Arm’s Software Chief Envisions Human Language as the New Programming Paradigm

At a glance:

• Arm, a chipmaker powering Macs, iPhones, and smartphones, is shifting focus to building its own hardware using an AGI CPU.
• Arm’s senior vice president for AI and developer platforms, Alex Spinelli, sees human language as the new way to program.
• Spinelli emphasizes the importance of understanding computer fundamentals and agile skills for engineers in the AI era.

Arm’s Strategic Shift to Hardware Development

For over 40 years, Arm has been a cornerstone of modern computing, providing processor designs that power Macs, iPhones, and the vast majority of smartphones worldwide. Queries made through AI platforms like ChatGPT, Gemini, and Claude all pass through Arm-based chips at some point. However, the company’s trajectory is undergoing a significant transformation: Arm is now developing its own hardware, utilizing an AGI CPU that OpenAI and Meta will incorporate into their systems. This move positions Arm to compete directly with industry giants such as Apple, Intel, Nvidia, Amazon, and Google.

Arm’s new direction is further supported by its Performix software suite, which leverages “recipes” and AI insights to assist engineers in identifying problematic code and CPU hotspots. This suite represents a crucial component of Arm’s strategy to enable application developers to fully capitalize on its new hardware from the moment it is released. The AGI CPU is at the heart of this transformation, marking a departure from Arm’s traditional role as a licensing company for chip designs.

The Evolution of Software Engineering in the AI Era

According to Alex Spinelli, Arm’s senior vice president for AI and developer platforms, the landscape of software engineering is undergoing a profound shift. Looking back at the history of computing, we can trace a gradual progression toward higher levels of abstraction—from punch cards and assembly language to higher-order languages, interpreted languages, and now, the era where human language is becoming the highest level of programming.

Spinelli explains that programming itself is not disappearing; rather, the way we express it is evolving. “Programming doesn’t go away, engineering doesn’t go away. The way we express it is going away,” he notes. This transition is leading to a greater blending of technical product management thinking, design thinking, and architecture thinking within a new programming model that relies on natural language to create programs.

Navigating the Future of Engineering Careers

Spinelli’s perspective on the future of engineering careers is clear: embracing the new model is essential. He advises engineers to understand their place in the evolving tool chain, with AI agents playing a pivotal role. “Where AI rubber really hits the road is with agents. Agents use a lot of AI and agents are software,” Spinelli says.

For new entrants to the field, Spinelli acknowledges that the ideal mix of learning is still unclear. However, he emphasizes that AI tooling serves as a powerful tool for mid-to-senior engineers to adapt to this new paradigm. “If you look at the biggest technical innovations in the world — electricity, assembly line, railroad — they’re automations. When you radically reduce the cost of production of something, humans in history have not used less of it. People are finding new roles and new businesses are launching,” he explains.

Addressing the “Death of the Engineer” Prediction

Spinelli challenges the notion of the “death of the engineer,” attributing his deep understanding of computer science to his lack of a traditional computer science degree. He argues that even as large language models (LLMs) become more prevalent, the foundational knowledge of how computers work remains critical. “Think of an LLM like the smartest, most informed, overconfident, eager, arrogant recent MIT master’s grad. They know every language, but they would need a senior engineer to guide and help them,” Spinelli says.

He stresses that great engineers are more valuable than ever, as AI requires guidance and oversight. “The importance of great engineers has been elevated. AI needs that guidance,” he notes. Additionally, Spinelli calls for a revival of agile skills, which are essential for adapting to the rapid changes in the industry.

Preparing for the Future: Education and Pitfalls

When asked about the best approach for new developers to learn, Spinelli reflects on his own journey, which involved diving deep into assembly language and understanding how memory works. He suggests that even in an era dominated by high-level languages like English, a deep understanding of how computers work remains invaluable. “You might never write C++ or C code, but fundamentally understanding what’s happening is really important,” he says.

Spinelli also highlights the importance of formal education and training, though he acknowledges that there are many ways to educate oneself if one is motivated. “Go deep, understand how computers work, understand what a compiler is. It’ll pay dividends,” he advises.

In terms of the biggest pitfalls for engineers today, Spinelli identifies cost and security as major concerns. The expense of tokens, as seen in his own OpenClaw instance, can lead to unexpected bills if not properly managed. Similarly, security challenges are less inherent to the frameworks themselves and more about how people use them, such as exposing passwords and tokens in clear text. Spinelli recommends institutionalizing policies within agent frameworks to mitigate these risks.

The Future of AI-Built Software

The future of the AI-built world, according to Spinelli, is characterized by fast software development, akin to fast fashion. With the drastic reduction in production costs, we are seeing a trend toward disposable software. “We’re going to build things quickly. If they don’t quite work, that’s okay. The agent remembers how to do it. I’ll just rebuild it,” he says.

However, Spinelli cautions that this shift does not mean we can accept any kind of failure. “Things might fail hilariously or catastrophically, and then we’ll fix it in an automated way,” he notes. He envisions a future where every engineer has an expert sidecar agent and a swarm of agent developers they can rely on. “You use Claude Code or Codex or Gemini to spin up agents, each with a specific role…designer, architect, coder, tester. Research says when you bind an agent to a role with procedures, policies, and standards around it, and you allow those agents to interact, the outputs are orders of magnitude higher quality than leaning on a single agent,” Spinelli explains.

Adapting to the Rapid Pace of AI Change

Spinelli emphasizes the need for diverse opinions and people with different ways of thinking to make accurate projections in an era where AI changes every week. Traditional approaches like component-based, modular-based architectures, user-centered design, and service-oriented design remain important, but flexibility and adaptability are crucial. “You need the ability to flex and bend,” he says.

He also stresses the importance of not future-proofing too much, as assumptions often need to change. “The pace is new. We almost went away from agile in the industry. Resurfacing those principles…ends up being pretty important now because stuff’s changing,” Spinelli notes.

Key Takeaways

Arm’s strategic shift to hardware development, led by its software initiatives, marks a significant departure from its traditional role as a licensing company for chip designs. The company’s new AGI CPU and Performix software suite represent a crucial component of this transformation, enabling developers to fully capitalize on Arm’s new hardware from the moment it is released.

The evolution of software engineering in the AI era is characterized by a gradual progression toward higher levels of abstraction, culminating in human language as the new way to program. This transition is leading to a greater blending of technical product management thinking, design thinking, and architecture thinking within a new programming model that relies on natural language to create programs.

The future of engineering careers, according to Arm’s senior vice president for AI and developer platforms, Alex Spinelli, is centered on embracing the new model and not trying to fight it. He advises engineers to understand their place in the evolving tool chain, with AI agents playing a pivotal role. Additionally, Spinelli emphasizes the importance of understanding computer fundamentals and agile skills for engineers in the AI era.

Finally, the future of the AI-built world is characterized by fast software development, with a trend toward disposable software. However, this shift does not mean we can accept any kind of failure. Instead, Spinelli envisions a future where every engineer has an expert sidecar agent and a swarm of agent developers they can rely on, leading to orders of magnitude higher quality outputs than relying on a single agent.