Legendary venture capitalist Don Valentine said “Great markets make great companies” and the market has proven him right time and time again. I’ve lived it myself, having joined companies pre-version-1.0 as well as working for mature leaders like Oracle and Business Objects.
Watching the pace of evolution in AI right now takes me back to early days of the web (more on this later). And much like those early days, the potential breadth and depth of new applications creates all kinds of interesting infrastructure challenges to solve. Anyscale is at the forefront of solving those challenges.
People slice it differently and there are plenty of sub-categories in the broader AI infrastructure space. But the overall AI infrastructure opportunity is enormous and growing quickly. Think about the investments that the cloud hyperscalers have made. It’s not unusual for them to pursue new markets where they see demand - offering a wide range of platform services, specialized database engines, and more. But for AI, Amazon, Microsoft, and Google are all producing their own specialized chipsets. They didn’t do that for Web 1.0 or 2.0, cloud computing, big data, or other multi-billion dollar markets, but they’re doing it for AI.
Positioned to Win
Enormous, disruptive market opportunities attract lots of attention and investment. There’s a global army of startups lined up alongside established leaders all trying to get a piece of the market by solving high-value user problems.
While we have an enormous amount of work to do, Anyscale is well-positioned for success in this environment.
Thriving open source communities are a strong signal of product-market fit, and Anyscale has that with the community around the company’s Ray project. The Ray community grows every day, adding users, contributors, and committers from academia, other open source communities, and large and small commercial companies. That drives market fit and product velocity, and can create a global fan-base that propels the technology as well as the category.
For example, I was so lucky in my last job at Neo4j to get to work with a global community of developers that did amazing things with the company’s graph database technology. It was 3 years ago now that the Covid-19 pandemic hit globally and I still get the chills when I look at how Neo4j’s open source community stepped up to build apps for contact tracing, drug repurposing, office reopening, supply chain analysis and more.
Production usage for business-critical applications is also a strong signal and it has been amazing in my first couple of weeks to learn about how companies like OpenAI, Instacart, and Spotify all rely on technology from Anyscale to power their ML applications. As an occasional Product Manager in prior professional lives, I also appreciate how those users push your technology and team and make you better, faster.
The Anyscale team is impressive by every measure. Powerhouse academics and industry veterans have embraced a “win as a team” culture that you can feel. In just my first 2 weeks, I’ve gotten help from Engineering, Community, Finance, each Founder, Sales, and more. You know amazing things are possible when you see that kind of talent working together at startup speed for a common purpose.
It’s Still Early, so Buckle Up
In pondering where AI goes from here, I can't help but remember the early days of Web 1.0….
When Amazon was an online bookstore. Yes, really.
When it made national news when a person sold a car(!) on the internet!
When people would look at you funny if you said “I’m dating someone I met…..online”
When a little dot-com company called Salesforce.com decided to put business applications online and offer contact management and salesforce automation via a web browser.
Now Amazon is a massive consumer and enterprise business, everything imaginable is sold online, online dating is pervasive to the point of being the “default” for many, and Salesforce started a SaaS revolution that forever changed the way companies buy and use software..
That’s where we are with AI. As impressive as GPT-4 is, it’s a matter of time, and not much time imho, until we look back on it as an important step towards a myriad of amazing and sophisticated use cases. We don’t know who the ultimate AI application winners and losers will be, or how AI applications and use cases will evolve, but we know they’ll need infrastructure, and that’s great news for Anyscale.
I want to be a part of that, and that’s why I joined Anyscale as Head of Marketing. I couldn’t be more excited to be a part of the team, community, and evolution of AI.
PS if you’re a vendor and want to sell me something, I’ve got your info, and I’ll reach out if I need you. Resist the urge to ping me - thanks.