The notion of tech founders attending etiquette classes reads as a punchline, but the underlying signal is real. As code generation, summarization and other "intermediate" engineering tasks become commodities, the work that remains is heavier on judgment, communication and relationships than the technical work that came before it. The founders best positioned to lead the next decade of company-building are not the ones with the deepest hands-on coding skills — they are the ones with the strongest mental models of customers, teams and decisions made under uncertainty.
Key takeaways
- AI compresses the value of skills that are easily codified and amplifies the value of skills that aren't.
- Customer empathy, narrative ability, and conflict resolution are the new scarce resources.
- The "soft skills" label undersells what is actually being rewarded — these are leadership skills with measurable economic returns.
What is actually being compressed
The intuition that AI threatens "low-skill" work is wrong, or at least incomplete. The work being compressed first is the codifiable middle: tasks with clear inputs, clear outputs, and a wide corpus of examples to learn from. That includes a great deal of standard engineering, basic legal drafting, formulaic writing, and entry-level analysis. Above that band, the work that survives best is the work that involves judgment under ambiguity: choosing what to build, knowing when to stop, persuading a team to commit, reading a customer's unstated needs.
| Skill class | Direction of value |
|---|---|
| Boilerplate code | Down |
| System design under business constraints | Up |
| Drafting routine contracts | Down |
| Reading a counterparty in a hard negotiation | Up |
| Standard performance reviews | Down |
| Coaching a team through ambiguity | Up |
Why the change reaches founders
Founders are professional decision-makers under uncertainty. As more of the executable work shifts to AI, the share of a founder's time spent making decisions — about products, hires, capital, narrative — only grows. The founders who scale through this transition are the ones who are good at decisions and at communicating decisions, in a way that lets thousands of humans across an organization actually act on them.
The founder's job is increasingly the part of the job that the AI cannot do — and most of that part is human.
What "soft skills" actually means
The phrase is a euphemism. The capabilities being rewarded are concrete and learnable:
- Reading the room. Calibrating message and tone to the audience without over-rotating.
- Disagreement without rupture. Pushing back on a peer or board member while preserving the relationship.
- Customer empathy as method. Not love-bombing the customer, but actually inferring what they're trying to do.
- Narrative discipline. Telling the same true story consistently to investors, employees and recruits.
- Decision hygiene. Knowing which decisions are reversible (decide fast) and which aren't (decide carefully).
Why this rebalancing is good news for some founders
Many of the most successful operators in modern tech were not the strongest engineers in the room. They were the people who could attract the strongest engineers, set them on the right problem, and keep them aligned. That archetype was always valuable; AI makes it more so. The label "engineer-first" loses some of its rhetorical halo and is replaced by something closer to "builder-first" — a more capacious definition of what good founders look like.
How to invest in these skills, concretely
- Structured feedback loops with people unlike you — peers in different industries, different functions, different age cohorts.
- Deliberate writing practice — long-form, internal, weekly.
- Customer time — repeated, unfiltered, by the founder personally.
- Negotiation practice with stakes — board, vendors, top recruits.
FAQ
Is technical skill becoming irrelevant?
No. Deep technical insight remains a moat. What is becoming less central is rote technical productivity — typing speed, syntax recall, boilerplate familiarity.
Can these skills actually be taught?
Most of them, yes — through deliberate practice, repeated feedback and structured exposure. Not through one seminar.
What does this mean for hiring?
Hiring rubrics that overweight resumé pattern-matching are about to age badly. Rubrics that test judgment, written reasoning and ambiguity tolerance will perform better.
The bottom line
It is not that founders are becoming softer. It is that the part of their job that survives compression is the part that requires the human capabilities that the previous generation of tech leaders sometimes failed to take seriously. That part of the job is now the job.
