Ethan Mollick on why employers should treat AI as what it is: weird

News
 |  
Apr 2026
 |  
The Economist
Save to favorites
Your item is now saved. It can take a few minutes to sync into your saved list.

What: Treating AI as conventional enterprise software limits its transformative potential, preventing organisations from reimagining business models and value creation.

Why it is important: Companies that move beyond incremental automation and invest in workforce upskilling and experimentation are more likely to achieve lasting competitive advantage.

Many organisations are making a strategic error by treating artificial intelligence as just another enterprise software tool, slotting it into existing processes and focusing narrowly on automation and efficiency. This approach risks squandering AI’s truly transformative potential, as it discourages experimentation, creative risk-taking, and the reimagining of business models. When AI is managed solely by IT departments and measured by conventional KPIs, its use often becomes limited to routine tasks, missing the opportunity to unlock new products, services, and markets. The most successful companies are those where leadership drives AI strategy, encourages broad employee experimentation, and invests in dedicated teams to explore new applications. Without these elements, organisations face information gaps, hidden AI usage, and a lack of institutional learning, all of which undermine their ability to adapt and compete. Embracing AI’s “weirdness” and fostering a culture of innovation are essential for realising its full value and building future-ready retail organisations. 

IADS Notes: Recent industry analysis underscores that while AI investment remains a top priority for retail CEOs, as reported by The Wall Street Journal in December 2025, only a minority of retailers have realised substantial value from these initiatives. Deloitte in September 2025 found that just 10% of retailers have succeeded in scaling advanced AI solutions, with operational, regulatory, and workforce challenges slowing transformation. BCG in July 2025 emphasised that CEOs who limit AI to efficiency gains risk missing out on transformative business model innovation, while Forbes in March 2025 documented that leading retailers have achieved up to 30% improvements in customer service efficiency and notable revenue growth through AI, but most struggle to move beyond pilot projects due to organizational barriers. Bain & Company in December 2025 further noted that successful scaling of AI depends on leadership commitment, workflow redesign, and workforce upskilling. Collectively, these sources confirm that realizing AI’s full potential in retail requires visionary leadership, systematic learning, and a willingness to reimagine both business models and organizational culture.

Ethan Mollick on why employers should treat AI as what it is: weird