📊 Vibe-Coding Is the New Excel

Hi, and happy Tuesday.

Recent advancements by Anthropic’s Claude and Google’s Gemini is creating buzz in AI circles that vibe-coding will replace software.

This viral tweet says it all:

We - at PreScouter - don’t entirely disagree. 

Coding models have reached a point where non-engineers can build meaningful internal tools.

That’s why, starting this week, we’re training all our staff on vibe-coding.

But the notion that “everyone will become a software developer” is not the reality we are seeing, and if you’ve spent time inside a large organization, you’ll recognize the tension immediately.

Many highly capable professionals still don’t want to:

• Build reports in Salesforce

• Join tables in a BI tool

• Touch anything that smells like “technical work”

Not because they can’t - but because they don’t have the interest or don’t want to take on the cognitive load.

Vibe-coding lowers ability barriers, not interest barriers.

What we’re seeing in practice is three persistent roles:

• Builders (a minority): These are the modern power users - the Excel wizards of the AI era. They enjoy shaping systems and workflows. AI turns them into force multipliers.

• Delegators (a majority): They don’t want to build and the responsibility of maintaining software. But they do want custom outcomes. Their interface to vibe-coding is an agent - or a colleague who knows how to use one.

• Recipients (another majority): They want a link, a dashboard, a button that works. For them, software should be stable, opinionated, and invisible.

This is also why non–vibe-coded software will always exist. In regulated, revenue-critical, or high-risk environments, people want guardrails, defaults, and accountability - not infinite flexibility.

We also believe some enterprises, or teams within them, will look to partners - such as PreScouter - to play the role of the builder.

Ultimately, we think of vibe-coded apps as the new Excel. 

However, enabling vibe-coding safely inside an enterprise requires discipline.

For example:

• Front-end AI-generated apps cannot talk directly to sensitive systems like Salesforce or SAP

• Credentials, identity, and access controls still matter

• Without a proper architecture, speed creates risk

We’re working on a secure “Universal Bridge” - a backend layer that centralizes access to sensitive data and credentials access to it using Identity-Aware Proxy (the user’s company login).

I’m sharing this story to give you a sense of how AI adoption is shifting.

It’s now about intentionally designing how AI is used, who builds, and how fast ideas can safely move from intent to impact.

That’s the work we’re doing - and the work we help leaders do every day.

If you want to give vibe-coding a whirl, for a personal project, there are many tools available. I recommend trying Google AI Studio using your personal Google account. 

https://aistudio.google.com/apps?source=showcase&showcaseTag=featured 

Best,

Dino