• The Inside Track
  • Posts
  • #7: Memory is the Moat & Using Projects to Organise your Work with AI

#7: Memory is the Moat & Using Projects to Organise your Work with AI

One idea and one tactic to get ahead in the AI race.

April 24, 2025

Hey AI explorer — welcome, especially to new subscribers! Most people still sleep on how much leverage smart use of AI gives us. That’s our edge.

Each week, I share what I’m testing, learning, and building — all through the lens of spotting new opportunities with an AI-first mindset.

Let’s jump in:

💡 One Idea

Memory is the Moat

Memory will be the kingmaker in AI as the technology continues to evolve.

Here's why it matters:

The base layer of AI is becoming commoditized. You can now access powerful assistants from OpenAI, Anthropic, Google, and Meta with relatively similar capabilities.

The differentiator? Memory.

As an AI assistant learns your preferences, work patterns, and specific use cases, it builds a personalized knowledge graph that compounds in value. While general AI delivers horizontal intelligence by default, memory creates vertical specialization that deepens with every interaction.

This shift will trigger two massive business consequences:

  • Unprecedented switching costs — Once your AI has accumulated months of personalized context, changing providers becomes increasingly costly

  • Authentication as leverage — Expect "Sign in with OpenAI" (or similar) to emerge as the new gateway for products offering deep personalization

This memory layer will eventually permeate everything we do online, with e-commerce being perhaps the most lucrative implementation.

If you want to understand how the AI landscape will evolve, follow the memory.

🛠️ One Tactic

Using Projects to Organise your Work

In ChatGPT and Claude, Projects let you organize chats with dedicated context, file access, and custom instructions. This transforms AI from occasional helper to strategic collaborator.

Three powerful applications:

  • Business Coach – Upload financials and strategy docs for tailored analysis and recommendations.

  • Writing Partner – Add reference writing and use meta-prompting to create instruction sets matching your style.

  • Product Lab – Consolidate user feedback and specs for a thinking partner that remembers your product's evolution.

Each conversation builds on previous ones, creating compounding value.

Projects are available on paid subscriptions only.

💡 Need personalized help? Book a call to explore how AI agents and automation can unlock efficiencies and accelerate growth in your business. Click here.

Thanks for reading the Inside Track. You can subscribe for free below:

That’s it for today. Thanks!

– Martin