Insights on Building AI Teams — Heath Black, SignalFire

Key Concepts

• De-credentialization: The decreasing importance of formal education (PhDs, prestigious schools) in AI hiring.
• AI Talent Mobility: The movement of AI professionals between companies, particularly within the "aiv League."
• Retention Rates: The percentage of employees who stay at a company for a specific period, used to gauge poachability.
• Generational Differences: Variations in job-hopping behavior and risk tolerance among different generations (e.g., Gen Z vs. Gen X).
• Historical Composition: A snapshot of a company's organizational structure and key personnel at different points in time.
• Narrative in Recruiting: Crafting a compelling story about a company's mission, culture, and opportunities to attract talent, beyond just salary and equity.
• aiv League: A group of companies that have seen a massive concentration of AI talent.

Filters: De-credentialization and the Importance of Experience

• Main Point: Traditional academic credentials (PhDs, top schools) are becoming less critical for AI engineering roles. Experience and a proven body of work are now more important.
• Data:
  • In 2015, 27% of AI engineer hires were from top schools, and 16% had PhDs.
  • In 2023, those numbers dropped to 15% and 7%, respectively (a ~50% decline).
  • Only about 40% of research scientists have advanced degrees (not necessarily PhDs).
• Explanation: The shift from foundational ML research to applying models to real-world problems (ML Ops, product-focused roles) has increased the value of practical experience.
• Actionable Insights:
  • Focus on a candidate's body of work (open-source contributions, projects) rather than just their education.
  • Consider whether a PhD researcher is truly necessary for the role or if an experienced engineer would suffice.
  • Remove or soften academic requirements in job postings to broaden the talent pool.

Location: San Francisco Still Reigns, But Others are Rising

• Main Point: While San Francisco remains the dominant hub for AI talent and funding, Seattle and New York are emerging as significant players.
• Data:
  • San Francisco accounts for ~29% of all startup engineers (slightly down from 33% in 2013 but up from 2021).
  • 50% of big Tech engineers reside in the San Francisco Bay Area.
  • San Francisco accounts for 35% of all engineers in AI. Seattle makes up about 22% and New York makes up about 10%.
  • San Francisco accounts for nearly 38% of all early-stage funding into AI startups, compared to 26% of all early-stage funding in the US.
• Actionable Insights:
  • Location still matters, even in a distributed world.
  • San Francisco, Seattle, and New York are the premier locations for AI talent.
  • Monitor location and funding trends to identify emerging talent hubs.

Timing: Poachability, Generational Differences, and Historical Context

• Main Point: Timing is crucial in recruiting, both in terms of when to reach out to candidates and understanding their potential impact on the team.
• Subtopics:
  • Retention Rates: Understanding how long employees stay at target companies helps determine their "poachability."
    • Example: Anthropic has a 66% 4-year retention rate, while Perplexity hovers around 43%.
  • Generational Differences: Different generations have varying job-hopping tendencies and risk profiles.
    • Data: 27% of Gen Z left their job in 2023, more than twice the rate of Gen X.
    • Gen Z has about 2.2 jobs within four years of graduating, compared to 1.1 for Gen X.
  • Historical Composition: Knowing when people joined and left companies, and the company's context at that time, helps assess their experience and potential impact.
    • Example: Being on the New York Knicks in 1973 (championship team) is very different from being on the New York Knicks in 2018 (worst record).
• Actionable Insights:
  • Track retention rates at competitor companies to identify potential candidates.
  • Monitor changes to individuals' profiles to identify potential job seekers.
  • Study generational patterns to understand job-hopping behavior.
  • Use tools like "historical composition" to understand a candidate's experience within the context of their previous companies.

Narrative: Beyond Salary and Equity

• Main Point: In today's competitive market, salary and equity alone are not enough to attract top AI talent. Companies need to craft a compelling narrative around their mission, culture, and opportunities.
• Vonnegut's Story Shapes: Kurt Vonnegut's visualization of story shapes is used as an analogy for understanding a company's journey and communicating its triumphs and challenges.
• Data:
  • From November 2022 to November 2024, the average tech salary increased by only 1.6%.
  • AI engineers command a 5% salary premium and a 10-20% equity premium over other engineering roles.
  • In Q2 2024, 33% of people exercised their vested shares, down from 55% a couple of years earlier.
• Reasons for Decline in Equity Appeal: Concerns over valuations, the cost of liquid capital to exercise shares, and market volatility.
• Elements of a Compelling Narrative:
  • Close-knit environment with founders
  • Collaborative teams
  • Speed and lack of friction
  • Big mission
  • Growth opportunities
  • Exploding markets
  • Solving complex problems
• Actionable Insights:
  • Understand and articulate your company's unique value proposition beyond just compensation.
  • Focus on creating a positive and engaging work environment.
  • Communicate the company's mission and the impact employees can have.

Synthesis/Conclusion

Recruiting top AI talent requires a data-driven approach that goes beyond traditional methods. By understanding the changing dynamics of the talent market, including de-credentialization, location trends, timing factors, and the importance of narrative, companies can gain a competitive edge in attracting and retaining the best people. The key is to use data to filter for the right skills and experience, target the right locations, time outreach effectively, and craft a compelling story that resonates with candidates' values and aspirations. Just as data is used to build and analyze products, it should also be used to build and nurture a company's most valuable asset: its team.