ShipTalk - SRE, DevOps, Platform Engineering, Software Delivery · By Harness

CTO Predictions for 2026: How AI Will Change Software Development (with Harness Field CTO Nick Durkin)

·41 min·2 clips

Nick compares software teams to football teams that think they play the same sport but use different rule books.

1. ShipTalk covers 2026 predictions for AI’s effect on software development, DevOps, platform engineering, and software delivery. 2. Devon Ahmed hosts the episode, and Harness Field CTO Nick Durkin is the guest because he gives the “spicy takes” on AI in production. 3. The episode asks whether AI will change how teams ship software faster without breaking security, quality, or employability. 4. Nick says an AI-caused meltdown is more likely to come from generative code or AI-assisted development that skips steps and misses checks. 5. He compares that failure mode to Waymos getting stuck in the street, where human intervention and preparedness still matter. 6. Nick says confidence in AI will come from multiple AIs checking one another, not from a single system acting alone. 7. He describes this as triangulation, where one model helps define confidence and another validates the output. 8. On cleanup work, Nick says teams are already hiring AI cleanup crews after moving fast and “gone a little over our skis.” 9. He does not think cleanup crews are the durable answer, because the real fix is harmony, policy, and controls in the pipeline. 10. He gives a Mario analogy where each level exposes the next vulnerability, and the pipeline documents what was tested before production. 11. Nick argues that DevSecOps has to become more integrated, because shifting security left or right still leaves teams segmented. 12. He says security should provide the rule book and blast-radius context at the right time, not two weeks later. 13. For MLOps, he groups models, agents, and prompts with other production artifacts that need deployment, testing, validation, and security scans. 14. He says customers do not want 17 different tools in 17 different places when the same platform can handle the workflow. 15. On agents, Nick says humans may start behaving more like agents by focusing on specific tasks rather than everything at once. 16. He estimates developers spend only about 20 to 30 percent of their time on the work they love, with the rest swallowed by toil, backlog, and delivery overhead. 17. Nick’s tone is conversational and analogy-heavy, with repeated references to football, Google Maps, Mario, and GPS-style routing. 18. The format is a back-and-forth interview with Devon Ahmed pushing scenario questions and Nick answering with concrete operational examples. 19. Engineers, platform teams, and security leaders interested in AI-native delivery will get the most value from this episode. 20. Listeners looking for a terse product announcement or a pure coding tutorial may skip it.

As heard by us

A grounded 2026 AI-and-security talk that uses sports rules to make a clear point about trust, speed, and control.

Nick Durkin frames the 2026 predictions discussion around a simple systems idea: software teams move faster when they are working from the same rule book.

Read the full review in PlayNext →

Why you'd press play

Hear why software teams move faster when they are playing the same game, and why AI security now has to be handled in-line, not thrown over the fence.

Read the full recommendation in PlayNext →

Listen to the show on

Apple Podcasts Spotify

Play clips from this episode More from ShipTalk - SRE, DevOps, Platform Engineering, Software Delivery