# The Future of Software Engineering for Tech Leaders ## Feat. Augment Code Co-Founder Guy Gur-Ari & Docker EVP of Engineering and Product Tushar Jain Jul 21 Written By [Laura Oppenheimer](/content/newsletter?author=685b252ea047d845b5ccd255/index.html) We were thrilled to host an in-person event at our office in San Francisco with [Docker CTO Tushar Jain](https://www.linkedin.com/in/tusharj/) and [Augment Code Co-founder Guy Gur-Ari](https://www.linkedin.com/in/guy-gur-ari/). Moderated by [Product Advisor Michael Sippey](https://www.linkedin.com/in/sippey/), we dove into the big topics facing technology leaders today: What’s the short term outlook for software engineers? How do we think about adoption of AI agents in software development across the enterprise? And why wouldn’t a team be all in on AI right now? **If you’d been there, here’s what you’d still be thinking about:** 1. **Things are moving so fast that driving any tops down edict about tools may backfire.** Tushar says at Docker, last year the primary tool being used was Copilot, but earlier in 2025, Cursor became the tool of choice. And then _in the last two weeks_ he’s seen a switch to Claude Code from his team. The advances are happening so quickly that it doesn’t make sense to pick one tool and sign a large contract for Docker — they are staying nimble and adapting to changes in the industry. 2. **Building with AI agents requires changing how engineers think about their craft in significant ways.** Guy notes that developers like “looking at code, thinking about code and being in the code.” Asking developers to build from a prompt box is a totally different skill set — how to structure a prompt, break down what you want to build, and give enough context. The skill of being a great engineer needs to adapt from defaulting to being in the code to being an expert on how to prompt — this is a significant change in the craft and for established senior engineers, will take time. 3. **Change is rapid, but we’re still at the beginning.** One of our event attendees asked how we should think about AI agents — are they more like employees? One-off task completors? Guy says that right now they’re like eager interns. The new agents will do whatever you prompt them to do — and do it quickly. But there still need to be code review, feedback, etc. For complete notes and to read all our takeaways (and to watch a recording of the event), [apply to become an Enrich member](/content/apply/index.html).