Aaron Levie Challenges the Fear: AI Will Not Replace Jobs

The fear that artificial intelligence (AI) will take over jobs has become a major concern for the public, fueled by frequent warnings from tech CEOs. In recent months, this fear has intensified, particularly in March, when tech companies announced nearly 46,000 layoffs, the worst monthly total in over a year. Many executives cited AI as a key factor in this workforce reduction, and this trend is beginning to manifest in Silicon Valley.

The Economic Index from Anthropic reveals that conversations about AI in the workplace are increasing across nearly all white-collar professions. Recent research suggests that entry-level jobs, particularly vulnerable to large language models (LLMs), are already showing signs of disruption. This phenomenon is prompting workers to react, as evidenced by the vote of Google DeepMind employees in the UK to unionize, highlighting the significance of the situation.

Despite these concerns, the impact of AI on productivity remains difficult to gauge. For instance, Amazon plans to hire a similar number of software engineering interns in 2026 compared to previous years. Furthermore, job postings for software engineering roles are currently at their highest level in three years, raising questions about the true nature of this transformation in the labor market.

In this context, a new mini-series of podcasts on Platformer aims to explore these issues in depth. Over the next seven weeks, discussions with CEOs, operators, and academics will examine the history of AI and jobs, with a particular focus on recent data. Ella Markianos, a colleague of the host, will participate in each episode to analyze the latest surveys and research on the intersection of technology and work.

For the first episode, the host spoke with Aaron Levie, CEO of Box, a well-known figure in Silicon Valley. Levie, who introduced the host to the concept of software as a service (SaaS) in 2010, continues to be a strong advocate for this model. Box has actually fared better than most of its SaaS peers over the past year, despite predictions that traditional enterprise software would soon be surpassed by AI agents.

Levie does not share this pessimistic view. During their conversation, he presented a detailed and sometimes provocative argument for why he believes the "SaaSpocalypse" and fears of massive job losses due to AI are exaggerated. According to him, AI agents will actually increase the number of workers using enterprise software rather than replace them. He argues that the "last mile" of human work is far more resilient than one might think.

Levie explains that if one believes AI is replacing professions like engineer, accountant, or lawyer, it is actually addressing the initial 80% of the work. The remaining 20%, which represents all the value creation, relies on expertise and domain knowledge, and cannot be replaced by AI-generated text. He predicts that tomorrow's engineer is more likely to work in a pharmaceutical company than at Meta.

Finally, Levie emphasizes the importance of data security and governance in a world where AI agents are increasingly interacting with systems. These aspects are crucial to ensuring that AI is used ethically and responsibly.

To listen to the full discussion, visit the Platformer podcast or watch it on YouTube. Your feedback is welcome at casey@platformer.news.