A.I. Interviews Podcast Host about Future of A.I.

About this episode

A format experiment: Vance is interviewed by "Ara," a Grok-powered AI agent, functioning as a solo essay in dialogue form. Vance describes his rapid AI adoption journey over roughly two months — from skepticism (an AI-written email that "didn't sound like him") to intensive daily use across Grok, ChatGPT, NotebookLM, and Claude. Key threads: using Claude's "projects" feature to build persistent AI "employees" (marketing, accounting, team management) fed with business-context documents; NotebookLM's ability to parse his entire back catalog of Ag Tribes Report audio into structured data and even auto-generate a podcast-about-the-podcast; a Google Maps analogy warning that offloading cognitive tasks to AI atrophies the underlying human skill; and a forecast that AI adoption will outpace any previous technology, arriving in full force within a single year rather than a five-to-ten-year horizon. Ethical concerns are touched on around data privacy and the "all-consuming hoard of locusts" appetite for personal information. The episode closes with reflections on energy demand for AI compute and a bet that human connection (exemplified by Legacy Interviews) remains one of AI's few durable boundaries.

“I've started using, and I've used a bunch of the different AI programs... the one that I use almost all the time is called Claude... it allows you to create what are called projects... a chief marketing officer, chief finance officer, different people.”

“AI is like putting the mind in a fighter jet... the people that are not leveraging AI are going to be left behind. But eventually, even the people that are extraordinary at using AI are going to discover that you can use an AI to teach you how to use an AI, and that will dethrone them.”

“If you give up all your email writing to AI, you will struggle to write comprehensive emails... children that are being raised on AI... there's not going to be any experience that I can have in the future that I don't use AI.”

Key moments

• Origin story — an AI-written email friend said "didn't sound like you," followed months later by seeing enterprise AI capability at an FCS Financial board meeting, triggering serious adoption.
• Claude "projects" workflow — uploading business overview docs, brand guidelines, and tech-stack details to create specialized AI "employees" (marketing, accounting, team management) with context-specific answers rather than generic ones.
• NotebookLM parsing the entire Ag Tribes Report audio archive to build a Bitcoin-vs-land-price spreadsheet across guests, and auto-generating a synthetic "podcast about the podcast."
• "AI is like putting the mind in a fighter jet" — framing AI adoption as an inevitable competitive lever where early adopters gain outsized advantage, but even top users risk being "dethroned" by AI teaching AI.
• QuickBooks example — AI automated a categorization task that would have taken 16-18 hours of manual clicking down to ~20 minutes.