Domo CDO: Stop AI FOMO, Start with Spreadsheets
Chris Willis, chief design officer and futurist at data platform Domo, says companies are wasting money on AI because they treat it as a solution rather than a tool. In an interview with The Register, he argued that the 'act now or be left behind' marketing creates anxiety, not real innovation.
Key points from the interview
- Product without a spec — 'The feature spec is: It'll do anything for anyone, anyway, anyhow, in any language.' That makes it impossible to define what the product actually does, leading to confusion.
- Tokenmaxxing is theater — Buying lots of AI tokens and pressuring employees to use them doesn't change the bottom line. It's a narrative to show action, not real strategy.
- Start with a spreadsheet — Instead of moonshot goals, Willis suggests automating a simple process tied to a spreadsheet. He cited a customer app that checks invoices for discrepancies and surfaces anomalies for human review. The team was thrilled with the result.
- Fear is not durable — 'Fear is not a durable strategy for innovating,' Willis said. Companies should understand their workflows before adding an AI engine.
What this means for developers
If you're being asked to build AI features without a clear problem, push back. Start with something small: automate a manual check, reduce toil, and prove value before scaling. The CDO's advice is essentially: understand the process first, then pick the right tool.
📖 Read the full source: HN AI Agents
👀 See Also
Benchmarks Show Distilled Models Match Frontier LLMs on Structured Tasks at 10x Lower Cost
A comprehensive comparison of small distilled Qwen3 models (0.6B to 8B) against frontier LLMs shows distilled models match or beat mid-tier frontier models on 6 out of 9 tasks at dramatically lower cost, with Text2SQL achieving 98.0% accuracy at $3/M requests versus $378 for Claude Haiku.
Cambridge researchers develop hafnium oxide memristor for low-energy AI chips
Researchers at the University of Cambridge have created a hafnium oxide-based memristor that switches currents a million times lower than conventional oxide devices, potentially reducing AI hardware energy use by up to 70%.
Measuring the Sloppiness of Code: Verbosity and Erosion Metrics
LLMs are almost perfect at generating correct code, but correctness does not mean the code is free of duplication, unnecessary abstractions, or bloat. Two metrics from SlopCodeBench, verbosity and erosion, help quantify code sloppiness.
