Investing in AI still remains a grey area for venture capitalists. Sequoia Capital shows a graphic with a big blank space, leaving the question unanswered. The Information's article, 78 Artificial Intelligence Startups That Could Be for Sale This Year, suggests the mood away from investing in late entrants or features is in full swing. However, one use case to invest in AI that feels certain is evaluating venture-backed companies. This does not use the Language Modelling Method (LLM) but rather combines heuristics created by and scored by humans with machine-learning pipelines.
SignalRank's platform is a prime example of this. It captures human attributes and quantitatively analyses investor behaviour and decisions rather than focusing on company-level data or financial fundamentals. It takes human factors into account and combines them with machine learning in order to selective private market assets.
Focusing on investors who play 'multiple hands across multiple tables' allows the model to aggregate consistent human judgement over time. Qualitative human attributes are turned into structured data, which the SignalRank algorithm uses to create objective signals.
Once listed, the SignalRank index will be available to retail buyers also, with the minimum purchase being a single share. As money flows into the index, it is used to support partners' pro-rata in the next set of companies. The goal is to have an index of around 200 companies at scale, constantly refreshed through exits and new investments.
Investing in AI for enterprise apps, angel and seed investing is still a question facing investors. But investing in AI to evaluate venture-backed companies is becoming more popular. Heuristics created by and scored by humans is combined with machine-learning pipelines in order to selectively choose private market assets. As this model becomes a combination of human insight and machine learning precision, private market assets will become more accessible to accredited and retail buyers alike.