The conversation highlights the potential and current limitations of AI in SaaS applications. AI is effectively used in onboarding and customer support, helping users adapt to new software through AI-driven tools. This has been particularly successful in presentation software, where AI aids in creating templates and easing user adaptation. Additionally, AI is automating basic manual tasks like transcription and note-taking, which enhances productivity and reduces the need for extensive human labor. However, fully automating roles and processes remains a challenge, as AI struggles with planning and complex operations, necessitating human oversight. The discussion also touches on the importance of focusing on vertical data to improve AI accuracy in specific fields like legal and accounting. Pricing and monetization of AI features are still evolving, with companies experimenting with different models to find what works best.