A multi-speaker YouTube discussion (uploaded April 17, 2025) about how agentic AI is reshaping software development. Speakers cite surveys and industry predictions (e.g., an Anthropic CEO prediction and a 250-developer survey) and argue that AI agents will shift developers from repetitive technical tasks toward higher-level strategic work, code reuse, and business-impact-focused projects. Examples mentioned include using agents for code reviews, compliance checks, and discovering internal reusable components.
Chapters
Software developers are going through a seismic AI power transformation. In fact, the CEO of Anthropic predicts that just six months from now, AI will write 90% of all code. There is a perception that all developers are concerned about being replaced by AI agents. But our recent survey of 250 developers found that 96% of them are enthusiastic about the potential of agents to assist with day to day work.
AI agents are shifting my role from being purely technical into a more strategic side of the business. So now, instead of spending time on, let's say, repetitive code analysis, I'm basically focusing on designing more scalable architectures, optimizing business processes and, you know, driving innovation.
In fact, another survey from Slack found that nearly two thirds of developers believe AI agents can help build better code, ship it faster and boost their personal creativity. For example, DevOps automation company Capado will start using agents for code reviews and compliance checks.
What that means is maybe less time on fixing customer bugs than reviewing pull requests, and more time on building net new applications or net new features.
In other cases, AI agents can identify code that's already in use in another part of the company.
I think one of the great things is the internal reuse of code. A tool, an architectural tool, can be like, listen, you already have a tax calculator tool, you already have an order management tool. How do I discover and reuse versus re implement?
Traditionally, developers have measured their productivity only using metrics like app build time, frequency of deployments and number of software issues addressed. But now developers can spend more time understanding business needs. If you work for a bank, for instance, you can spend more time on customer centric projects like writing code that reduces loan approval time instead of just measuring the amount of code you deployed that week. With AI agents, developers can focus more on impact and not just output.
