Redis Creator Salvatore Sanfilippo Advocates AI-Assisted Coding, Citing Efficiency Gains

Salvatore Sanfilippo, the creator of Redis, has publicly stated that manual coding is no longer a wise approach, advocating for the use of artificial intelligence in software development. Sanfilippo, known online as Antirez, shared his perspective in a blog post titled "Don't fall into the anti-AI hype," asserting that AI has fundamentally altered the landscape of programming.

Human hand interacting with a glowing, abstract digital interface representing AI.

Sanfilippo's shift in perspective comes after a career marked by a preference for meticulous, hand-coded C language development for Redis. He described his current view as an "ultimatum to all programmers," emphasizing that despite his personal affection for traditional coding and reservations about AI's economic impact, the technological shift is undeniable.

From Coder to AI Orchestrator

Sanfilippo's engagement with AI's implications dates back to late 2018, when he began to step back from daily Redis operations. He subsequently authored "Wohpe," a science fiction novel exploring AI, climate change, and societal transformation, and publicly discussed the need for Universal Basic Income to address AI-induced unemployment. He initially believed the full impact of AI was years away but now acknowledges that technological evolution has accelerated beyond his earlier projections.

Conceptual image showing transition from manual coding to AI-driven project management.

He noted that for him, a developer who built his career on crafting code line by line, this admission is significant. Sanfilippo now argues that unless coding is pursued purely for enjoyment, manual development is no longer logical. He envisions programmers transitioning from "coders" to "product managers," conceptualizing solutions and directing AI to execute them. He also cautioned that human programmers cannot compete with the 24/7, low-cost output of AI-driven development.

Demonstrating AI's Capabilities

Sanfilippo presented three case studies to illustrate AI's efficiency in coding and debugging:

Server rack with glowing lights and overlaid C code, symbolizing rapid prototype generation by AI.

Rapid Prototype Generation: He tasked Claude Code, an AI, with creating a pure C language BERT model inference library. The AI produced 917 lines of C code in just five minutes. Testing revealed the output was consistent with PyTorch, with a performance difference of only 15%. Sanfilippo highlighted the stark contrast between this five-minute AI output and the weeks a human engineer would typically require for a similar task.

Advanced Debugging: Sanfilippo described a complex, transient Redis kernel bug involving TCP deadlocks and subtle timing issues. He reported that Claude Code autonomously investigated system states, reproduced the environment, analyzed the logical chain, and precisely located and fixed the bug.

Digital hand pointing to a highlighted bug in a complex network diagram, symbolizing AI debugging.

Accelerated Development: Sanfilippo had previously spent several weeks redesigning the internal implementation of Redis Streams. He provided his original design document to Claude Code, which replicated weeks of his work in under 20 minutes. He noted that much of this time was spent by him reviewing the code and authorizing commands.

Sanfilippo characterized these examples not as incremental improvements but as fundamental changes in development speed, describing them as a "time black hole" and "time warping." He stated that human experience, often valued in programming, is being compressed to milliseconds by silicon-based computing power.

Abstract digital vortex symbolizing a 'time black hole' or 'time warping' effect in development.

Addressing Resistance to AI Adoption

Sanfilippo criticized what he terms "anti-AI delusion," referring to the tendency among some programmers to focus on AI's errors or limitations. He compared this resistance to the skepticism faced by early automobiles from horse-drawn carriage operators. He warned that this mindset could be detrimental to careers, as the market values efficiency and cost-effectiveness.

He outlined two potential paths for programmers: "old humans" who adhere to manual coding and risk becoming obsolete due to higher costs and slower output, and "centaurs" who integrate AI into their workflow to amplify their capabilities. Sanfilippo emphasized that denying AI's impact will not control the future and urged programmers to embrace AI tools.

Allegory of two paths: one resisting AI, one integrating with AI, represented by a centaur.

He concluded by advising developers to thoroughly test AI tools, dedicating weeks to refactor their workflows rather than dismissing them prematurely. Sanfilippo suggested that AI, by automating tedious tasks, allows programmers to focus on the creative aspects that initially drew them to the field. He framed the current period as a critical juncture, offering a choice between resisting technological change and adapting to become a "new creator."