“Slop cannons” is Darragh Curran’s term for the fear that AI-generated code will degrade craft. The fear is real. The same fear runs through design: AI-generated interfaces will be derivative, generic, indistinguishable from each other. Curran is Intercom’s CTO, and he published a detailed report on what happened when Intercom went agent-first across their entire R&D org. The result: 3x productivity in 16 months, tracked across nine metrics. The code quality results were not what anyone expected.

Curran:

A legitimate worry with the use of coding Agents, is that they won’t write high-quality code and the craft we’ve fought to protect will be undermined by slop cannons. We have a system to rate the structural quality of code contributions using static analysis and various rules/heuristics. It’s clear that prior to agentic coding, this metric would oscillate up and down above the line. As we started to use AI for writing more and more of our code, the overall quality (by this measure) declined. My intuition was that this was inevitable in the short term, but correctable in the medium term, as models and harnesses get better. We are starting to see this and recently had possibly our first ever five-week streak of net positive code quality overall.

Quality did dip. He confirms it. The slop cannon fear describes a real phase: at 93.6% agent-driven PRs, when agent-generated code degrades, the whole org feels it. But there’s a second finding:

There is huge latent potential. Some people are really pushing the limit of what is possible, tokenmaxxing, doing really interesting things, while others have only really made incremental changes to how they’re working and don’t see much change in their personal throughput. Ultimately one of the biggest bottlenecks to progress is with humans; how we work together, how we change behavior, etc.

Intercom’s top 5% of contributors produce 6x the median PR throughput. Those are the people spending over $1,000 a month on tokens. That spread is the real finding from going agent-first. The slop cannon fear is about whether agents can execute well. The 6x gap is about who’s learned to orchestrate them, and Curran’s candid that most of his org is still finding out.

For design, we worry about going too fast, of solving the wrong problem, and building the wrong thing. Those are legitimate fears. Nonetheless, if you’re working in startupland as a designer, acceleration and automation are coming.