In the age of AI and approaching AGI, where models are getting as good as or better than humans at many tasks, the question becomes how to remain valuable in the economy. It’s hard to predict exactly which jobs or companies will be most impacted, but one thing feels clear: the only real option is to skillmax, continuously building skills to become as capable as possible.
The path I’m choosing for skillmaxxing, or what you could also call talent stacking, is to develop a wide range of skills and combine them to maximize your toolset.
Going Broad
I want to develop a wide range of skills in different areas. If I’m good at 5 different things, I’ll likely bring value to the economy in one way or another. Not because I’m the best at any single one, but because the combination becomes hard to replace.
AI is extremely good at isolated tasks. It’s still weaker at stitching things together into something cohesive, especially when it involves taste, tradeoffs, or ambiguous goals. Breadth is a bet on synthesis.
Here are the skills I’m focusing on:
Agentic Engineering
I really like building AI tools and services. Staying at the cutting edge here is less about knowing the latest model and more about understanding how to orchestrate systems. How do you design loops, memory, tools, and feedback so that the system actually does something useful? The leverage isn’t just in calling models, it’s in structuring problems in a way models can solve.
Product Management
Product skills will remain important for the foreseeable future. Getting really good at prioritization, understanding user pain points, and deciding what to build. AI can generate options, but it doesn’t have stakes. It doesn’t care which direction you take or what tradeoffs you make. Someone still needs to decide what matters and what is worth pursuing.
Writing and Communication
Writing and communication help with everything else. AI needs someone to tell it what to do. Clear thinking shows up as clear writing. If you can articulate problems well, you can both use AI better and coordinate humans better. This becomes a force multiplier.
Taste and Design
This one is underrated and hard to define. Taste is the ability to say “this feels right” or “this feels off” and actually be correct. As AI makes it easier to generate infinite variations, the bottleneck shifts to selection. What should exist? What should be shipped? What actually resonates? Taste compounds over time and is difficult to shortcut.
Growth and Marketing
It will always matter whether people care about what you put out in the world. Distribution is leverage. You can build something great, but if no one sees it, it doesn’t matter. Understanding channels, incentives, and human psychology becomes even more important when supply is effectively infinite.
Systems Thinking
How do all these pieces fit together? Most real-world problems are systems problems. Product affects growth, growth affects product, engineering affects both. Being able to zoom out and understand second-order effects is increasingly valuable.
Positioning
Skillmaxxing is not just about capability. Positioning matters just as much.
Two people can have similar skills and end up with completely different outcomes based on:
What problems they choose to work on
Who they surround themselves with
Where they apply their skills
Skillmaxxing without positioning can lead to a lot of competence with no leverage. The goal is not just to get better, but to get better in the right arenas.
How are you skillmaxxing?
Great read Scott.
Maybe this falls under systems thinking, but I think “Management” could be another dimension.
Do you have the skills to build and recruit a team of talented people … and then get them all working in the same direction towards a common goal?