Bask Iyer has held the global CIO seat at Johnson & Johnson, Honeywell, VMware, and Dell, and now advises over a dozen AI companies including Zoom and ThoughtSpot. In this episode, Russ gets into how enterprise technology decisions actually get made, and how AI is exposing the real cost of CIOs, CFOs, and heads of HR operating without a shared accountability model.

Bask makes the case that most AI pilots are theater, that cost avoidance is not a real business result, and that the safest-looking vendor decisions are quietly destroying a CIO's ability to differentiate. He also shares the specific framework he uses to get CEOs out of their own way without winning the argument.

Topics Discussed:

• Why CIOs at tech companies can't lead with technical authority, and the Eric Clapton framework Bask uses to hold his position anyway
• The agent accountability gap: who owns the bot that just discounted your inventory 30% on Thanksgiving, and why CIO, CFO, and HR have to answer that question together
• Why task-specific agents are breaking the case for monolithic ERP, and what that means for how CIOs should think about build vs. buy today
• What enterprise founders consistently get wrong about selling outside Silicon Valley, and why Bask has to shock them into leaving their zip code
• "Pretotyping": the one-month framework Bask uses to pressure-test a CEO's AI conviction without a nine-month proof of concept
• Why "the tool saves me time on weekends" fails the CFO test, and what actual productivity impact has to show up on the bottom line
• The case against splitting the CIO and Chief AI Officer roles, and what Bask tells boards when they ask
• Why letting Gartner pick your stack leaves you with no budget left to differentiate, and why Bask tells younger leaders to validate vendors over coffee in Silicon Valley instead

Dan Zhang has been CFO of ClickUp for five years, with an unusually broad remit spanning finance, accounting, legal, HR, and operations. That vantage point shapes everything he shares here, from his 2x2 vendor evaluation framework built around autonomy and attribution, to his three-bucket system for measuring AI ROI across a finance team, to the exact methodology his team used to shift skeptical analysts from "I don't think AI can do my job" to "I can test my hypothesis in seconds instead of hours."

He also gets into two new hiring profiles he is actively recruiting for, why curiosity and humility have replaced domain expertise as his primary criteria, and how AI is already rewriting the corporate math on enterprise AE quotas, sales engineer ratios, and back-office headcount models.

Topics Discussed:

• The autonomy-attribution 2x2 for deciding which AI vendors get scaled, monitored, or sandboxed

• Three-bucket finance ROI framework: 1-to-10 automation, new capability creation, and quality boost

• Why top-down alignment combined with bottom-up organic adoption is the only way to roll out AI successfully across a company

• The live offsite methodology that converted finance team skeptics into believers through real-time edge case testing

• Why the board doesn't want quick wins: tailoring AI impact messaging across frontline employees, business leaders, and investors

• Two new hiring profiles replacing traditional functional expertise: the context optimizer and the insights interrogator

• How AI is already rewriting corporate math on enterprise AE quotas, sales engineer ratios, and back-office headcount models

• Why curiosity and humility, not domain expertise, are now the hiring criteria that determine who succeeds in an AI-first environment

Listen to more episodes: 

Apple 

Spotify 

YouTube

Jesse Vincent is the creator of Superpowers, the most installed skills framework for Claude Code, with over 100,000 GitHub stars in its first five months. Superpowers isn’t a prompt or a collection of tips—it’s a complete software development methodology built for a world where agents do the building. Jesse has been shipping open source for 30 years—he created Request Tracker, led Perl 5, built K-9 Mail (now Thunderbird for Android), and co-founded Keyboardio. His last hand-written line of code was October 2025. He’s never been more productive.

In our in-depth conversation, Jesse shares:

• The full Superpowers workflow: Socratic brainstorming → design spikes → spec writing → spec review → implementation → code review—and why skipping any stage breaks the whole thing

• Why the brainstorming phase is designed to trick you, not the agent—and the consulting technique behind it

• Why specs have replaced code as the only artifact humans should review—and how cross-reviewing specs between multiple Claude instances catches what a single pass misses

• The orchestrator pattern: a coordinating agent that dispatches pinned tasks to sub-agents, each with an ephemeral spec-review agent that enforces quality before any code ships

• Why most implementing agents could run on Haiku instead of Opus—and what that means for cost and speed

• The “fresh eyes” prompt that causes agents to step back and actually think critically—and why managing agents is eerily similar to managing humans

• Why saying “I love you” to Claude measurably improves output—and how persuasion principles from Cialdini’s Influence transfer directly to frontier AI agents

• His advice to developers: the spec is your leverage now—learn to write clearly, decompose problems well, and let the agents grind

Karl Mosgofian built his career in IT from the ground up, starting as a teenager and eventually serving as CIO at Gainsight, where he watched AI reshape what the CIO role actually means inside a company. In this episode, he and Russ Fradin get into the mechanics of how great CIOs operate, why the CFO relationship matters more than reporting lines, and what changes when agents start doing real work inside your org. The conversation is direct, specific, and covers territory most AI podcasts skip entirely.

Topics discussed:

• Why the CIO is often the right starting point in any enterprise technology sale

• How CFOs became the natural home for IT to report into

• Where CIO and HR functions are headed as agent workforces grow

• Why agents need their own directory identity separate from the humans who authorize them

• What separates AI projects that scale from the ones that stall

• The small-group peer learning method that actually moves AI adoption

• Why most orgs are focused on the wrong AI security risk

• Which job functions will see the clearest AI impact based on what LLMs are actually built for.

Listen to more episodes: 

Apple 

Spotify 

YouTube

Ajay Sabhlok, CIO and Chief Data Officer at Rubrik, runs one of the more disciplined AI governance operations you'll hear about from a practitioner at this level. He set up a formal AI council with his CFO, co-founder, head of engineering, and general counsel to review every tool request coming into the organization, and he walks through exactly how that structure works in practice, including how Legal now publishes a list of conditions attached to every approved tool.

What makes this conversation worth your time is where Ajay is today versus where most CIOs still are. He's past the risk debates. His workforce is educated, excited, and flooding his team with use cases. The bottleneck is now capacity, not buy-in, and that shift reveals a lot about what a mature AI program actually looks like from the inside.

Topics discussed:

• How Rubrik's cross-functional AI council routes, reviews, and approves tool requests

• Legal publishing conditional approval lists tied to data sensitivity requirements

• Why capacity, not skepticism, became the real internal obstacle

• The data quality foundation Ajay says must exist before any AI use case scales

• Why pay-as-you-go contracts are replacing multi-year agreements in AI procurement

• Why Ajay is more bullish on AI now than he was two years ago

• Sales and marketing as the next major AI productivity frontier after coding and call centers