Can AI Just Do My Tax Return? The Three Bridges That Make It Possible in 2026

AI in 2026: What to Expect

Introduction to AI Advancements

• The speaker expresses excitement for the year 2026, anticipating significant advancements in AI capabilities, particularly in automating tasks.

• Acknowledges that 2025 was a foundational year focused on selecting tools, creating policies, and training teams to become AI literate.

Current Capabilities of AI

• Identifies three key gaps between current technology and full automation of tasks like tax returns or contract reviews.

• Emphasizes the importance of understanding these gaps to discern valuable AI solutions from those that are not yet viable.

Gap Analysis: Playbooks

• The first gap involves providing clear playbooks for AI; it needs specific instructions tailored to individual preferences and policies.

• Highlights progress made in creating agents capable of managing tasks such as triaging emails and reviewing documents using established systems.

Gap Analysis: Contextual Understanding

• The second gap is the need for deep contextual knowledge about clients and past work, which is often stored across various systems.

• Discusses ongoing developments towards universal connectivity (MCP servers), allowing AI tools to access necessary information from multiple sources.

Gap Analysis: Physical Interaction

• The final gap pertains to giving AI the ability to perform physical actions within software environments, such as filling forms or manipulating data directly.

• Mentions emerging technologies like AI-powered browsers that could facilitate these interactions but notes they are still developing with inherent risks.

Summary of Key Gaps

• Summarizes the three essential requirements for effective AI task execution:

• Clear playbooks detailing how tasks should be performed.

• Deep contextual understanding through interconnected systems.

• Capability for physical interaction with software applications.

AI Innovations and Their Practical Applications

Exploring AI Capabilities

• The discussion highlights the emerging technology of giving AI physical capabilities, such as the ability to touch and interact with objects. This technology is still in its infancy but shows promise for future applications.

• Although current implementations are slow and not time-saving, there is an emphasis on the importance of experimentation for innovators and early adopters who wish to explore these advancements.

Strategies for Effective AI Utilization

• The speaker advises engaging with various AI tools, suggesting that while innovators should experiment, most users should focus on established methods that cater to both early and late majority adopters.

• A key strategy involves creating AI agents based on documented playbooks. This approach allows teams to teach AI how to perform specific tasks effectively.

• There is a call-to-action for those interested in collaboration or further exploration of these technologies, indicating a community-driven approach towards implementing AI solutions.