One of the fastest growing, yet least understood, areas of the tech sector is artificial intelligence (AI). AI includes the development of computer systems that perform tasks which, up until recently, required human intelligence. This might include jobs dependent on speech recognition, visual perception, and/or decision-making, and AI is currently being put to use in a wide range of innovations, from driverless cars to Alexa to cancer detection technologies. AI makes it possible for machines to learn from experience, a task that they can master faster and with more accuracy than humans. But with so many companies now using AI, and so many business leaders and valuation experts uncertain of AI’s worth, an unprecedented number of tech companies are left uncertain of the actual value of their companies.

There has been no shortage of investment in AI startups, and tech giants like Google, Twitter, Salesforce, Apple, and others, have been aggressively acquiring AI startups for the past five years or more. Meanwhile, Forrester Research predicts that Cognitive Computing Technologies will be worth $1.2 trillion by 2020, with AI investments tripling by that time. And Accenture predicts that the market will be worth $8.3 trillion in the US alone by 2035. That is a lot of money exchanging hands.

And yet, we are still debating how to value AI in the first place. The challenge lies in determining whether valuation methodologies should follow a strategic approach or an operational one. AI can certainly be considered proprietary technology, but in instances where the “vision” of the company (the way in which the intellectual property, or IP, is being put to use) plays into the valuation, these opposing views must be negotiated.

Founders in the AI space push for their companies to be valued from a strategic point of view, placing the emphasis of the valuation on the revolutionary idea and corporate vision that has made the company a success. Meanwhile, more traditional operational valuations tend to favor investors and acquiring companies, basing appraisals on more standard sales and profit growth formulas. While this is a fine methodology in many instances, in today’s technology age, it does not really work to ignore the value of a cutting-edge AI application or profound corporate vision. This is especially true for start-ups. For established companies, it is much easier to look at their historical financials and use those as a basis for cash flow forecasts. Start-ups, however, don’t have the benefit of historical data. Valuations must rely more heavily on future potential.

Unfortunately, there is no single correct way to value AI. The fact of the matter is that it depends on the inputs and assumptions that are unique to the company in question. Still, audit firms will demand an approach that is documentable and replicable. That is why it makes sense to engage a third-party expert like Appraisal Economics to value your AI company, so that the problem can be approached in a logical, methodical way that is defensible. Valuing technology companies is a challenging undertaking that requires much more than a simple formula. An appraisal expert must use judgement to assess both strategic vision and operational value.