If you are a software developer or coder, you may have heard that AI could soon replace your role. Recently, Mark Zuckerberg announced a timeline regarding this shift, at least for his company. In a podcast, Zuckerberg mentioned that within the next 12 to 18 months, most of the code for the company’s Llama project will be written by AI. He stated that AI is already performing at the level of an "average very good person" in a team and can automatically complete sections of code. He believes that soon, AI will surpass even the best coders and will be capable of independently writing high-quality code, running tests, and finding bugs for Meta.
Zuckerberg in a podcast interview with Dwarkesh Patel explained that several coding agents are being developed within Meta. The focus is not on transforming the company into an enterprise software manufacturer; rather, these tools are being created for internal use with specific objectives in mind. The intention is to develop a coding agent and an AI research agent aimed at advancing Llama research. This initiative is fully integrated into Meta's toolchain and is considered significant. He suggested that this effort will play a crucial role in how work is accomplished.
Zuckerberg has previously indicated that AI will eventually write all the code for applications, replacing mid-level software developers. This perspective is also echoed by Anthropic CEO Dario Amodei, who predicts that by the end of 2025, AI will be able to generate all code, with an initial estimate of 90 percent in the next three to six months. The tech industry is increasingly relying on AI for coding, as noted by Google CEO Sundar Pichai, who stated that 25 percent of its code is AI-generated, and OpenAI CEO Sam Altman, who reported that AI accounts for 50 percent of the code in some companies.
Should you be concerned?
Zuckerberg has revised his timeline for when AI might replace mid-level engineers. Initially, he predicted that by 2025, AI could write code comparable to that of a mid-level engineer at Meta and other companies. However, he later stated that such systems would not be ready until at least mid-2026. This change underscores the importance of critically evaluating the promises made by the AI industry, as predictions often serve as marketing for technologies that are still in development and may not perform as expected.
ALSO READ: OnePlus 13R now available for Rs 25,999 along with free Buds 3: Here's where to buy
