Legora CTO Criticizes Tokenmaxxing as Ineffective Incentive
Business Insider reports that Legora chief technology officer Jacob Lauritzen criticized 'tokenmaxxing' on the '20VC' podcast, calling it 'a really stupid way to do anything.' Lauritzen, who joined Legora in 2024, said hack days, demos, and reward structures tied to measurable efficiency gains are better ways to encourage AI adoption, according to Business Insider. He also acknowledged a trade-off for fast-growing companies, saying that spending tokens can be worth it when the efficiency uplift is large. The remarks, also covered by other outlets and available on the 20VC podcast, speak to a practical question many teams face: how to measure AI adoption without rewarding hollow usage.
What happened
Business Insider reports that Legora chief technology officer Jacob Lauritzen criticized what he called 'tokenmaxxing' on an episode of the '20VC' podcast. Business Insider quotes Lauritzen saying that people 'bring up token usage at performance reviews,' which 'leads to tokenmaxing, which is people just burn tokens just to look good,' and that this is 'a really stupid way to do anything.' The article reports Lauritzen joined Legora in 2024 and recommended hack days and demos as better mechanisms to surface productive AI work.
Editorial analysis - technical context
Reporting frames tokenmaxxing as inflating AI-tool usage metrics through heavy, often low-value calls to models such as Claude, Codex, and Cursor. In organizations that track raw token consumption, such dashboards can encourage optimizing for the metric rather than the outcome. That creates two operational problems for practitioners: token volume alone is a weak proxy for productivity, and unchecked token growth raises cloud and API costs without proving value.
Industry context
The comments land amid a broader debate over how to measure AI adoption inside firms, with many teams shifting from raw-usage metrics toward outcome KPIs such as time saved, defect reduction, or throughput. For teams with high opportunity cost per engineer-hour, the calculus for exploratory token spend differs from lower-cost settings; Lauritzen's on-record remark about trade-offs illustrates that tension without prescribing one approach.
What to watch
Signals to monitor include whether internal dashboards move from token counts toward outcome-linked metrics, adoption of structured experimentation such as hack days, and tooling that correlates model calls with business KPIs and attributes spend to specific projects, which is necessary to judge when exploratory usage produced real gains.
Key Points
- 1Token-focused dashboards can reward volume over value, encouraging superficial model calls rather than outcome-driven work.
- 2Alternatives such as hack days, demos, and outcome KPIs surface practical efficiency gains more directly than raw token counts.
- 3For high-opportunity-cost teams, bounded exploratory token spend can be rational, but only with attribution to measurable productivity gains.
Scoring Rationale
A substantive, widely covered take on a real operational problem, how to incentivize AI adoption without rewarding hollow usage, now corroborated by Business Insider and the primary 20VC podcast. It is useful framing for engineers and managers but remains a single executive's opinion rather than a product launch or research result.
Sources
Public references used for this report.
Practice with real Ad Tech data
90 SQL & Python problems · 15 industry datasets
250 free problems · No credit card
See all Ad Tech problems