NEA’s Tiffany Luck says enterprises are still figuring out their AI ROI
TL;DR
TechCrunch’s Equity podcast talks to NEA partner Tiffany Luck about the hangover after Silicon Valley’s Tokenmaxxing push: teams were told to maximize AI usage, and now finance teams are checking the bill. The article names concrete warning signs: Uber reportedly burned through its annual AI budget in a few months, some companies cut Claude licenses for parts of their org, and Meta ended an internal usage leaderboard.
Nauti's Take
Tokenmaxxing was a convenient management shortcut: push usage up and hope productivity becomes visible later. That works briefly while budgets are loose.
The healthier phase is less glamorous: AI has to attach to real workflows, not usage leaderboards. The VC perspective is naturally interested, but the core point holds: the next enterprise race will not be won by the loudest models, but by systems that show cost and impact in the same spreadsheet.
Briefingshow
The shift is not from hype to rejection, but from hype to procurement discipline. AI vendors will be judged less by how many tokens a company consumes and more by whether a task gets cheaper, faster, or measurably better. Tools that cannot prove that risk being treated as experiments instead of infrastructure.
Sources