Knowledge-worker compensation hasn’t caught up to Enterprise AI

The smarter AI gets, the more valuable human contributions to AI systems become, and I do not think corporations are pricing those contributions appropriately right now.

The OpenAI Sales team just debuted an impressive tool they built with ChatGPT Work that gathers all of the necessary internal and public data for their sales pipeline and throws it quickly into presentation-ready decks, spreadsheets and webpages. But you know what’s specifically missing from this content? The person who pipes up in the meeting and says, “I know someone higher up in this org chart from a past role and I can approach them.” Or the person who says, “This isn’t a regular customer. This is a strategic account for X, Y and Z reasons.”

Those contributions can create tens of millions of dollars in revenue. They come from an individual’s relationships and pattern recognition developed over decades in different roles and companies. And increasingly, employees are being encouraged, often through the design of these AI interfaces themselves, to funnel that knowledge into new corporate AI systems without retaining proportionate ownership of, or economic participation in, the value it creates.

I fundamentally disagree with the way employees are expected to interact with enterprise AI systems today. Historically, a lot of your individual know-how left the company when you did. Now companies have the ability to capture it, structure it, reuse it and compound it across the organization.

That is a very different value proposition, and I’m building something that allows people to retain more of the value they contribute.

The question I keep returning to is - as companies capture and compound more of our individual knowledge through AI, should we continue contributing it under the same salary and equity structures we accepted before these systems existed? Because AI-native companies already pay tens of millions of dollars for proprietary external data. They clearly understand that differentiated information has economic value. So what happens when the differentiated information is their own employees’ relationships, general know-how, and instincts? If the technology has fundamentally changed how much value companies can extract from individual knowledge, why are we assuming the economics for the individual should stay exactly the same?

Part of me would like to believe the AI-related and -adjacent layoffs we’re seeing are evidence that executives understand this, that they plan to run leaner organizations and reward the people who remain far more generously for the knowledge and leverage they bring. But I suspect that may be giving them too much credit.