The AI Stack: A New Layer in Engineering? | Users Embrace Agents in Development
Edited By
Carlos Mendez
A growing trend points to the integration of AI tools in software development. Many developers are adapting to new agents like Aider and Copilot as an essential part of their workflow. The spontaneous rise of this "AI stack" raises questions about its permanence.
The Current State of AI Integration
Reports from various development communities highlight how common it is for people to use multiple AI agents daily. People are swiftly jumping between platforms such as Cursor, Windsurf, and Cody, viewing these tools as standard components in their coding practices.
"Our entire team uses Augment Code. It has become a very important aspect of our dev stack," one developer noted.
This casual adoption suggests a paradigm shift in how software engineering is approached. While some developers view it as a technological advance, others speculate about its long-term impacts on the development process.
Diverse Tools in the Development Toolkit
Feedback from various forums shows a range of preferred AI tools:
GitHub Copilot: Users emphasize how it helps automate tasks efficiently.
Claude Code with GLM: Favored by some for its backup capabilities against other models.
VS Code with Copilot: A combination that many have grown fond of for its versatility.
One coder shared, "I’ve been using GitHub Copilot in the GitHub UI lately, and it’s been great. I find it easy to spin off a bunch of tasks in parallel with it."
Are We in Experimental Stage?
The crucial question arises: is the integration of AI tools a temporary trend, or is it a fundamental change in engineering? As reliance on these tools increases, developers share mixed insights about their effectiveness.
While many embrace the efficiency they bring, discussions indicate a cautious optimism about potential over-dependence on AI for programming tasks.
"It might be better options, but I’ve settled on GLM for the foreseeable future. I like having other state-of-the-art models to fall back on if GLM can’t do something," a developer remarked.
Key Insights into the Shift
⚙️ Adoption of AI agents is becoming routine among developers.
🔄 Many prefer specific tools that enhance their productivity.
❓ The long-term sustainability of these AI layers remains uncertain.
As the conversation around the "AI stack" grows, it's clear that this new layer of engineering is reshaping how software is developed, but whether it will last remains to be seen.
What Lies Ahead for the AI Stack?
Looking forward, there’s a strong chance that AI tools will further embed themselves in standard development practices. Experts estimate around 70% of developers will depend heavily on AI agents within the next three years. As companies aim to increase productivity, expectations to enhance automation are high. However, there's an undercurrent of caution among engineers, indicating that while the convenience may thrive, the risk of over-reliance could hinder problem-solving skills in the long run. This shift could lead to a bifurcated landscape where some developers thrive with AI assistance while others may struggle to keep pace without it.
A New Chapter in Tech Workforce Evolution
This scenario has parallels with the rise of personal computers in the 1980s. Initially, many office workers feared that these machines would undermine their careers, yet they soon found themselves empowered by the technology to boost productivity and creativity. Just as office culture transformed with desktop computing, the influx of AI tools in software development could redraw professional lines. Programmers today might feel the same trepidation as accountants did back in the day, faced with novel systems that forever changed how their jobs are performed—all while creating entirely new opportunities for innovation and growth.