AI super agents will reshape work, not replace jobs
The premise of this episode is we're going to go through what else you predict will happen. The AI job apocalypse is not really a thing. I am super, super bullish on PMs and full-stack designers.
Automation is a lie. Every agent needs a human. We have so much automation, so much AI, and I also work way more. Creativity, it just feels like it's going to be more and more valuable to stand out from all the slop that people are shipping and launching constantly.
I honestly would prefer it to say that it's coming from GPT 5.5 and I make change it to do that. But I had this, I had this experience the other day where I had this I had this send an email to one of our investors and I asked Codex go do it and use usually Codex knows to ask me and it usually does, but this time it didn't and it just sent the email and I didn't look at it at all and I was like, 'F***!' And so I went to my scent and looked at it and I was like, oh, this is exactly what I would have sent.
- Dan Shipper previously predicted the rise of Claude Code for non-engineering work, and the hosts confirm his prediction was 'unbelievably right,' noting it was a major factor in Anthropic's success.
- The optimal AI experience will be an AI agent with a browser integrated into it, allowing the agent to see everything a user is doing on their computer. This will become super common within a year.
- SaaS tools will eventually run *within* AI-powered environments like Codex or Claude Co-work, rather than AI being baked directly into SaaS applications. This puts SaaS companies in a position where users bring their own AI tokens.
- The 'AI job apocalypse' is a myth; AI automation will actually increase the amount of work for humans, particularly in managing AI agents, and increase demand for SaaS products rather than eliminating them.
- Model progress is growing exponentially, but despite this, humans will find more work to do, creating a paradox where increased automation leads to more human effort, especially in managing and refining AI outputs.
- Agent-generated bug reports are far superior to human-generated ones, providing precise reproduction steps and even code suggestions, leading to a much faster and more efficient bug-fixing process.
- The CLI (Command Line Interface) era is effectively over for most professional work, as Graphical User Interfaces (GUIs) combined with AI agents offer a nicer and more efficient work surface.
- Companies will adopt a 'super agent' model, with one central AI agent for the entire company, managed by forward-deployed engineers. This top-down approach will later decentralize as AI models become more independent.
- Every doubled in size from 15 to almost 30 people (The company, Every, has doubled its team size in the last year, demonstrating growth in an AI-forward company.)
- GPT 5.5 scored 62 out of 100 (On a 'Senior Engineer Benchmark' designed to test AI's ability to rewrite vibecoded slop code from first principles, GPT 5.5 scored 62, compared to 30 for previous models and 80-90 for a human senior engineer.)
- AI agents can read 10,000 pages in a second (Highlighting the speed and capability of AI agents to process large volumes of information, enabling deeper research and analysis.)
RevBots.ai View:
- AI Sprinkler teams will struggle with 'bring your own AI' SaaS models.
- ARM-stage companies will architect around centralized AI agent orchestration.
- Tab Hoppers face disruption as CLI-era tools give way to AI-native interfaces.
- SaaS Hoarders risk tool sprawl as AI agents demand new integration patterns.
Join The RevBots ARMy
The insider daily for Autonomous Revenue Masters.