MMarketing Against The Grain
← All frameworks
Leadership

High-Slope AI Talent Scorecard

Hire fast learners who care deeply, collaborate well, and adapt with AI

Difficulty
Advanced
Time to result
~months to results
Steps
5
Confidence
95%

The High-Slope AI Talent Scorecard evaluates candidates on the rate and quality of their development rather than treating their current specialization as the full measure of future value. Leaders look for evidence that a person repeatedly learns unfamiliar skills, experiments without waiting, and converts learning into useful action. That learning slope is balanced with four cultural and performance requirements: genuine care for the mission and users, obsession with improving the product and team, humility and kindness in collaboration, and clear competence at the work. The framework does not reject specialists; deep experience remains valuable for complex systems. It instead recognizes that rapidly shifting AI environments reward people who can preserve sound judgment while expanding what they can do. The intended output is a compact, adaptable team that improves its coordination every week.

Origin

Anton Osika explained the attributes he prioritizes while building Lovable's unusually small team during a period of rapid growth. Extracted from Marketing Against The Grain.

Core principles

  • 01Learning velocity matters when required skills change rapidly
  • 02Care for the mission, users, product, and team predicts sustained effort
  • 03Humility and kindness strengthen collaboration
  • 04Functional competence remains necessary even in generalist roles
  • 05Team composition is a founder's highest-leverage responsibility

How to run it

  1. 1

    Measure learning slope

    Review situations in which the candidate had to acquire an unfamiliar skill quickly. Determine how fast they moved from ignorance to useful output.

    Pro tip Ask for a timeline, actions taken, and independently verifiable result.

    Watch out Do not infer learning agility from confidence or prestigious credentials alone.

  2. 2

    Test proactive experimentation

    Look for a habit of trying tools, building artifacts, and validating ideas without waiting for exhaustive direction. Use a practical exercise when possible.

    Pro tip Give the candidate a bounded unfamiliar problem rather than a rehearsed domain question.

    Watch out Unstructured action without judgment can create costly rework.

  3. 3

    Assess depth of care

    Explore whether the candidate pays sustained attention to users, product quality, the mission, and team effectiveness. Seek evidence of behavior rather than declarations.

    Pro tip Ask what they improved after noticing a problem outside their formal remit.

    Watch out Intensity focused only on personal status can damage the team.

  4. 4

    Verify competence and judgment

    Confirm that adaptability is supported by the baseline expertise required for the role. Examine how the candidate anticipates consequences for users and systems.

    Pro tip Use work samples that include trade-offs and downstream effects.

    Watch out Generalist potential does not remove the need for role-critical expertise.

  5. 5

    Evaluate collaborative character

    Assess humility, kindness, receptiveness to correction, and the effect the person has on group performance. Reference checks can help distinguish polish from consistent behavior.

    Pro tip Ask former colleagues how the candidate behaved when wrong or under pressure.

    Watch out Do not excuse destructive behavior because a candidate appears technically exceptional.

In the wild

Staffing an AI-native product company

A rapidly growing AI startup compares two capable candidates. One has narrower but prestigious experience; the other shows repeated evidence of learning new disciplines quickly, shipping experiments, responding humbly to feedback, and improving team workflows. Using the scorecard, the company selects the second candidate while verifying the necessary technical baseline.

The company adds a contributor equipped to expand with the role and strengthen a small cross-functional team.

Common mistakes

Confusing youth with adaptability

Anton explicitly rejected the idea that he generally tries to hire teenagers; curiosity and learning behavior matter more than age.

Hiring curiosity without competence

Learning agility complements rather than replaces the ability to perform the role and understand complex consequences.

Ignoring interpersonal impact

A fast learner who lacks humility, kindness, or care can reduce the effectiveness of the entire team.

Is it for you?

Best for

It is best for small AI-native companies that need autonomous people to learn quickly and contribute across functional boundaries.

Not ideal for

It is not ideal as a substitute for mandatory deep expertise in highly regulated or technically unforgiving roles.

From the transcript

I generally bet on people that have a high slope so that you can see that they're very, very good at learning new skills and…

Anton Osika · 10:30

But most importantly, I mean how I think about building the company is to build a team of people that uh really care about what…

Anton Osika · 11:00

And people who are generally humble and nice, you want to be able to work with good at their job, of course.

Anton Osika · 11:00

From the episode

The Startup Letting 99% of People Build Apps Without Code