Go and tokenmaxxing.
Laura Gonzalez, head of human resources at AI video company Synthesia, says focusing on the use of AI rather than what it offers risks encouraging the wrong initiatives.
A better goal is impact, she said. Her view aligns with a growing number of leaders who believe that heavy use of AI does not necessarily lead to business wins.
Evaluating software developers primarily based on token usage is like evaluating salespeople based on the number of calls they make rather than the number of deals they close, she says.
“It’s not an indicator of success,” Gonzalez said of the use of the token. Instead, companies should focus on the business impact of AI, she said.
Rather than looking at the number of tokens someone has spent, which are common metrics for measuring AI usage, Gonzalez said it’s better to ask whether the technology is helping teams ship faster, better prepare for customer calls, or reduce the time it takes for new hires to start making an impact.
London-based Synthesia develops software that allows companies to create AI-generated videos for training, marketing and other purposes, so her employees use AI extensively across all roles, she said.
Although the company does not reward employees who are enthusiastic about using the token, it still hopes that employees will use the technology.
“We absolutely expect everyone to leverage and use AI as needed within the company,” Gonzalez said.
changing trends
The lack of focus on tokens marks a departure from the messaging of some companies, where leaders treated burn rate as a metric for AI adoption. In some cases, leaderboards may highlight employees who are generating the most AI activity.
Now, the rising cost of AI for some heavy users is forcing some bosses to rethink what they tell their employees.
Coinbase saw a spike in AI usage in February after Claude rolled out a new coding model. Business Insider previously reported that crypto exchanges currently have price caps ranging from $500 to $5,000 per week, depending on employee level and job.
Marty Kausas, CEO of customer operations platform Pylon, said Wednesday that his company’s annual billing for Anthropic will jump from $400,000 to $1.4 million as it moves into the enterprise pricing tier. Kausas, who said he inadvertently spent $4,000 on Claude Code in three days, predicted that TokenMax’s era is “coming to an end.”
Gonzalez said cost concerns are not driving the company’s token-focused approach. In January, Synthesia announced it had raised $200 million in a Series E round, valuing the company at $4 billion.
Not encouraging the use of tokens at scale does not mean abandoning innovation, she said.
“There is still a culture of giving people the tools to experiment and do things without rewarding them for maximizing their tokens,” Gonzalez said.
When tokens matter
There are still times, such as during hackathons, where companies use leaderboards to show token usage. But those are the exceptions, she said.
The corporate push-pull around the use of tokens comes as many employers are looking for ways to encourage employees to use AI to increase productivity. In some cases, managers are using dashboards to track how much AI is being used by staff in roles such as software development. In some cases, employers are considering using AI to evaluate workers.
Token allocation can still be important, especially for new employees who may join the company with different expectations about access to AI tools.
Jeff Hyman, a recruiter with nearly 30 years of experience, said future employees may still want to know what AI resources they have access to.
He compared asking about token allocation to the question workers routinely asked in previous eras: “Does the job include a laptop?”
Do you have a story to share about using tokens? Contact this reporter at: tparadis@businessinsider.com.