Lenny's PodcastThe AI paradox: More automation, more humans, more work | Dan Shipper
At a glance
WHAT IT’S REALLY ABOUT
AI work paradox: automation expands work, empowers PMs and designers
- Work will bifurcate into (1) a company “super-agent” used in Slack and (2) a desktop agent workspace (Codex/Claude Cowork) where most daily work happens inside the agent’s environment.
- Agents won’t eliminate human work because they require continual human ownership, maintenance, and systems-building—creating new roles like forward-deployed/agent engineers.
- SaaS won’t die; instead, agents will increase SaaS usage while shifting token costs away from SaaS vendors as users “bring their own AI” into products via agent workspaces.
- Organizations will experience a throughput explosion (more PRs, more analyses, more documents), making review, coherence, deletion, and governance the new bottlenecks.
- The biggest near-term winners are PMs and full-stack designers who pair strong judgment and creativity with AI-assisted building, while the feared job apocalypse is unlikely if people “ride the models.”
IDEAS WORTH REMEMBERING
5 ideasExpect two primary AI modes at work: Slack delegation and desktop co-working.
Shipper forecasts a company-wide super-agent (often in Slack) for delegated tasks, plus a desktop agent environment (Codex/Cowork) where you and the agent operate in the same workspace via an embedded browser and shared context.
“Personal agents for everyone” is later; “one super-agent” is now.
He initially believed in personal agents mirroring each employee, but flipped after seeing setup/maintenance friction; today’s agents need a dedicated human owner, which pushes companies toward a central agent with clear stewardship.
Automation increases total human work because oversight becomes the job.
Even as models automate execution, humans must frame the right problems, verify outcomes, manage failures, and continuously improve the system—similar to how managers stay busy despite delegating work.
The real bottleneck shifts from building to coherence, review, and deletion.
AI makes producing code/docs/analysis cheap, so teams drown in pull requests and low-quality outputs; the scarce skill becomes deciding what fits the product/system, what to merge, and what to remove to avoid bloat.
SaaS is positioned to grow, not collapse—agents become new SaaS “users.”
Rather than replacing SaaS, agents drive more SaaS interactions and higher volume usage; Shipper argues the “SaaS apocalypse” is wrong and predicts increased demand plus new infra/pricing challenges for vendors.
WORDS WORTH SAVING
5 quotesI'm simultaneously extremely AI pilled and very bullish on humans. Automation is a lie. Every agent needs a human. We have so much automation, so much AI, and I also work way more.
— Dan Shipper
It's going to bifurcate in two main ways. One is everyone's gonna have at least one agent that they talk to that they can offload work to. Second is that most of the work that you do is actually going to happen on your computer in an environment like Codex or Claude Cowork.
— Dan Shipper
We speed ran the CLI era. It was nice while it lasted, but I think CLIs are over.
— Dan Shipper
I would buy SaaS stocks right now. Um, I would, I think the SaaS apocalypse is dumb... And I think that what agents do is increase the number of users of SaaS, not get rid of it.
— Dan Shipper
What models do in general is they make yesterday's human competence cheap, and so it becomes commoditized. It's not valuable anymore. What humans do is we go in there and we're like, "Yeah, we, we have all this frozen human competence from yesterday. How do I use this to, like, make something new and interesting?"
— Dan Shipper
High quality AI-generated summary created from speaker-labeled transcript.
