AI-First Companies: Flipped
The companies that win with AI will be the companies that start with AI. The winners of the web era were companies that started on the web. The winners of the e-commerce era were companies that started online like Amazon and Dollar Shave Club; not Sears and L’Oreal. The winners of the intelligence era will be AI-first companies.
We run Zetta, the first investment fund to exclusively invest in AI-first companies. We were the largest investors in the world’s biggest community of AI practitioners (Kaggle) and the first investors in the leading AI collaboration product (Domino). AI-first companies collect valuable data from day one, use that data to train predictive models, then use those predictions to automate work. In the software era, companies focused on building product features, getting qualitative feedback from human users and manually iterating on those product features. Today, in the intelligence era, AI-first companies focus on building predictive models, getting quantitative feedback from machine agents and automatically iterating on predictive features.
The big tech companies of today can’t win in this era because they’re not AI-first companies. Incumbents are focused on adding features, reducing cost and locking in customers. Incumbents started in an era where customers were afraid of putting their data in the cloud. Without the data, you can’t build predictive models and without predictions you can’t automate anything. Competing with the AI-first companies would require the big tech companies to go back and negotiate their customer contracts, change the way they build products to collect more data and hire whole new teams of machine learning engineers.
This is an opportunity for startups: be AI-first and dominate an industry. There are opportunities to create AI-first companies in every industry. Taking stock of what you have on your shelves using AI-based computer vision (automating stock-taking), figuring out what damage was done to a car by matching the AI-based analysis of a photo with repair/replace rules (automating insurance claim processing), learning from stoppages on a production line to predict & avoid the next outage (automating manufacturing), combing the spare parts log so that you order just enough (automating procurement), watching how translators correct words to do it automatically the next time around (automating language translation), and monitoring all the data coming out of health devices to predict if your heart condition is going to get better or worse (automating cardiology). We support companies tackling each of these opportunities.
Here’s the twist: AI-first companies will dominate industries more completely than ever before because their advantage will be unassailable. The first phase of the intelligence era will be dynamic, in that we will see AI-first startups challenge incumbents in every industry. Startups will beat big companies. The concern is that we may then see a lifeless, second phase where the new incumbents - the startups of yesterday - can’t be challenged by the next generation of startups thanks to their supreme data advantage. In a sense, this isn’t too different from the last era - the companies that dominate the technology industry today were startups 20 years ago. The difference is that the leaders of the intelligence era could rise more quickly and entrench more fully. This could stifle innovation, leaving whole ecosystems of startups for dead.
Governments may react by redefining monopolies, and re-thinking anti-trust laws. The Chicago school of economics informed our current definition of a monopoly: if it’s good for the consumer, it’s not a monopoly. More efficiency is good for the consumer, so companies can continue to commandeer market share as long as there’s downward price pressure. However, that is perhaps a definition that only works in the short term. AI-first companies can price their products low to maximize usage, thus maximizing the collection of valuable data then, when their models achieve supreme accuracy, increase prices to capture all the value in the market. This is a sort of data-enabled monopolistic equilibrium, beyond which no one can catch up and the AI-first company can charge whatever they wish for their products because it’s so much better than others.
We are in a new era of technology — the intelligence era — and with that comes a new type of company: the AI-first company. Previously, the companies that won managed land, labor or capital better than anyone else. Today, the companies that win manage data better than anyone else. This presents tremendous opportunity for startups to beat incumbents by building products that generate deep moats around their business. However, once the startups of today become the new incumbents of tomorrow, their moats could be so deep that they force us to rethink our notion of fair competition.
- AI-First Companies: Flipped
- The AI-first startup playbook
- AI adoption is limited by incurred risk, not potential benefit
- AI-First Companies
- Data rights are the new IP rights
- The Intelligent Enterprise Stack
- Beating Behemoths
- Don't sell your data
- Framework to grasp industrial analytics opportunities
- Beyond systems of record
- Positioning a machine learning company
- The intelligence era and the virtuous loop
- Vertical beats horizontal in machine learning
- Zetta Bytes AMA: Questions to ask about pricing
- Zetta Bytes AMA: Hiring a CTO
- AI Entrepreneurs: 3 things to check before you pursue a customer
- There are more data scientists than you think
- Stages of funding in the intelligence era
- Could data costs kill your AI startup?
- Measuring AI startups by the right yardstick
- Finding the Goldilocks zone for applied AI
- Data is not the new oil
- Machine Learning in the Deployment Age
- 10 innovations of the next decade
- Zetta Bytes: Privacy Preserving Machine Learning
- GDPR panic may spur data and AI innovation
- Computing like a human
- New opportunities for hardware acceleration in data analytics
- Hardware acceleration in data analytics