NEA's Tiffany Luck on AI IPOs, Personal Agents, and the ROI Reckoning

Earlier this year, Silicon Valley was swept up in the craze of "tokenmaxxing"—a trend where CEOs aggressively encouraged employees to push AI usage to its absolute limits. The goal was to maximize productivity and innovation, but the strategy has quickly collided with a harsh economic reality: the bill has finally come due. Reports indicate that major tech players are feeling the financial pinch. Uber allegedly burned through its entire annual AI budget in just a few months. Other companies have been forced to scale back, with some organizations cutting Claude licenses for specific departments, and Meta quietly killing its internal AI usage leaderboard. This growing tension between boundless AI ambition and finite budgets is at the heart of a new discussion featuring NEA partner Tiffany Luck, who recently shared her insights on AI IPOs, the future of personal agents, and the impending ROI reckoning. According to Luck, the enterprise AI landscape is at a critical inflection point. The days of unchecked spending on large language model tokens are rapidly ending as CFOs demand tangible returns on their massive AI investments. This ROI reckoning is forcing startups and established tech giants alike to pivot from mere usage metrics to demonstrating actual business value. If AI tools cannot prove they are driving revenue or drastically cutting operational costs, their budgets will be slashed. However, Luck remains deeply optimistic about the long-term trajectory of artificial intelligence, particularly in the realm of AI-driven personal agents. She envisions a near future where personal agents move beyond simple chatbots to become indispensable, autonomous assistants that seamlessly manage complex workflows. This evolution, she argues, will be what finally justifies the massive compute costs, delivering clear, measurable productivity gains that far outweigh the token spend. On the investment front, Luck is closely watching the AI IPO pipeline. As the market matures, investors are no longer satisfied with hype; they want sustainable business models and clear paths to profitability. The startups that will successfully go public in this upcoming wave will be those that have navigated the tokenmaxxing hangover by building highly efficient, specialized AI solutions rather than relying on brute-force, generalized compute. Ultimately, Luck's perspective underscores a broader maturation of the AI industry. The transition from the speculative tokenmaxxing era to a disciplined focus on ROI is not a setback, but a necessary step toward building a sustainable AI economy. As personal agents become more capable and efficient, and as companies learn to do more with less, the industry is poised to emerge stronger, leaner, and far more resilient.