Uh-oh: Someone senior wants to announce an "AI breakthrough." Your engineering colleagues look nervous. Here's how to navigate this tension.
The reality of AI "breakthroughs":
True breakthroughs are rare. Most AI advances are incremental improvements, not paradigm shifts. Genuine breakthroughs typically take months or years.
Engineers fear overpromising. They've seen many "revolutionary" AI announcements followed by disappointing performance. They protect credibility by being cautious.
CXOs want competitive differentiation. In a crowded AI market, incremental improvements don't generate buzz or investment interest. They need to stand out.
Snazzy words have a credibility problem. Years of overhyped announcements have made audiences skeptical of superlatives like "breakthrough," "revolutionary," and "game-changing."
How to manage the CXO-engineer divide:
Ask for specific metrics: "What exactly improved, and by how much?"
Focus on customer impact: "How does this change the user experience?"
Armed with this information, you can recommend a communications approach better than relying on superlatives - one anchored in specific, measurable impact:
❌ "Our breakthrough AI system revolutionizes customer service"
✅ "Our new AI system reduces customer wait times by 60% while maintaining 99% satisfaction scores"
❌ "Game-changing AI breakthrough in medical diagnosis"
✅ "Our AI diagnostic tool identifies conditions 15% faster than previous methods, with validation from three major hospitals"
Lasting credibility in AI won’t be achieved by bold claims swiftly proven to be false. It’ll be by making claims you can consistently deliver on.
More rapid reads
Never claim 100% accuracy with AI
Today's AI is the worst you'll ever use