The Rise of AI and the Changing Face of Programming
Recently, a writer experimented with an AI co-pilot, prompting it to analyze the contents of his refrigerator and create an app to help with lunch planning. This innovation exemplifies the ongoing changes that AI is bringing into the world of programming.
In 2023, AI guru Andrej Karpathy made a bold statement, declaring English “the hottest new programming language.” This claim emerged from the capabilities of large language models (LLMs), which allow humans to instruct computers in natural language, like speaking to a gardener. The transition has been rapid, with programmers being the earliest adopters of LLMs, using them as assistants or “co-pilots.” However, it’s important to remember that, like interns, they make mistakes, and human experience is still needed.
Karpathy suggests a new era of “vibe coding,” wherein programmers rely heavily on LLMs to generate code, embracing a more intuitive approach. “When I get error messages I just copy [and] paste them in with no comment, usually that fixes it … I’m building a project or web app, but it’s not really coding – I just see stuff, say stuff, run stuff, and copy paste stuff, and it mostly works,” he claims.
This sentiment has resonated with tech columnist Kevin Roose, who, despite not having coding experience, reported that he has “been coding up a storm” using AI. He used an AI co-pilot to build “LunchBox Buddy,” an app that suggests lunch ideas.
However, some experts, like Gary Marcus, scrutinize this perspective, noting that Roose is celebrating regurgitation of pre-existing code.
This AI shift may divert attention from a deeper understanding of what an AI-influenced future looks like. LLMs’ software-writing abilities give an excellent way to assess the technology’s potential for augmenting human capabilities. Programmers have explored the advantages and disadvantages of this. That is where Simon Willison has become an indispensable source for AI analysis, and uses LLMs. Willison’s insights show how these AI co-pilots aid in coding, providing a valuable framework for understanding the technology’s effectiveness and its limitations.
The Future of Programming
Programming is an elite trade. AI is affecting this, as with other elite professions. Though AI will change the environment, it will not make programmers redundant. As Tim O’Reilly noted, AI will not replace programmers; instead, their jobs will transform, as well as other elite trades.
