▶Jesse Zhang consistently argues that AI agents will become the primary interface for consumer-brand interactions, eventually supplanting websites and mobile apps [28, 46].Mar 2026

▶He believes customer service is the most significant and immediate market for enterprise AI adoption, citing large cost reductions and existing escalation infrastructure as key drivers [12, 36, 43].Mar–Apr 2026

▶Zhang posits that the primary business of foundation model providers like OpenAI is their application layer (e.g., ChatGPT), not their API, which he views as a commoditized wedge for adoption [17, 39, 45].Apr 2026

▶Across both interviews, Zhang emphasizes that the key competitive moat for AI agent companies is the continuous improvement loop from client-specific data, which makes incumbent agents increasingly difficult to displace [14].Mar–Apr 2026

▶Zhang champions an intense, mandatory five-day in-office work culture as crucial for productivity [6, 47], yet Decagon is also expanding internationally with a new European office, which may create operational and cultural challenges for this rigid model [44].Mar 2026

▶He describes the current VC climate for AI as having 'too much excitement' and being 'way too easy' to raise money [54], while simultaneously noting that his own company, Decagon, has received preemptive funding offers almost immediately after closing rounds, indicating he is a major beneficiary of this environment [20].Apr 2026

▶Zhang advocates for a 'forward-deployed engineer' model for customer success but qualifies that the true Palantir-style model is only viable for massive ($10M+) deals, suggesting most startups, including potentially his own, are using a more hands-on but less intensive version [24, 27].

▶He states that AI coding agents are used to augment engineer productivity rather than replace them [51], yet his broader thesis is that AI will replace low-cost labor in areas like customer service, creating a nuanced view on AI's overall impact on the workforce [18, 33].