Thank you for subscribing to our content. You will now receive the latest news in your inbox.

Accessible AI
October 04, 2022

Launching With Conviction

Occasionally, a technology comes along that changes everything. AI is that kind of foundational technology. We believe this so deeply that we’re designing an entire organization around it. Conviction is purpose-built to serve “Software 3.0” companies, and to help them achieve their potential as the most important companies of tomorrow. 

We’ve raised a $100M first fund to back extraordinary founders building these enduring companies.

A Little Backstory

My parents are founders of a company, and of a new life for our family. My father immigrated from China to Wisconsin in 1987 when he was 24 years old. My mother joined him a year later. Shortly after, I was born. They both worked at Bell Labs/Alcatel-Lucent out of grad school. They are both engineers turned entrepreneurial businesspeople. They both barely spoke English upon arrival. They are two of the bravest people I know.

I quite literally grew up inside their tech company – their head of hardware engineering supervised my fumbling build of their first website when I was 14. That website is exactly what you might picture from a budding teenage developer seeped in early 2000’s web aesthetic and thrilled about gradients. I grew up understanding that company-building is Type II fun. “Sleeping over” in the office for bug bashes and building diet coke-can castles with the CTO was my idea of a good time. I remember worrying about my parents being stressed about the innumerable moments of crisis: a big deal falling through, a mammoth competitor trying to murder us with litigation, a product reliability issue, a key hire disappointing. I remember about the number of times we all thought the company wasn’t going to make it.

My (abnormal) adolescence forged my view of founders and startups – it gave me the deepest respect for the courage and resilience required to build an enduring company, the permanent expectation that everything is going to be hard, the pervasive sense of fear that it’s just you and your small band of pirates against Goliath and the uncaring world at large. Your only advantages are speed, undeniably good product, a technical thesis*, risk appetite, cleverness, the focus that comes from always being in wartime. And sometimes, insanely, that is enough.

More recently, I was lucky to spend almost a decade incubating and investing in startups at Greylock Partners, the very first - and still a legendary - generalist venture capital firm. I worked with thoughtful, experienced, well-connected partners. I had the privilege of learning what made great businesses spring from nothing: first in my initial zones of comfort, infrastructure and security, but then across consumer and enterprise, fintech and crypto. I collaborated with inspiring founders at all stages, but discovered I was, at the core, a glutton for punishment. I loved the challenge of finding product-market fit, and the essentialism of early-stage team building. In the course of hundreds of late night whiteboard sessions, Sunday morning meetings, moments of crisis (and celebration of company milestones) – I built conviction in the kinds of companies I cared about, and the way in which I would partner with them.

Software 3.0

Innovation happens inexorably, but at uneven speeds. After years of nothing foundationally new happening (outside of crypto) we are at a special moment in technology history. There has been radical, accelerating progress in machine learning over the past decade, with unexpected core breakthroughs: the shift from classification to generation, and the unreasonable effectiveness of ever-larger, transformers-based general models.

We believe the use cases for more general intelligence are, well, general, and will expand over the coming years to every industry. The dominant consumer company of today (TikTok) rethought the content experience with personalized AI. We expect the dominant companies of tomorrow to rethink every industry with intelligence. We are excited to invest up and down the intelligence stack: in chips, data infrastructure and developer tools, research labs, and AI-native applications in every vertical. We believe that better tooling, cheaper and more abundant compute, more data, more open source models, and more AI-educated talent will lead to ten thousand experiments. We believe (despite the recent excitement) that AI's impact is still underestimated.

Arming the Builders

These intelligent “Software 3.0” companies will both need to be great software businesses in the traditional sense (and can thus benefit from great networks and tribal knowledge), and will also be something entirely alien. If Software 1.0 was about human-written code, and Software 2.0 is about dataset labeling, Software 3.0 will be about manipulating foundation models.

Conviction is a new kind of investment firm that aims to bridge that gap, by helping extraordinary founders harness what is enduring, and designing from a clean sheet of paper a firm purpose-built to enable AI-natives. Here are a few of the ways we expect to operate differently:

  1. We are hands-on. That means operational contributions measured in outcomes (talent, customers, data and compute access), as well as leading operators and technologists in-network and in-house.
  2. We are deep in our domain. AI-native companies encounter entirely novel challenges. We are building the new playbooks together – for both product challenges (gracefully handling the error matrix for a model, designing product to manage hallucinations or abusive prompts, reducing per-customer fine-tuning) to the commercial (packaging and selling category-creating products that replace human labor). We expect to learn with our companies, but always be on the bleeding edge.
  3. We will create community, and contribute to our community. Conviction will serve as a Schelling Point for the brightest minds in AI. We care about industry progress. We know there are more great companies to be built than we have capacity to back.
  4. We are long-term oriented, true believers. AI has come into fashion before, and also gone through troughs of disillusionment. We expect this to continue. We expect to take technical risk. We believe the prize is worth the patience.

Finally, we believe that AI is crucial – not only for our own efforts but for our nation and for our global society. There are important policy and public opinion considerations in privacy, information integrity, national security, safety. We will partner with our portfolio and the ecosystem to contribute to these discussions. We believe AI can be a force for abundance, good, and compounding innovation; we have a responsibility to invest in enabling that.

First Steps and First Principles

Our LP base includes ~50 Founders, CEOs and COOs of the most important companies of today, who believe that AI is critical to their businesses’ success tomorrow. We are grateful to our early institutional investors, and to our individual entrepreneurial backers as we take our first steps.

Conviction is a distinct choice of name for a venture firm, and intentional. Most venture firms are designed for optionality, not conviction. However, we do not believe in a one-size-fits-all approach to venture investing. When we have conviction´╗┐ that a technology market is generational, we design and build an entire organization around it from first principles. The first market for us is Software 3.0.

If you are a founder or investor and want to shape the future of AI, please reach out. We are growing our team and always on the hunt for people starting enduring companies. We also welcome interested people to apply for our programs: Communities (our event series), Conviction Fellowships (for young talent), Visiting Researchers (full and part-time), and Visiting Builders (operators and founders, interested in AI).


*Part of their technical thesis was (and is) more intelligent networks.