The unlikely winner of the AI gold rush

PLUS: Dell's AI tipping point, a new agent benchmark, and Latin America's own GPT

In partnership with

Good morning, AI enthusiast.

The AI gold rush has an unlikely winner: the tiny Caribbean island of Anguilla. The nation has struck digital gold, earning tens of millions from its country-code domain, .ai, as tech companies rush to secure AI-branded web addresses.

This windfall now accounts for nearly a quarter of the government's total revenue, fundamentally diversifying its tourism-dependent economy. What does this unexpected boom signal about the future value of digital real estate in the AI era?

In today’s AI recap:

  • How to build an Gmail AI Agent using n8n (watch here)

  • Anguilla’s surprising .ai gold rush

  • Dell’s AI datacenter tipping point

  • A new benchmark for testing AI agents

  • Latin America's culturally-aware GPT

  • 8 trending AI tools

Anguilla's AI Gold Rush

The Recap: The tiny Caribbean island of Anguilla has struck digital gold, earning tens of millions from its country-code domain, .ai, as tech companies rush to secure AI-branded addresses. This unlikely windfall is now a core part of the island's economy.

Unpacked:

  • In 2024, domain sales generated $39 million, making up nearly a quarter of the government's total revenue and diversifying its tourism-dependent economy.

  • The number of registered .ai domains has surged past 850,000, a huge leap from fewer than 50,000 in 2020, with premium names auctioning for hundreds of thousands of dollars.

  • To secure this windfall, Anguilla has partnered with Identity Digital to move its registry to a global server network, ensuring stability against local disruptions like hurricanes.

Bottom line: Anguilla's story is a powerful example of how digital real estate can create immense value for small economies in the tech era. It also solidifies the .ai domain as a clear signal of a company's focus within the artificial intelligence landscape.

The Key to a $1.3 Trillion Opportunity

A new trend in real estate is making the most expensive properties obtainable. It’s called co-ownership, and it’s revolutionizing the $1.3T vacation home market.

The company leading the trend? Pacaso. Created by the founder behind a $120M prior exit, Pacaso turns underutilized luxury properties into fully-managed assets and makes them accessible to the broadest possible market.

The result? More than $1B in transactions and service fees, 2,000+ happy homeowners, and over $110m in gross profit to date for Pacaso.

With rapid international growth and 41% gross profit growth last year alone, Pacaso is hitting their stride. They even recently reserved the Nasdaq ticker PCSO.

The same VCs that backed Uber, eBay, and Venmo also backed Pacaso. Join them as a Pacaso shareholder before the opportunity ends September 18.

Paid advertisement for Pacaso’s Regulation A offering. Read the offering circular at invest.pacaso.com. Reserving a ticker symbol is not a guarantee that the company will go public. Listing on the NASDAQ is subject to approvals.

AI Tools of the Day

  1. 💻 Aider - Pair program directly with LLMs in your terminal to edit existing codebases using the power of git integration.

  2. 🎮 nunu.ai - Deploy multimodal AI agents that see and interact with games just like human players to revolutionize QA and testing.

  3. 🌐 Airtop - Build powerful AI agents that can automate complex web tasks in real browsers, effortlessly handling logins and dynamic content.

  4. 🎨 Instant Character - Generate perfectly consistent character designs from a single reference photo, giving you precise control over your AI-generated art.

  5. 🐞 Metabob - Uncover and fix complex bugs that traditional static analysis tools miss by using an AI that understands the logical context of your code.

  6. 🎭 Charisma.ai - Power emotionally-aware virtual characters that can hold realistic voice conversations, creating truly interactive stories and experiences.

  7. Latent AI - Optimize and deploy sophisticated machine learning models efficiently on resource-constrained edge devices like sensors and cameras.

  8. 📲 InboxBites - Transform your cluttered newsletter subscriptions into a swipeable feed of 50-word insights to consume 10x more content in less time.

Explore the Best AI Tools Directory to find tools that will 10x your output 📈

Dell's Datacenter Tipping Point

The Recap: Driven by a massive surge in AI server demand, Dell’s datacenter division has officially outgrown its iconic PC business for the first time. The company reported a staggering $8.1 billion in AI server sales for the quarter, marking a major shift in the hardware landscape.

Unpacked:

  • Dell’s Infrastructure Solutions Group (ISG) hit $16.8 billion in quarterly sales, a 44.3% year-over-year increase, while its PC division posted $12.5 billion.

  • The growth engine is AI-specific hardware, with sales jumping 2.6x to $8.1 billion, fueled by major deals with customers like xAI and CoreWeave.

  • Despite the revenue boom, ISG’s operating margin has dipped, suggesting that while Dell is selling more hardware, partners like Nvidia are capturing the lion's share of the profits.

Bottom line: This milestone shows how enterprise AI demand is fundamentally reshaping the tech hardware market. The gold rush is on for building AI infrastructure, and traditional hardware giants are adapting to become the critical suppliers.

AI Training

The Recap: In this video, I'm going to show you how to build an AI Gmail chatbot agent for local businesses using n8n that is able to respond to customer questions and inquiries around the clock, 24/7. This is perfect for businesses that don't have the resources to hire a full-time employee to answer these questions, and it's going to allow them to lead to more sales, less refunds, and less disputes.

P.S. We also launched a free AI Automation Community for those looking to build and sell AI Automations — Come join us!

The Agent Benchmark

The Recap: A new open-source benchmark called Tau² offers a new way to test how well AI agents use external tools to complete real-world tasks. OpenAI recently highlighted the benchmark to showcase the advanced agent capabilities of its GPT-5 model family.

Unpacked:

  • Tau² works by simulating conversations between two LLMs: an “Agent” that must use tools like APIs and databases, and a “User” that presents a problem within domains like Airline or Telecom.

  • The evaluation goes beyond simple success or failure; in some tests, another LLM acts as a judge to evaluate if the agent’s communication was clear, helpful, and aligned with human expectations.

  • More than just a leaderboard, its methodology provides a blueprint for how to reliably test and validate the performance of increasingly complex and non-deterministic agentic systems.

Bottom line: This benchmark signals a move away from measuring simple model accuracy toward evaluating an AI’s ability to reason and execute complex, multi-step actions. This gives developers a much-needed framework to build agents that are not only powerful but also reliable in practical applications.

Where AI Experts Share Their Best Work

Join our Free AI Automation Community

Join our FREE community AI Automation Mastery — where entrepreneurs, AI builders, and AI agency owners share templates, solve problems together, and learn from each other's wins (and mistakes).

What makes our community different:

  • Real peer support from people building actual AI businesses

  • Complete access to download our automation library of battle-tested n8n templates

  • Collaborate and problem-solve with AI experts when you get stuck

Dive into our course materials, collaborate with experienced builders, and turn automation challenges into shared wins. Join here (completely free).

Latin America's Own GPT

The Recap: A new collaborative initiative called Latam-GPT is building an open-source large language model trained specifically on Latin American languages, history, and cultural contexts.

Unpacked:

  • The project aims to solve the cultural bias of existing models, which often lack regional relevance and use examples that don't resonate with Latin American users.

  • Backed by a $10 million investment, the model is being trained on a supercomputer in Chile equipped with NVIDIA H200 GPUs and will launch with 50 billion parameters.

  • Future plans include incorporating indigenous languages like Mapuche and Guaraní, creating a family of models that other institutions can adapt for specific sectors like education or agriculture.

Bottom line: This initiative empowers the region to develop its own AI tools, reducing reliance on models built elsewhere. It also serves as a powerful blueprint for creating more culturally and contextually aware AI on a global scale.

The Shortlist

BlackRock developed its own agentic AI platform, Asimov, to gather research, monitor markets, and scan company filings in real time, reshaping how its leadership makes strategic decisions.

Alpha School opened enrollment for Virginia's first AI-powered private elementary school, where students use AI-guided academics for two hours daily, with tuition set at $65,000 per year.

Google reported that a median text prompt to its Gemini system uses just 0.26 milliliters of water, highlighting the massive efficiency gains and environmental impact differences between AI models.

Meta announced it will add new guardrails to its AI chatbots to stop them from discussing sensitive topics like suicide and self-harm with teen users, instead directing them to expert resources.

What did you think of today's email?

Before you go we’d love to know what you thought of today's newsletter. We read every single message to help improve The Recap experience.

Login or Subscribe to participate in polls.

Signing off,

David, Lucas, Mitchell — The Recap editorial team