Wade Foster observes that users consolidate their AI work into a single primary tool (like ChatGPT or Cursor). This means other applications, like Zapier, must become "headless" and integrate into these chosen environments rather than trying to become the primary destination themselves.
Wade Foster argues that for most business automation, delegating tasks to an AI agent is inefficient. Instead, an AI should be used to *build* a deterministic workflow (i.e., traditional code) which is cheaper and more reliable for execution. AI's reasoning is best reserved only for parts that truly require it.
The old no-code model of manually connecting boxes is obsolete. According to Zapier's CEO, the new paradigm is prompting an AI to build the workflow. The visual interface's new role is not for building, but for helping humans quickly verify the AI's work and serve as documentation for the team.
When an automated workflow fails, Zapier internally spins up five independent AI agents to diagnose the problem. They found that if four out of the five agents agree on the cause, it is highly likely to be the correct diagnosis. This multi-agent consensus model improves the reliability of automated troubleshooting.
To solve the "what should I automate?" problem, Wade Foster runs a weekly personal automation that reviews his activity across Gmail, Slack, and other tools. The AI then proposes specific, personalized workflows he should build. This moves beyond generic recommendations to hyper-contextual suggestions based on actual behavior.
Once a company achieves baseline AI fluency, the biggest hurdles become people-related: rewriting job descriptions, reskilling teams, and managing organizational change. Zapier appointed its Chief People Officer to lead AI transformation because these human-centric challenges became more critical than the initial technical adoption.
Zapier encourages teams to move communications from private DMs to public Slack channels. While this aids human transparency, the primary driver is strategic: it makes a much richer dataset of context available to the company's internal AI agents, dramatically improving their effectiveness.
Wade Foster believes AI companies over-focus on competing in a "sea of sameness." The real challenge and massive opportunity lies in educating the average user, who is still largely unfamiliar with AI's potential and hasn't adopted any advanced tools yet. The battle is against apathy and habit, not other startups.
Wade Foster argues that fixed, seat-based pricing is dying because it doesn't align with the variable nature of AI-driven work. He sees the market splitting into two models: usage-based for commodity-like AI services and outcome-based for enterprise tools that deliver a clear, measurable result (e.g., "price per resolved ticket").
The problem with using AI in communications isn't its use, but the user's lack of ownership. According to Wade Foster, you must be prepared to stand behind, explain, and verify every detail of what you send. Using AI to transfer the burden of thinking and editing to the recipient is a misuse of the technology.