Google DeepMind CEO Affirms Developer Value Amidst AI Advancements; Google Unveils Subscription and Notebook Enhancements

Demis Hassabis, CEO of Google DeepMind, has reiterated the essential role of software developers in the era of Artificial Intelligence, emphasizing that AI should be leveraged to augment productivity rather than prompt layoffs. Speaking to Wired, Hassabis highlighted the immense opportunity for companies to scale operations by making engineers “three or four times more productive” with AI tools. This perspective aligns with broader industry trends, as major AI players like OpenAI and Anthropic are establishing their own consulting firms to facilitate AI integration into core business processes, signaling a significant demand for AI implementation expertise. Hassabis specifically praised the coding capabilities of Google’s latest Gemini 3.5 Flash model, which excels at complex agentic coding tasks and code translation, yet stressed that this advancement does not signify an obsolescence of human developers. He argued that businesses seeking to replace developers with AI exhibit a “lack of imagination and understanding” of AI’s transformative potential.

In related developments, Google has updated its AI Plus subscription plan, making it more accessible and robust. The monthly price has been reduced to $5 (or €5 in Spain, down from €8), and cloud storage has been doubled from 200 GB to 400 GB. This plan includes access to Gemini, AI features in Google Photos, and NotebookLM, though advanced features, potentially referred to as “Antigravity” in the transcript, remain exclusive to the AI Pro tier. Concurrently, NotebookLM, Google’s AI-powered research assistant, has received significant enhancements. New functionalities include code execution for Gemini 3.5-generated code, a “thread of thought” visualization, and vastly improved export capabilities, allowing users to save notebooks to formats like PDF, Word, Excel, Markdown, PowerPoint, and CSV for images. Agentic research features have also been introduced, enabling the AI to dynamically search the internet and YouTube to add sources. For those seeking open-source alternatives, the “Open Notebook” project (boasting 28,000 GitHub stars) offers a completely free, private notebook solution compatible with various AI models (OpenAI, Anthropic, Llama, local models) and supporting multi-language podcast generation with up to four voices, surpassing NotebookLM’s two-voice limit.