There’s huge excitement around AI’s ability to write code. Tools like v0, Loveable, and Cursor are rapidly turning concepts into fully functioning products in days, not months.
But does this mean we’ll all become software developers overnight? Are traditional SaaS apps now obsolete, as we’ll simply prompt our personalized versions into existence?
Not exactly.
Imagine having a Star Trek replicator. Would you replicate a hammer and saw? Of course not—you’d skip straight to replicating the table you wanted.
Yet many are approaching AI precisely this way: replicating existing SaaS tools, rather than designing truly AI-first solutions.
Take CRM software. Today’s CRMs largely revolve around structured data entry: tracking relationships, logging activities, managing pipelines. But why replicate the tool, forcing manual data entry, when AI could analyze your calls, parse the internet for insights, qualify leads automatically, and even suggest exactly what to say?
Traditional SaaS tools focus heavily on structured inputs, data manipulation, and alerting. AI changes everything by handling massive, messy piles of unstructured data, distilling it into clear answers, and pairing those insights with AI agents to automate action.
The takeaway? Don’t replicate today’s tools. Take an AI-first approach and go straight to the results.
