together). Only one of those four is the AI's part.

And that connects to your point about the market. If the requirements are the hard part, then what this forum is good for is not code answers, it's people in other countries who can tell us if the problem we are solving is one somebody will actually pay for. I'd rather have that than another bug thread.