According to Sundar Pichai, CEO of Google and its parent company Alphabet, the company is now utilizing AI to generate over 25% of its new code, as revealed during their latest earnings call. Pichai explained that Google is "employing AI internally to enhance our coding processes, leading to increased productivity and efficiency. " He noted, "Currently, more than a quarter of all new code at Google is AI-generated, subsequently reviewed and approved by engineers. This allows our engineers to accomplish more in less time. " While the concept of using generative AI to write code is not entirely new, it could potentially lead to a reduction in software engineering jobs, particularly for entry-level positions, as remaining employees might take on more responsibilities. Earlier this year, startup Cognition Labs unveiled its AI "software engineer, " named Devin, which can execute entire engineering projects with minimal human involvement. Additionally, last year, ChatGPT reportedly demonstrated the capabilities of a Level 3 software engineer by passing a Google coding interview. However, utilizing AI for code generation could raise copyright concerns and create security vulnerabilities, especially if the AI is trained on proprietary or outdated code. Some companies that have employed AI-generated code have already faced outages and cybersecurity issues, primarily due to insufficient human oversight. Google is fully committing to AI initiatives, such as featuring AI Overviews at the top of Google Search, enhancing YouTube with AI tools, and developing a new Gemini model. Reports suggest that Google is planning a December launch for its latest Gemini model, potentially called Gemini 3. 0, coinciding with the release of a new model from its main competitor, OpenAI. The upcoming AI model may function as an autonomous agent, capable of completing tasks by controlling a computer, navigating various applications, typing, and clicking to finalize projects or carry out research.

This could enable Google to delegate even more internal tasks to AI models in the future. If this becomes a reality, it would build on Anthropic's introduction of a comparable feature, Computer Use. However, AIs can sometimes generate false information or be misled, raising significant security concerns and highlighting the essential need for ongoing human oversight. Moreover, the growing reliance on AI tools contributes to increased energy consumption and elevated data center costs, particularly in regions facing power shortages or unreliable grids. The current substantial electricity and water demands of AI could significantly impact the environment, and Google has already seen a 50% surge in emissions over the past year due to its AI usage. In response, the company is considering nuclear energy solutions to mitigate this energy strain. Currently, AI appears to be beneficial for Google's financial performance. Alphabet's stock has risen approximately 2% in the past day and 36% in the past year. Meanwhile, Nvidia has experienced significant gains within the AI sector, although competitors like AMD have not benefited as much from the generative AI trend.