Domo’s CDO argues AI hype is driven by fear marketing, producing theater over results, and companies should start with specific business problems instead.
Key Takeaways
LLMs marketed as doing “anything for anyone” give companies no clear product spec, leading to misaligned expectations and wasted spend.
“Tokenmaxxing” – forcing high token usage to signal AI adoption – doesn’t move the bottom line even if individual productivity rises.
Klarna’s cycle of replacing customer service staff with AI then rehiring humans is cited as the canonical cautionary example.
Willis recommends starting with narrow, verifiable automations: invoice discrepancy checks surfaced for human review, not moonshot replacements.
CFO scrutiny is rising; budget reckonings are coming for orgs that bought AI tooling without measurable outcomes.
Hacker News Comment Review
Commenters flagged the credibility gap: Domo itself rebranded to “Governed Data for AI Agents,” undermining the CDO’s anti-hype message, and at least one commenter cancelled their Domo subscription calling it obsolete.
A counter-thesis gained traction: because LLM switching costs are near zero and there is no durable first-mover advantage, FOMO is largely irrational on competitive grounds.
Some commenters pushed back on the “failed POCs are waste” framing, arguing that failed experiments are how engineers calibrate trust in a new technology before putting it in load-bearing systems.
Notable Comments
@kvgr: Near-zero distribution and switching costs mean any first-mover advantage collapses within months when better tools ship.
@himata4113: The real threat is small teams now able to replicate paid products cheaply, not large companies outpacing each other with AI.
