It was all about getting employees to use AI. The more you used, the more applause you got. Now that the bill for AI usage is due, employers are going to use it to pay for ideas and innovation
From Tokenmaxxing to Tokenomics
Employers have encouraged employees to use AI for everything. Add to it the constant reminder that if you don’t use AI, someone who uses AI will replace you. Driven by that stress of uncertainty, employees have gone on overdrive to use AI. The vast majority are using it to write emails, plan their kid’s tennis lessons and car pool, plan vacations with often hilarious outcomes. Many use AI to write deep thought leadership articles on LinkedIn. Everyone has their own AI moment.
When we ask AI to write an email, we pay it in TOKENs. AI companies sell artificial intelligence in tiny units called tokens. Every prompt that is written and rewritten is costing tokens. AI thinking is bought and sold in tokens. The longer the conversation, the more tokens you pay. All the emails AI wrote for you has a cost. Every cat video you watched was made at a cost. The employers are suddenly faced with the bill.
Electricity is measured in kilowatt-hours. Mobile data usage is measured in GBs. AI is measured in tokens.
Tokenmaxxing: Because The Cost Is Invisible
Homer Simpson’s family was once given three wishes by the tooth fairy. He cleared his throat and asked for the first wish: “I want a thick juicy burger right now.” When the family groaned in protest he corrected himself and said, “My second wish is to add fries. The third wish is to add ketchup – I forgot that.”
Behavioral Economics experiments tell us that when you swipe the credit card, it doesn’t feel like you’re giving anything up to make the purchase, unlike paying cash where you have to hand over bills. Handing over cash activates the same areas of the brain when we experience pain. Credit cards bypass that area.
Employers incentivized heavy AI usage. More usage of AI showed up in rising cost of tokens. The leaders patted themselves on the success of the rollout. This is like the early birds at a buffet piling their plate high and wasting food while the ones who are further in the queue don’t get anything. Tokenmaxxing is the free-buffet phase of AI adoption where you are piling your plate high because nothing shows a price tag. Then the bill arrives and someone asks how much you actually needed.
Not everyone is using it to generate AI slop. Plume Network launched a payroll pilot that lets employees receive part of their compensation as tokenized financial assets rather than conventional cash. Finance pays employees in tokens. Meta’s internal leaderboards rank employees by token use, with titles like “Token Legend.” AI tokens is a status symbol.
It is sobering to remind people about Goodhart’s law. The moment we start measuring something, the metric tends to be the end goal. Who can burn up more tokens is the current day’s equivalent of “move fast and break things” that Zuckerberg had announced with a chest thump and later blamed it as a folly of youth.
Tokeneconomics asks 4 questions
Uber’s CTO moving the company “beyond tokenmaxxing” toward Phase 2 which is all about spending smart. Tokenomics is all about figuring out whether the money spent on AI was actually worth it. This is no different from the CEO asking, the Chief Learning Officer, yes every employee attended the workshop and you got rated 5 out of 5. But did it change the bottom line?
- Person or AI? Don’t just use AI to write an email. There may be better uses of AI
- Which AI model do you need? AI models go from simple ones to complex ones. You don’t need to use a high powered cannon to kill a mosquito. Use the same logic for choosing the AI model.
- Which team gets the most out of every AI dollar? All employees may work for the same number of hours, but some people have a bigger impact than their peers. Organizations are tracking which function is a cost center when it comes to AI tokens. Because some functions will become profit centers.
- When does more AI stop helping? AI models often leave you with another suggestion which you would have laughed at if a colleague suggested it. When AI suggests it in a very important tone, don’t get tempted to go down the rabbit hole.
Token of Appreciation
Tokens will transform business models and organizations. Hav
That is why tokennomics matters. It marks the moment when AI stops being an exciting technology and starts becoming an economic resource that must earn its place on the balance sheet. Innovation and impact may now have a new price tag – the number of tokens needed to innovate.
