Insider Interviews: Media and Marketing Pros

A handful of genuinely interesting distinctions emerge (causal AI predicting 'next number' vs. LLMs predicting 'next word', the nodes/edges graph framing for enterprise causality, the Delta Olympics example), but these are surrounded by significant promotional filler and repetitive high-level language that dilutes the idea-per-minute rate.

“LLMs were not really designed to predict next number. They were designed to predict what the next word should be.”

“you cannot hallucinate your way through a multi-million dollar strategic budgeting or pricing decision”

The 'next number vs. next word' framing is a clean, moderately fresh articulation of causal AI's differentiation from generative AI, but the rest of the conversation rehashes familiar measurement-industry pain points (CMO/CFO disconnect, silos, attribution) without producing genuinely contrarian or first-principles arguments.

“you cannot hallucinate your way through a multi-million dollar strategic budgeting or pricing decision”

“there's no differentiation in the causal ai graph between brand versus performance it's all connected”

Hitesh Wadhwani has legitimate practitioner credentials - 12 years at Google, hands-on work building Google Meridian and contributing to Google Analytics - but he is only three weeks into Alembic at time of recording, visibly limiting the depth and authority of his answers about the product he represents.

“I spent 12 years, which was definitely a very interesting, enriching journey”

“contributing to Google making or building its own measurement stack, building products like Meridian, which have been used by thousands of advertisers globally”

The Delta Airlines Olympic campaign ($30M spend, specific insight about the medal-ceremony Eiffel Tower logo placement driving Paris flight bookings) is a concrete, memorable case study, but beyond that single example the episode is largely devoid of hard metrics, customer outcome numbers, or verifiable technical benchmarks.

“Delta Airlines had like this huge investment of almost 30 million dollars for the Olympic campaigns”

“during the medal ceremony in front of the Eiffel Tower during the medal ceremony the placement of the Delta logo drove most of the flies to Paris than any other moment in time”

The host is engaged and did light pre-research (reading a LinkedIn quote back), but the conversation is dominated by softball framing ('you're so smart,' 'that's brilliant,' 'too good to be true?') with no genuine pushback on product claims, no probing on competitive differentiation, accuracy validation, or pricing - leaving promotional assertions entirely unchallenged.

“I researched you because I'm very anxious about speaking with you. You're so smart.”

“So all of this almost sounds too good to be true.”