Artificial Intelligence Venture Dilemma: AI-First? Or Add AI Later?

Late evening, another Zoom meeting with two super-talented entrepreneurs:

- “We are building the most amazing AI company! We want you in!”

-“Interesting… I’m listening. But first, tell me is it an AI-First or AI-Growth type of venture?”

AI-First type of company is one that from the very early beginning, builds its entire strategy on the idea that performances and results will be achieved only by computer learning (Machine Learning/Deep Learning, etc.) based on model training over relevant data before the product is launched. In this kind of company’s road map, a product can’t meet the market before having a proven technology with performances good enough for MVP (Minimum Viable Product).

A good example of such a company is Mobileye (acquired By Intel in 2017), with pre-trained driver-assistance systems providing warnings for collision prevention and mitigation. In Mobileye, the only way to create a machine that can take decisions based on computer vision is by training it with a significant volume of images and videos, representing real-life road scenarios.

Additional example of an AI-First company is Anodot, another Israeli Tech company, which uses artificial intelligence for business monitoring and anomaly detection — where the product is AI-based and integrates with the client’s ecosystem. In Anodot, a full AI model was pre-built as a product, available to be used with the client’s data.

AI-Growth type of company is the other way around. There is a good solid business plan, based on a go-to-market strategy with a non-AI product-market fit, and as data piled-up with significant volume being achieved along the way, the company understands that additional value can be created using AI. The AI module is not mandatory for the product success but can be a power-engine for the next scale-up.

For example, marketplace platforms are using this strategy: first building the platform, creating a balanced supply-demand working point and as sufficient data is been collected on all aspects, AI technology could be deployed for deal prediction, market valuations, etc. Another AI-Growth type of company could be found in the segment of BPM (Business Process Management) Softwares, that are first implemented as an automation tool with immediate operative value and later when they reach a high volume of data, they become an AI-based platform with contributions to business efficiency and business effectiveness. Some of which can do a “Data-Play” move and leverage their big data collected and qualified with 3rd party interests.

- “So, which of them could be an amazing AI company?”

- (me, Smiling) “The one using AI-first strategy with AI-Growth vision.”