Intro and TL/DR

Hello, and welcome to our Q1 2023 report. This quarter has been the most exciting one yet for our team, and we can't wait to share everything with you. First, let's talk about AI. As we stated in our previous report, we were well-prepared for the AI revolution. However, we were not quite ready for the number of startups that emerged like mushrooms in its wake. We anticipate that the number of deals we consider will continue to increase next quarter as well, so we continue adding new team members and Deal Captains to stay on top of our game.

This quarter, we invested just over $1M in 11 deals - 8 new and 3 follow-ons. Our portfolio now has 49 companies (34 in AI/ML space). Our strategy of investing in repeat founders alongside bigger firms still looks excellent in today's fast-changing environment. This approach allows us to identify the most promising companies and double (and sometimes triple) down on the biggest winners. It is still too early to tell, but we’re seeing great early signs: our follow-on deals are made alongside prominent investors such as the Open AI Fund and Lachy Groom, as our first-time deals are closely tracked by many bigger VCs such as a16z, 8VC, and Matrix Partners.

This quarter we have also welcomed a record number (14) of new investors into our community and grew our quarterly fund size to $1.8M, which is well on our way to reaching $2.5M by the end of the year.

We have hosted several AI events that help us stay at the forefront of the early-stage community, meet new founders, and increase the exposure of our portfolio founders to top VC players. Our All-AI-alignment Dinner for founders and scholars and the Crosspolin-AI-tion Soiree for founders and VCs have been both received incredibly well. They have allowed us to build even stronger relationships with partners of other VC firms, further improving our access to the pipeline and our portfolio founders' ability to raise funds. With the help of our partners from OpenAI, Microsoft, and Google, we plan to make these events a regular occurrence.

There are also lots of updates on our internal systems and products, a new playbook for Deal Captains, a new website, and new merch that you should get. You can also look at photos from our epic LP event and learn about the upcoming activities. But first, let’s talk about AI and its effect on VC.

The Effect of LLMs on Investing in SaaS Businesses

The technological progress and the boom of cloud solutions have successively collapsed the costs of connecting, computing, and storage. Now LLMs can do the same to the cost of the software itself.

We believe that one of the most significant impacts of Large Language Models (LLMs) will be on how people create software. Co-pilot solutions improve the productivity of engineers by autocompleting their code and unit tests (like our portfolio company Machinet), turning a 10x engineer into a 50x engineer. Additionally, LLMs can help these engineers navigate aspects of their profession that they may not be familiar with. This allows people proficient in front-end development to build the back-end and vice versa.

But LLMs are not only for 10x engineers. We expect a new revolution in no-code and low-code development, as it will be increasingly less difficult to build complex solutions using natural language. As Andrei Karpathy said, "The hottest new programming language is English."

This has a profound effect on how software startups are built and evaluated. Traditionally, SaaS investors have looked for five types of moats that could protect startups from competition:

1. A strong brand or unique access to channels or distribution. Anything that allows the company to acquire new users more cheaply than competitors.
2. High switching costs, especially when paired with a first-mover advantage.
3. Technological moats such as software complexity or unique IP; some software can be difficult and costly to replicate.
4. Network effects and data flywheels: As more users adopt a product, either its value for new users increases directly, or the product can benefit from getting more user data.
5. Economies of scale or capital intensity: Some businesses need to be huge to be profitable, and some business models, like marketplaces in mature markets, can only be built by significant investments in either chickens or eggs.

Since most of these moats are unachievable for companies early, early-stage VCs have primarily focused on technological moats and early signs that a startup can solve distribution, create a strong brand, and/or have a network effect.

It appears that the era of tech moats has come to an end. Technology is driving down every type of cost on startups' PNL statements, which allows more players to enter the field and compete. The only cost that is steadily increasing is the cost of user acquisition. As a result, it's time for startups to adopt a new SaaS business strategy.

The key to success in this new paradigm is being closer to customers than their competitors. Engaging and retaining users, as well as creating high switching costs, will be the factors that set businesses apart. Creating and leveraging network effects and data flywheels will be much more important than most technological know-how. Distribution is the new SaaS moat that investors will be looking for.