By: Francois Aubin.
Back in 2011, Marc Andreessen wrote his now-famous essay, Why Software is Eating the World. At the time, it was a bold prediction. Andreessen foresaw a future where software wouldn’t just support businesses — it would become the business. Traditional industries were being reimagined as software-driven platforms. And he was right.
Think about it: Blockbuster was devoured by Netflix, which turned video rental into a software service. Kodak collapsed under the rise of digital photography and photo-sharing apps. Even the smartphone became a software-first device. When Steve Jobs introduced the iPhone in 2007, he didn’t just embed an MP3 player into a phone like Nokia did — he wrapped powerful software around it, enabling a whole new app-driven ecosystem.
From ride-sharing and logistics to entertainment and finance, the past two decades have shown us that software can consume and reinvent entire industries. The apex of that transformation? Fully automated factories and self-driving cars like those made by Tesla — machines where software literally drives the world.
But Now, Something New is Happening:
AI is Eating Software.
Today, we’re watching a second-level transformation unfold. If the first wave was software replacing analog systems, this new wave is AI replacing software — or more accurately, absorbing it.
Take the example of Customer Relationship Management (CRM) systems. Traditional CRMs like Salesforce or HubSpot provide structured interfaces for tracking leads, sales funnels, customer interactions, and so on. They are tools — complex, multi-screen environments that require teams to update fields, schedule reminders, and generate reports.
Now enter AI.
Instead of manually interacting with a CRM through forms and dashboards, AI models like those from OpenAI, Anthropic, or Meta can now be used as natural language interfaces — a conversational layer between you and the data. The AI can parse emails, meetings, and messages to automatically populate and manage CRM fields. You can ask your CRM assistant questions like:
- “Who are our top leads this month?”
- “Summarize our last call with Acme Corp.”
- “What’s my pipeline looking like this week?”
Even more importantly, these models can connect directly to your calendar, email, contacts, and other tools, enabling autonomous updates and task creation. Tools like OpenAI’s ChatGPT with plugins and actions, or Anthropic’s Claude with integrations, make it possible to operate an entire sales workflow without traditional CRM interfaces.
This isn’t just automation. It’s absorption. The AI is doing the job that used to be spread across multiple apps, dashboards, and processes — and doing it in a way that feels natural and intelligent.
What’s Next?
The logic is recursive:
- Software ate analog.
- AI is now eating software.
What comes next may be AI autonomously constructing, refining, and replacing software itself — not just acting as a layer on top of it. As GitHub Copilot and GPT-4o already hint, AI is becoming an integral part of software development. The tools that used to require engineers to build may soon be conceived and assembled by models themselves.
We’re not at the end of this story. We’re in the middle of a new chapter — one where tools disappear and intelligent behavior emerges. The interface is no longer a window. It’s a conversation.
References
- Andreessen, Marc. Why Software is Eating the World. The Wall Street Journal, 2011.
- Isaacson, Walter. Steve Jobs, Simon & Schuster, 2011.