My Journey in Radical Transparency

What are you the best at that you enjoy doing the most?

By

The Agentic Endgame: Shifting an Investment Thesis for the AI Platform Era

The tectonic plates of enterprise software are officially shifting. We are moving decisively out of the SaaS/Cloud Infrastructure era and into the Agentic AI Platform era.

For mid-market incumbents, the opportunity is huge, but it’s a true Innovator’s Dilemma. They must actively cannibalize their high-margin, sticky-tool revenue streams to build the actual solution layer. Most won’t. This creates the ‘white space’ for disruptive startups we look to fund.

Here are the biggest opportunities:

1- Shifting from Tools to Solutions
The core mandate has changed: Customers are no longer buying tools; they are investing in guaranteed outcomes.

For the last decade, SaaS has been a tax on the enterprise—paying per seat for unused features, siloed dashboards, and data summaries that still required expensive human analysis to act upon. That model is now structurally defunct.

We are actively seeking startups that completely abstract the interface layer and focus solely on solution delivery. No more “data visualization for decision support.”

The biggest initial challenge will be organizational change management. The first enterprises that can embrace the loss of the human-in-the-loop intermediary role will achieve exponential, asymmetric gains.

2- Defensible Business Models of the Agentic Era
A) Business in a Box or Service as a Software

Thesis: Deep vertical automation that achieve quantifiable results.

The Play: Building a closed-loop system that doesn’t just manage a process (like ERP) but executes the complete workflow from ingestion to resolution. The moat here is not the code, but the proprietary execution graph—the deeply engineered understanding of the business infrastructure necessary to guarantee a quantifiable result. This is about delivering a linked chain of automated solutions, not just a set of discrete features.

B) The Agentic Network Layer

Thesis: Creating a standardized, low-latency protocol for inter-agent communication and task decomposition.

The Play: We’re betting on the emergence of foundational network infrastructure. These platforms will enable individual, sovereign agents (owned by users) to securely expose the minimal necessary data to coordinate with other agents to solve complex, multi-step problems in real-time. This is the dynamic B2B marketplace of the future.

C) Data Co-operatives & Federated Learning

Thesis: Solving the data silo problem with aligned economic incentives.

The Play: Aggregating “dark data” that is individually non-material but collectively powerful. Think of specialized small data creators (SMBs, niche researchers) whose aggregated, high-fidelity data pool can train a model that no single entity could afford to replicate. The key is the shared value creation mechanism: A protocol that compensates data contributors via fractionalized ownership or micropayments, while charging solution providers for access. This creates a powerful data-moat flywheel that increases in defensibility with every new participant.

3- Outcome-Based Pricing (OBP)
The true signal of a mature agentic business model is the shift to Outcome-Based Pricing.

If you’re still charging a flat monthly fee per seat, you’re building a tool, not a solution. We invest in companies so confident in their execution engine that they can virtually guarantee the result.

Model: Fee = Value Created. If the defined outcome is not met, the customer pays nothing.

Initial Traction: This will immediately disrupt the incumbent SaaS market.

Future: This model will fully converge with the Agentic Network, where real-time micro-transactions—multiple per second—will be executed for solving fractionalized problems, replacing the entire concept of a “subscription” with dynamic, utility-based value exchange.

————————————————————————————————————-

This isn’t just an upgrade or a shift. It’s a fundamental architectural replacement that dictates a completely new investment thesis.

If you or someone you know are in San Diego and are interested in any of the above topics please reach out. We are focused on supporting the growth of this future in San Diego.

– This is the changing point for my writing. I had AI help writing the post. I started with a detailed transcript and then used the support for formatting and clarification of the concepts. It’s still all original ideas in one place, just cleaned up a bit from previous works.

By

We now have a name for it: Agentic

This is a blog post written by Claude that summarizes a keynote presentation that I made back on April 20th, 2024. At the time there were still so many unanswered questions. I’m not sure we have enough answers yet, but things are starting to come into focus. It’s crazy how far we’ve come. I’ll attach the Podcast that Google’s Notebook LM generated from this as well. Have a look and listen.

The Future of AI in Startups: Beyond Apps and SaaS

In a recent presentation to professors and entrepreneurs, a seed-stage investor shared compelling insights about how AI is reshaping the landscape of entrepreneurship and technology. Here are the key takeaways about where we’re headed and what it means for startups, education, and society at large.

The Current State: Big Tech’s Data Monopoly

We currently live in a world dominated by big tech companies built on an unwritten rule: if you collect and store data, you own it. These companies have developed increasingly sophisticated ways to gather our data, which they then monetize through advertising and services.

A recent example illustrates the value of this data: Reddit’s IPO filing revealed they generate $800 million in annual revenue, with $60 million coming from Alphabet specifically for AI training data. This presents a crucial question: Are we reaching a breaking point in our awareness of personal data’s value?

The Vision: Three Key Shifts

1. Personal Data Ownership

The future will enable individuals to own and control their data through personal AI agents that authenticate and allocate data access to services. Instead of giving complete access to service providers, zero-knowledge proofs will allow verification without exposing unnecessary personal information.

2. The End of SaaS

Traditional Software-as-a-Service models will evolve dramatically. Currently, most SaaS products are essentially different windows into similar data operations. The future will break these solutions into individual components that can be dynamically assembled based on specific needs.

3. An “Appless” Future

We’re moving toward a future without traditional apps. Instead of downloading separate applications and creating multiple accounts, AI agents will dynamically assemble solutions from component services. This shift could potentially return 30% revenue to industries currently paying app store fees.

Critical Challenges in Education and Human Development

Redefining Knowledge

As AI augments our ability to recall and access information, we must reconsider what it means to “know” something. When we can instantly access not just our own memories but potentially the collective experiences of others, how do we define knowledge?

Preserving Critical Thinking

Two uniquely human capabilities must be protected and developed:

  • The ability to measure information quality
  • The skill to identify what should be questioned These capabilities become increasingly crucial as AI-generated content proliferates.

The Future of Startups

Smaller, More Focused Teams

  • Companies will operate with smaller teams
  • Less physical space will be needed
  • Certain business functions may be eliminated entirely
  • Stronger culture can be maintained with smaller groups

Faster Iteration and Specialization

  • Development cycles will accelerate beyond current capabilities
  • Teams can focus on specific components of larger solutions
  • Compensation will be based on value added to the network
  • New funding models will emerge, including “Shared Earnings” approaches

Looking Forward

Sam Altman recently predicted we’ll soon see the first billion-dollar companies with fewer than 10 employees. Looking further ahead, there’s speculation about when we’ll see the first single-employee billion-dollar company.

This future requires significant infrastructure changes and a shift in how we think about business and technology. While AI brings tremendous opportunities, we must remember that humans are at their best when working together. The challenge lies in maintaining human connection while leveraging AI’s capabilities to augment rather than replace human creativity and critical thinking.

As we navigate this transformation, success will come to those who can balance technological advancement with human elements, focusing on quality measurement while preserving the irreplaceable aspects of human collaboration and creativity.

Listen to the Google Notebook LM Podcast Summary