Nov 20, 2017 · 29m · mad
The Evolution and Impact of AI // Ash Fontana, Zetta Venture Partners
0:00 / 0:00
▶ Watch on YouTube →
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
In this DataDrivenNYC fireside chat hosted by Matt Turck, Zetta Venture Partners' Ash Fontana shares his venture capital investment framework for artificial intelligence, highlighting data network moats, vertical AI strategies, operational execution, and the broader societal impact of machine learning.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. Matt holds 8.5% of the talking time here. How this is scored →
Matt as informed peer 2.6 Guest teaching 4.5 Guest disagreement 1.0 Matt pushing back 0.8
speaking balance: gold is Matt, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 2:33 Reframing the host's general question
Ash politely rejects Matt's open prompt to list sectors, opting instead to dictate the conversation structure using his own framework.
Hardest push from Matt ▶ 11:22 Challenging the VC firm's technical depthMatt directly probes whether the partners actually evaluate complex machine learning technology themselves or outsource the evaluation.
Biggest teaching moment ▶ 7:00 Explaining the economic threat of free cloud AI toolsAsh provides a masterclass on why horizontal AI tools struggle to compete against cloud providers, backing it with the historical Omniture example.
Matt holds his own ▶ 6:25 Framing the horizontal vs vertical AI investment thesisMatt demonstrates clear sector knowledge by introducing the structural debate between horizontal and vertical AI strategies.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | Matt as informed peer | Guest teaching | Guest disagreement | Matt pushing back | Why |
|---|---|---|---|---|---|---|
| Event Opening and Speaker Welcome | 2 | 2 | 0 | 0 | Matt opens the episode with standard background questions regarding Zetta Venture Partners, fund timing, and target stages. Ash politely outlines their thesis around predictive data assets without any friction. | |
| The AI Adoption Risk Curve Framework | 1 | 6 | 2 | 0 | When asked a general question about interesting sectors, Ash reframes the conversation to present his structured adoption risk framework. Matt listens passively as Ash walks through consumer, additive, AI-centric, and AI-enabled categories. | |
| Strategic Focus on Vertical vs Horizontal AI | 4 | 5 | 1 | 1 | Matt demonstrates industry knowledge by framing the horizontal versus vertical AI thesis. Ash validates Matt's distinction and elaborates on why horizontal models fail against free cloud infrastructure, referencing historical precedents like Omniture. | |
| Investment Evaluation Criteria and Scientific Process | 5 | 4 | 1 | 4 | Matt presses Ash directly on whether the partners conduct deep technical due diligence themselves or rely on third parties. Ash explains their scientific evaluation of experiment stability and dataset uniqueness. | |
| Data Asset Case Study: Constructor.io | 3 | 5 | 1 | 0 | Matt prompts Ash for case studies and key lessons learned. Ash delivers detailed advice regarding data network moats, interactive ML interfaces, and hiring talent from fields outside computer science. | |
| Societal Impact of AI and Job Creation | 3 | 4 | 2 | 1 | Matt introduces the topic of societal impact and job displacement. Ash rejects apocalyptic narratives, calling himself an optimist and providing portfolio examples of direct job creation in data labeling. | |
| The Next Horizon: AI-Enabled Optimization in Complex Systems | 3 | 6 | 1 | 0 | Matt asks where the industry sits on the hype cycle. Ash provides a detailed technical and philosophical explanation of complex systems, ensemble modeling, and resource distribution. | |
| Audience Q&A: Assessing Data Moats and Synthetic Data | 0 | 4 | 0 | 0 | The host yields the microphone to the audience. Ash enthusiastically addresses an audience question regarding generative synthetic data and dataset evaluation. |
Statements from this episode (13)
Fontana: Zetta was the first fund focused on intelligent systems
“We lay claim, which sounds like a funny claim to lay today as being the first fund completely focused on intelligent systems.”
Fontana: Zetta invests when startups prove dataset predictive value
“We're really interested when a company has a unique data set, and they've started to show that it has some predictive value. So it can, you know, they've run an experiment to show that it's predictive of something at a very high level of accuracy or they've pu…”
Zetta avoids consumer AI and superficial AI overlays due to data disadvantages
“What we're not spending time on are things where AI, well, we don't spend any time in the consumer stuff because a lot of that data is owned by bigger companies, and it's very hard to get a data advantage there, and we don't really spend time on things that ha…”
Fontana: Zetta maintains a firm thesis against horizontal AI investments
“From the very beginning of our fund, we just said we're not going to invest in anything that's horizontal.”
Fontana: Vertical AI wins through specialized data collection and model tuning
“To make us only invest in vertically focused applications, because that's where you can really get ahead of everyone else by focusing on tuning a model for a very specific purpose, getting data to train a model for a very specific purpose.”
Fontana: Zetta was the largest investor in Kaggle when it sold
“We were the, Biggest investors in Kaggle when they sold on the board there.”
Fontana: 2017 data science tools are like software tools in the 1980s
“The tools that data scientists have available today are like what software developers had in the eighties.”
Ash Fontana: Zetta does not look for revenue traction in seed AI startups
“What we don't look for is traction, right? Like by the time a company has revenue and whatever else, it's sort of, One, if you think of what my job is, it's to find things that other people can't find that are valuable, and it's pretty obvious at the point whe…”
Ash Fontana: Data isolation prevents Algolia and Elastic from improving search algorithms
“Now, Algolia, Elastic, like, all these companies do that. They make it very easy to deploy a very fast search box on your website, but the thing is, because they guarantee you that they're not going to share any data in any way with anyone they can't really im…”
Ash Fontana: Search-as-a-service incumbents are AltaVista while Constructor.io is Google
“A lot of the, all the companies in the market today for search as a service are AltaVista, and constructor is Google.”
AI startups must reject on-premise deployments to build defensible data moats
“And some of those customers will ask you to do an on-prem private deployment, will ask you to not use their data to train anything that you do, and you need to reject those customers. Because you are not going to build a company that has any sort of moat aroun…”