Software finally eats services - Aaron Levie
The a16z Show
2025/09/24
Software finally eats services - Aaron Levie
Software finally eats services - Aaron Levie
The a16z Show
2025/09/24
A panel of tech leaders and investors explores the transformative impact of AI on talent, productivity, and market dynamics, while questioning whether current immigration policies are aligned with the needs of a rapidly evolving industry. The conversation sets the stage for a deeper look at how technological and structural shifts are reshaping who wins—and who gets left behind—in the new era of computing.
The discussion examines the flaws in the H-1B visa lottery system, arguing it favors large tech firms over startups and proposing a wage-based merit system to better allocate talent and protect wages. The panel highlights how AI is driving substantial productivity gains, with companies like Box shipping 30% of their code via AI, marking a shift from writing to reviewing code. Bottom-up, engineer-driven AI tools are proving more effective than top-down corporate platforms. Early adopters—especially startups—are achieving 3x to 10x productivity boosts, compressing development cycles dramatically. The rise of AI-native startups is disrupting incumbents, who face organizational inertia despite their resources. Historical platform shifts show that market leadership rarely transfers across eras, suggesting today’s AI frontrunners may not dominate long-term. Consumer-led AI adoption is now accelerating enterprise integration, with younger generations expecting AI tools at work. While early brand recognition helps, rapid innovation means new entrants can quickly challenge established players.
01:07
01:07
Box ships a third of its code using AI
02:11
02:11
Setting a price for allocation can improve market efficiency.
02:33
02:33
Amazon and Google might capture most talent under a $100K salary policy
08:51
08:51
Replace H-1B lottery with a high-merit, wage-based system to attract top talent and raise sector wages.
12:28
12:28
At Box, about 30% of code is now generated with AI assistance.
14:48
14:48
The shift from code writing to code review due to AI background agents is significant and will change computer science.
31:28
31:28
AI’s biggest gains go to those with area-specific expertise who can judge its output
33:52
33:52
AI creates a third category of users who gain product utility without monetization.
40:25
40:25
Startups can achieve scale quickly in the AI era as software can go viral and neutralize incumbent advantages.
40:58
40:58
Historically, the advantages of incumbents have been overestimated during platform shifts.
51:42
51:42
75% of adults use AI many times per week, indicating rapid mainstream adoption.
53:03
53:03
Younger workers will demand AI tools at work due to personal experience with AI for tasks like homework.
54:44
54:44
Past early leaders like Excite and Yahoo were replaced by unknown names such as Google.
55:22
55:22
Agents will create new opportunities for disruptors in AI.
