The Infinite Backlog

NLW's term for the condition agents create: because an agent never tires and only stops when you stop feeding it work, work no longer has a natural end — and the undone remainder feels like a personal failure to delegate.

The idea

In the "assisted AI" paradigm there was a reasonable end to a workday: you couldn't do any more, so you stopped. Agents break that ceiling — "agents aren't you doing a thing… they don't get tired." The only reason an agent isn't working is that you haven't given it something to do. Since the amount of work to be done is effectively unbounded, "there is always something next." NLW: agents make it feel like "we have beaten the end boss of time."

The result is a new kind of overwhelm that wasn't on most people's radar as an AI risk. The promise (AI frees your time → more leisure) and the fear (AI negates your value) both proved "very far away from the lived reality of the most advanced users." The real lived experience: not finishing at 3pm, but forcing yourself to stop at 3am — "tearing themselves away from the next thing they could accomplish." A single person can now go deeper into the backlog than was ever possible, which is exhilarating — and also makes any gap feel like a tooling failure on your part.

NLW introduced the concept in a prior episode, "Why Agents Make Every Job a Startup."

Connections

• The Human Sandwich — one answer to the backlog: humans manage the start and end of each agent task, choosing "the next piece of work to do."
• Token Shortage — an infinite backlog run at full autonomy is also infinite token consumption; the economics force prioritization.
• Agent Autonomy Spectrum — OpenClaw-style "heartbeats" tried to chew through the backlog autonomously; many found it burned tokens without doing what they wanted.
• AI and Employment — Replacement vs Expansion — the backlog is NLW's intuition pump for why work expands.
• The Bottleneck Shift — the backlog's "what's next?" choice is the same scarcity Ethan Mollick names: when execution is cheap, deciding what to do is the bottleneck.

Sources

• The AI Daily Brief — Why Agents Still Need Humans — NLW recaps the concept; origin episode flagged as AIDB — Why Agents Make Every Job a Startup #to-read.