Andrew Ng Criticizes Term 'Vibe Coding,' Emphasizes Its Intellectual Demands and Benefits

Andrew Ng, a prominent figure in artificial intelligence and a former Google Brain scientist, has spoken out against the term "vibe coding." Despite its catchy and seemingly relaxed connotation, Ng believes the term is misleading and detracts from the rigorous nature of AI-assisted coding. Speaking at the LangChain Interrupt conference in May, Ng emphasized that using AI to write code is far from a casual activity. Instead, it is "a deeply intellectual exercise" that can be as taxing as traditional coding, leaving him and his team exhausted after a full day's work. The term "vibe coding" was coined by Andrej Karpathy, co-founder of OpenAI, in February. Karpathy described it as a process where developers can give AI prompts to generate code, allowing them to "fully give in to the vibes" and "forget the code even exists." This idea has sparked significant discussions within the tech community, leading to concerns about job security among engineers and debates about the necessity of technical skills for startup founders. Ng is optimistic about the benefits of AI-assisted coding, noting its potential to accelerate software development. He highlighted that developers can write software more quickly, often with minimal direct interaction with the code itself. However, he is critical of companies, particularly those led by CIOs and CTOs, that continue to resist the adoption of AI coding tools. "We have to get past that," he urged, stating that his own teams are unwilling to return to traditional coding without AI assistance. Moreover, Ng debunked the notion that AI will render coding obsolete. In a recent discussion, he stated, "Over the last year, a few people have been advising others to not learn to code on the basis that AI will automate coding. I think we'll look back at some of the worst career advice ever given." Instead, Ng strongly advocates that everyone should learn at least one programming language, emphasizing the importance of understanding how to direct a computer's actions accurately. According to Ng, learning to code enhances one's ability to communicate effectively with AI tools. It enables individuals to create more precise prompts and interpret errors correctly, driving significant productivity gains. He cited his startup AI Fund as a prime example, where employees from diverse roles, including the CFO, general counsel, and receptionist, all possess some level of coding knowledge. Their proficiency, though not at the level of professional software engineers, aids them in their respective tasks by allowing them to leverage computational tools more effectively. The shift towards AI-assisted coding has also democratized app development, making it accessible to non-technical people. For instance, a product designer at A Block, who lacks formal engineering training, managed to build a dog identification app in two months using AI prompts. This anecdote underscores the potential of AI to empower individuals who may not have a background in software development. Industry insiders generally agree with Ng's stance, highlighting that AI is a tool rather than a replacement for human expertise. While AI can automate repetitive and mundane tasks, creative problem-solving and deep technical understanding remain crucial. Companies that integrate AI-assisted coding effectively are likely to see substantial benefits, including increased efficiency and innovation. Ng's background includes notable achievements such as founding Google Brain and contributing significantly to the field of deep learning. His current venture, AI Fund, focuses on investing in and supporting AI-driven startups, further solidifying his role as a key influencer in the tech sector. His insights on the evolving landscape of software development reflect not only his technical expertise but also his broader vision for how AI can enhance productivity and inclusivity in the tech industry.