This week, an interesting idea surfaced in Silicon Valley: AI tokens as part of engineers’ pay. The basic idea is simple. Instead of just salary and bonuses, companies would also provide AI tokens—units that run powerful tools like ChatGPT and Claude. These tokens can help automate tasks and boost productivity.
Jensen Huang, the CEO of Nvidia, recently sparked excitement by suggesting that engineers could receive about half their salary again in tokens. By his estimates, top engineers might spend around $250,000 a year just on AI compute. He sees this as a way to attract talent, predicting it could become a common practice in the tech industry.
The concept isn’t new. Tomasz Tunguz, a well-known venture capitalist, discussed AI tokens earlier this year. He noted that tech startups are already considering these tokens as a significant part of compensation. For instance, a high-paying software engineer could see their total earnings jump from $375,000 to $475,000 when including a $100,000 token budget.
The rise of “agentic” AI tools, like OpenClaw, has accelerated this conversation. These tools perform tasks on their own, which means they can consume tokens quickly. An engineer managing multiple agents can use millions of tokens daily, far more than someone just writing text.
A recent article from the New York Times highlighted this trend, revealing that engineers at big companies like Meta and OpenAI are competing based on their token usage. Many engineers now view generous token budgets as a new job perk, much like dental insurance used to be. One engineer even shared that they spend more on tokens than their salary, although their company covers the cost.
However, while more tokens can seem beneficial, they might not guarantee job security. Large token allocations often come with high expectations. If a company is funding a significant amount of AI compute for one employee, there’s an unspoken demand for increased productivity.
There’s another concern too. When the cost of tokens approaches or even surpasses an engineer’s salary, companies may begin to question the need for more staff. This shift could change how finance teams view workforce size.
Jamaal Glenn, a financial expert, warns that these tokens might be a clever way for companies to make compensation packages appear more attractive without increasing real cash or equity. Unlike salary or stock options, token budgets don’t vest or appreciate over time. If companies successfully normalize tokens as part of pay, they may keep cash compensation stagnant while promoting growing token allowances as a sign of investment.
Ultimately, while AI tokens can seem like a new opportunity for engineers, it’s essential to think critically about their long-term value. As the landscape of work continues to evolve rapidly, understanding how these changes affect compensation is crucial.
For more detailed insights, you can check out the New York Times article on this trend.
