Alphabet Inc.’s venture capital arm GV, formerly Google Ventures, has replaced humans with artificial intelligence when it comes to investment decisions, at least according to a report published Thursday.

Axios made the claim, saying that GV uses an algorithm called “The Machine” that effectively permits or prohibits both new and follow-on investments.

The Machine, which appears to use a mix of machine learning and artificial intelligence to make investment decisions, makes its decisions based on data feed into it by staff members. The data consists of information about the particular company, market data, round size, syndicate partners, past investors, industry sector and the delta between prior valuation and current valuation.

The algorithm is then said to use a traffic light-style system of indicators on a given investment. A green light indicates that an investment opportunity is a go, a red light means don’t invest and a yellow light indicates proceed with caution — though Axios’ sources noted that a yellow is usually the practical equivalent of red.

If replacing humans with AI for multimillion-dollar investment decision isn’t enough, The Machine is also said to have been used to source deal opportunities but apparently wasn’t overly successful in doing so.

The report claimed that knowledge of the machine is well-known among GV’s partners, who “have been known to try gaming the machine, manipulating inputs to get the desired results.” Those same investors apparently aren’t particularly fond of its decision-making prowess either, noting that The Machine “has ripped out their guts, possibly costing them lucrative opportunities (particularly some of the more risky bets that often turn into venture’s biggest wins).”

A move to robots, or in this case an AI-based computing platform, in place of human decision-making isn’t a huge surprise given predictions that robot automation will replace as many as 800 million jobs by 2030.

The Axios report does note, fairly, that venture capital is as much an art as a science and hence it’s questionable whether a robot can always make the right decision.

Photo: Pxhere