I am an Investor at Sapphire Ventre’s early-stage fund, Saphire Sport, where I invest in early-stage companies at the nexus of technology and culture. Before joining Sapphire, I worked my whole career at technology start-ups and large technology companies such as Cisco and SAP.

Quick bio

Please tell us a bit about yourself. Your background, current role and how did you get started in AI?    

I am an investor at Sapphire Ventures. At Sapphire, we have three investment vehicles- Sapphire Ventures (a growth stage B2B focused enterprise fund), Sapphire Partners (early-stage fund investing), and Sapphire Sport (early-stage vehicle). I focus on investing out of our early-stage vehicle, which partners with startups that are consumer, prosumer, marketplace, or B2B2C focused. 

My experience in the technology industry spans multiple functional areas. I began my career working at start-ups and then held positions at Cisco and SAP. Throughout my career, I served in roles such as corporate development, sales strategy in APJ (Singapore), and corporate strategy for the Office of the CEO. I have also supported social entrepreneurs in Indonesia and Sri Lanka in developing and starting businesses.

Many of the companies I worked at dealt with large amounts of data so while ML was the near-term focus, AI was always a concept that was being explored and tested. With the advent of OpenAI, Stable Diffusion, Anthropic, and Cohere it seems we are at the beginning stages of a Cambrian explosion of AI-based technologies. ChatGPT in many ways opened people’s imagination and willingness to explore AI for their personal and business needs.

🛠 AI Work  

Sapphire recently announced its plan to invest $1B in generative AI companies. Could you please elaborate on your investment thesis and vision for the generative AI landscape?

Our $1B+ commitment was made by our growth fund, but we anticipate allocating capital to this space from all our investment vehicles. In the longer term, I believe AI will most likely become table stakes for your app just like the cloud is today. 

As it pertains to the announcement, our firm has recognized how AI is a technology paradigm shift for the enterprise tech landscape. We believe the opportunities for organizations large and small to benefit from AI technologies are massive and we want to help the most ambitious entrepreneurs build technologies that will define this next platform revolution in technology; AI will not only create efficiency and productivity gains for companies, but also unlock new revenue channels for organizations.   

One of the most challenging questions when investing in generative AI is determining where the value would accrue in the long term. How do you differentiate potentially disruptive companies from mere features of other products powered by generative AI?

It’s a great question. To be honest it’s a question all investors and founders are wrangling with. The app layer is where the value accrual question becomes the most challenging in my opinion. I believe that the foundation models, infrastructure tooling, and security that support AI frameworks will have somewhat clear market opportunities.

At the app layer though, the worst place seems to be as a company that is a thin wrapper around a foundational model. Their products are too general to compete with niche providers and not differentiated enough for existing technology companies to offer them as bolt-on features. Unlike past technology platform changes, the technology incumbents of today are not asleep at the wheel per se. They are typically cloud-native, microservice-architectured technology companies and as a result are far better positioned to embrace AI and thwart some of their competitors.

Frankly, if a new AI technology does not reimagine and fundamentally change a workflow process or industry, it will most likely have a tough time in becoming an enduring company. There is some irony here because in some ways SaaS gave rise to the point solution, but with AI it very well could mean those companies and startups that take a more platform approach to their solution may be best suited. For example, Roll is an interesting example of a company reimagining video quality studio production with an iphone, the cloud, and AI from an end-to-end process perspective. Another exciting example is Wonder Dynamics, which uses AI technology to automatically animate, light, and compose CG characters into a live-action scene, and in doing so they are rapidly condensing the VFX process for professionals and democratizing the capability of this technology for creators. Our latest early-stage investment, SportVisio is redefining with AI how sports statistics are captured and recorded enabling all levels of athletic competition to be able to obtain robust sport analytics data.

From a more macro perspective, I am excited about the impact AI can have on categories such as advertising, UX, and social. Online advertising has become less effective and expensive for companies due to Apple’s IDFA and general privacy changes by large technology companies. With the use of AI, companies have the chance to better adapt to this new advertising paradigm by doing predictive targeting, continuous iterative content generation, and AI-augmented UX (i,e, Coframe). On the social front, I believe AI will open up new possibilities for UGC. Can of Soup is an exciting consumer social company that empowers individuals and friends to make images together in any imaginary setting they can think of. The output and creativity the app is inspiring in its users is incredibly exciting.   

In addition to generative AI, you also invest in the Web3 ecosystem and tweeted recently that generative AI represents a systemic risk to Web3. Could you elaborate on this thesis?

Yes, in some ways I do think AI is a risk to crypto, but as you have enlightened me on in our conversations it could also be a tailwind.

As it pertains to the headwinds and as I mentioned in my tweet I always believe whatever projects young developers are spending their nights and weekends working on is where you should be focused on as a venture investor. Up until recently, I would argue that if a young developer wanted to work on something novel or hard, they would work on a crypto project. Recently AI has captured everyone’s attention and unlike crypto, it is a tech platform immediately ready for scale and mass adoption. Take YC as a proxy. In the S22 batch ~10% were crypto companies and ~5% were AI. In the latest batch S23 ~1% were crypto and ~30% were AI. YC was never crypto heavy and it’s influenced like any firm or accelerator by selection biases, but I wouldn’t be surprised if applications they received followed a similar trend.

Unfortunately (for crypto), the allure of AI seems to be drawing young developers away from working on crypto projects. With fewer young developers focused on crypto projects, the pace of innovation and development within the crypto ecosystem could suffer, which could prolong the adoption cycle of the technology.

For the record, I think Crypto is not going anywhere and will be a part of our finance system, technology infrastructure and culture (via NFTs) for the foreseeable future, but I do believe the advent of AI puts pressure on the ecosystem’s developer activity.

As it pertains to tailwinds, AI will undoubtedly lead to an explosion of digital media. Whether it be images, music, or video and as a result having a verifiable and immutable history of who created that digital media property and who owns it could be quite powerful. In addition, we are going to see AI agents (virtual knowledge workers) who will do tasks and interact with other agents autonomously. It would make sense that if these digital agents/workers needed to do commerce with each other they could default to using a digital native currency such as Bitcoin to remit payment.

One of the most intriguing frictions in the generative AI space is the balance between open-source and closed-source/API-based models. How do you see the evolution of these two distribution models over the next few years? Who will emerge as the winner in the end?

I am biased toward open source. Historical surveys have shown that 98% of the world's enterprise-level companies use open-source software offerings in some capacity. It has resulted in significant software and technology advancement and is behind some of the most important internet protocols we use today.

The benefit of open-sourcing AI is two-fold in my mind. The first is to prevent the concentration of power by a few over these AI models. The second is it increases the distribution and accessibility of this technology, which will lead to new and novel use cases for this tech. Just like the open internet acted as a great equalizer for the world’s potential open-source models, AI can have a similar impact, enabling the potential for greater prosperity and economic accessibility for individuals across the globe. The counterargument is that while open models allow more people to contribute and utilize AI research and development, the powerfulness of these models makes the potential for harm and misuse from malicious actors increase with access.

The likely outcome is there will be a spectrum of models that are open or closed source based on their capability and the potential broader impact they could have on society.

💥 Miscellaneous – a set of rapid-fire questions 

Do you think there will be trillion-dollar native generative-AI-based companies? If so, would you venture to predict that OpenAI will be one of them?

Trick question. :) Many, if not all of the companies moving forward will be generative-AI enabled- similar to how cloud is pervasive today in all tech companies. With that being said, my answer is inevitably yes, but I will go out on a limb and say it won’t be any of the obvious names that we all know today.

What do you see as the biggest roadblock for the mainstream adoption of generative AI?

Longer-term compliance, security, and safety. As Gen AI becomes more and more prevalent in our technology stack, the scrutiny of its security, compliance, and safety will increase as well. As consumers, we now know if the product is free, then you are the product, so I think people will be a bit more hesitant to fully embrace AI in all facets of their life or business without better understanding of the implications of doing so. This isn’t the same concern as others who are sounding the alarm on AI being an existential threat, but more acknowledgment of how consumers are more perceptive to how their data is being used.

What are the most significant mistakes that you see entrepreneurs making while building in the generative AI space? How about the most common mistake investors make?

It’s probably the same for both. Investors and entrepreneurs not fully considering their competitive moat, business durability, and execution advantage for the product they are building.