Philip Su on Career Growth to Distinguished Engineer (IC9) at Meta & OpenAI: Insights on Management, Leadership & AI

Ryan Peterman

Early Career and Growth Factors [01:02]

• Rapid Promotion at Microsoft
  • Philip Su achieved a fast promotion to Senior Staff (equivalent to Microsoft's E7).
  • This was attributed to:
    • Mentorship and Team Environment: Learning from experienced managers and teammates, especially on the Tablet PC team, which had three distinguished engineers.
    • Intense Work Ethic: Working extremely long hours, including sleeping in the office and following Da Vinci's sleep theory (biphasic sleep) to maximize coding time.
    • Three Pillars of Career Growth: Luck (right place, right time), talent (natural aptitude), and hard work (outworking others). He notes that while hard work accelerates growth, it often came at the cost of work-life balance and isn't universally recommended.
  • Experience Acceleration: Long work hours provided significantly more exposure and experience than the average junior engineer.
  • Diminishing Returns: Acknowledges that while more hours generally yield more output, the productivity per hour decreases beyond a certain point.
  • Surgeon Analogy: Compares extended work hours to surgical residencies, where long shifts rapidly accumulate experience, making one a more skilled professional.

Navigating Management and Individual Contributor (IC) Paths [06:38]

• Fluctuating Roles: Su transitioned between individual contributor (IC) and management roles at least six times throughout his career.
  • Initially, he pursued management due to a cultural belief that managers gained more respect and career advancement.
  • His preference is to be an IC, but he would step into a management role if a team was poorly managed or needed a temporary leader.
  • Remarkably, on four occasions, former managers later asked to report to him, highlighting his effective leadership style.
• Strategic Career Transitions:
  • Two-Way Door Philosophy: He consciously kept the door open for switching between IC and management. Many engineers, especially those who become managers of managers, lose their coding skills and find it difficult to transition back to an IC role at a comparable level, either due to ego or financial constraints.
  • Maintaining IC Skills: Even as a manager, he actively engaged in coding to keep his technical skills sharp.
  • Readiness for Leadership: A crucial question before becoming a manager is "Would you want to work for you?" True leadership requires inspiring people to willingly follow and contribute, rather than merely holding a title.
• Career Level Transitions (Diamond Shape of Skills)
  • Early Career (E1-E2 to E5): Easier to make lateral moves (e.g., junior dev to junior product manager) as foundational skills are still developing.
  • Mid-Career (E5 range): Most difficult transition point. Deep specialization (e.g., E5 dev to E5 product manager) is challenging unless the previous role already incorporated aspects of the new one.
  • Senior Career (E8+): Skills converge again. Highly effective senior ICs (like E8) often possess leadership qualities, technical vision, understanding of infrastructure, strategic planning, and the ability to coach others, blurring the lines with management roles. This creates a "diamond shape" for career paths, widening at senior levels where leadership and strategic thinking are paramount.

The Voluntary Demotion from E9 to E7 at Meta [17:32]

• Decision to Re-level: Philip voluntarily requested a demotion from E9 (Distinguished Engineer) to E7 (Senior Staff Engineer) when he transitioned from being the Site Director of Facebook London to an IC role at Oculus.
• Rationale: He understood he could not meet the E9 expectations as an IC in the new domain, and he preferred to be appropriately leveled and compensated relative to his actual contributions and skills in the new role.
• Ego and Influence: Accepting this significant re-leveling required humility, as he went from leading a site of 400-500 people to reporting several layers down. He had to consciously avoid using his past influence and instead support the new leadership.

Understanding Engineering Levels: E5, E6, E7, E8, E9 [20:28]

• Scope of Influence: The primary differentiator between engineering levels is the scope of a person's influence and their ability to guarantee project delivery.
  • E5/E6 (Senior Developer): Expected to reliably deliver the work of approximately 10-15 people.
  • E7 (Senior Staff Engineer): Expected to regularly impact the work of around 50 people, drive technical direction, be consulted on key changes, and personally deliver projects with a 6-month scope.
  • E8/E9 (Director/Distinguished Engineer): Influence multiplies significantly (e.g., a Director might influence 100+ people). The difference becomes more qualitative than quantitative.
    • Qualitative Distinction: At these levels, individuals are irreplaceable; their unique expertise, whether deep specialization or broad generalist capabilities, cannot simply be replicated by adding more lower-level engineers.
• Promotion Beyond Bullet Points: Promotions at E8 and E9 are less about a checklist of tasks and more about the collective judgment of peers and senior leadership.
  • Diverse Archetypes: Successful E9s come in different "shapes" (deep specialists, broad generalists), excelling for varied reasons, making a standardized promotion rubric difficult.
  • Importance of Coaching: A key to advancing to these high levels is having a great coach (a decision-maker at or above those levels) who can provide qualitative feedback and guide development.
• Team Selection for Growth: Avoid "weak teams" where you might be the best, as it limits learning opportunities. Seek environments where you are "the middle house" to observe and learn from stronger peers.

The IC9 Promotion Story: Leading the Facebook London Office [28:58]

• Building from Scratch: Philip was promoted to E9 based on his role as Site Director of Facebook London. He started the engineering portion of the office with 12 people.
  • Exponential Growth: Over four to five years, he scaled the London site to 400-500 engineers.
• Strategic Leadership: His promotion reflected his ability to manage the increasing complexity and make critical strategic decisions during this rapid growth.
  • Long-Term Planning: An example cited was his ability to forecast H-1B visa volumes 18 months in advance to accurately predict and advise on future intern and full-time headcount needs in Menlo Park. This demonstrated forward-thinking, data-driven strategic impact.
  • Talent Pyramid Management: He was the first at Meta to analyze the "pyramid of seniority" across sites, focusing on ensuring the right balance of experience levels and fostering commensurate engineering growth velocity.
• Culture Building:
  • Learning from Others: He leveraged lessons from Meta Seattle's growth (where he was the second hire) and interviewed site directors from other companies like Google London (David Singleton, later Stripe's CTO).
  • The "Landing Team" Concept: Inspired by Google, he ensured the initial 12-person landing team committed to a minimum of two years. This commitment, though not legally binding, was crucial for establishing and maintaining a strong company culture during rapid scaling, as it prevented high turnover among early cultural anchors.
  • Strategic Hiring: He personally interviewed nearly 50 volunteers, selecting 12 based not only on technical skill but also their ability to be "good culture carriers." These individuals were instrumental in building a positive and cohesive engineering environment.

Working with Industry Legends: Zuckerberg, Boz, and Carmack [36:16]

• Impact of Leaders: He was fortunate to work under highly influential figures:
  • Boz (First Manager, later CTO): Provided initial guidance.
  • Shrep (Manager, then CTO): Mentored him during his time leading the London office.
• Mark Zuckerberg & Boz's Growth Mindset: What impressed him most about Zuckerberg and Boz was their profound personal growth.
  • Despite being younger than Philip when he joined Facebook, they were remarkably open to coaching and deliberately sought to improve themselves.
  • Example: Zuckerberg consciously refined his public speaking, visibly adjusting his posture over time to better connect with audiences during AMAs.
• John Carmack's Technical Acumen: While working on Oculus, Philip had indirect exposure to John Carmack.
  • Prolific Coder: Carmack was known for his immense coding output.
  • Deep Insight: More impressively, Carmack had an extraordinary ability to dive into any codebase (even after months away) and provide precise, actionable, and unsolicited technical advice on complex issues (e.g., photo distortion, performance optimizations for Oculus 360 photos).
• Scott Renfro: A Force Multiplier:
  • Described as an "interesting shape person" and a "slam dunk" case for high-level compensation.
  • Exceptional Mentorship: Scott's peer feedback was consistently glowing, highlighting his dedication to taking time to make others great through sensitive and effective code reviews, never publicly shaming colleagues.
  • Massive Impact: He was a force multiplier for teams, capable of personally guaranteeing huge scopes of work (e.g., leading major infrastructure changes like GDPR for Meta).

Leaving Facebook and Joining OpenAI [45:35]

• Motivation to Leave Facebook:
  • Social Impact Concern: A primary driver was his deep concern about the rising income gap in America and the impending threat of "technical unemployment" due to automation and AI. He felt compelled to act on these concerns, even contemplating starting a B Corp.
  • Diminishing Personal Impact: He recognized that his unique value as a generalist engineer was less impactful in a massive, specialized company like Facebook (now Meta) with thousands of engineers, compared to when it was smaller. He felt his contributions were no longer exceptional in that environment.
• Transition to OpenAI:
  • Non-Profit Work: After leaving Facebook, he started a local non-profit called Outere, building free software for global health (e.g., malaria rapid testing), funded by the Gates Foundation.
  • Missing Technological Revolutions: Having been at Microsoft during the dot-com and mobile revolutions (which Microsoft largely missed), he did not want to miss the AI revolution, which he saw as potentially "as big or maybe bigger than the internet's impact on humanity."
  • Market Leader Strategy: His decision to join OpenAI (and only OpenAI) was based on a learned philosophy: "Join the market leader or nobody at all."
    • Reasoning: Market leaders can afford to experiment and fail because of their dominant position, taking "brave big shots." Market followers, by contrast, are forced to "fast follow" the leader, limiting their innovative freedom (e.g., Google+ trying to catch Facebook).
• OpenAI vs. Facebook Culture:
  • Talent Density: OpenAI, with ~120 engineers when he joined, had the highest "talent density" of any group of engineers he had worked with, surpassing even early Facebook.
  • Product vs. Research Focus: While OpenAI's technical capabilities were exceptional, he noted that in its early days (pre-viable competitors), the product could be "lousy" and still see adoption. He anticipates that as competition intensifies, OpenAI will increasingly need to prioritize strong product experiences over pure research to succeed, a strength he observed in Facebook.

The Power of Writing and Continuous Learning [55:38]

• Developing Writing Skills:
  • Initial Disrespect for Soft Skills: He was initially a poor writer, as he undervalued "soft skills" like English and history, viewing them as subjective compared to "hard sciences."
  • Shift in Perspective: Through rereading classics and developing a love for language, he came to respect writing as a valuable, essential skill, comparable to technical skills.
  • Key Practices: His advice for writing well: read quality literature ("to write well, you need to read well") and be willing to rewrite multiple times, discarding weaker drafts. He also actively observes clever rhetorical techniques in others' writing.
• "The Great Post Explorer" (Internal Facebook Tool): He built an internal Facebook tool to rank and expose highly commented or reacted-to posts, ensuring valuable "evergreen" content wasn't lost in the temporal feed. This stemmed from his own benefit of discovering insightful past writings.

Reflections on Career Longevity and Age [1:03:00]

• Aging and Performance in Software Engineering:
  • Continued Growth: He believes engineers can become more valuable over time, though everyone eventually reaches an asymptote where growth naturally slows.
  • Shift from "Running Serpentine" to "Jogging Straight": Younger engineers may outwork others with sheer speed, but often waste effort going in the wrong direction. Experienced engineers learn to "jog in a straight line" by developing a "societal sense" to identify and stop bad projects early, saving significant team time.
  • Value of Alignment: Wisdom also brings the understanding that a team united and moving mostly in the right direction (e.g., 98% correct) is more effective than a divided team debating the "perfect" direction (100% correct). This aligns with Amazon's "disagree and commit" principle.
  • Physical vs. Wisdom: While physical stamina for long coding hours may decline with age, strategic wisdom and the ability to steer projects efficiently compensate, leading to sustained or increased value.

Credibility as a Young Manager [1:07:00]

• Early Challenges: As a young manager, he lacked humility and was overly ambitious, admitting he wouldn't have worked for his 24-year-old self.
  • Difficult Decisions: He faced tough situations, like having to fire a 40-year-old at age 24, a task made harder by his self-awareness of his own junior status relative to the employee.
• Learning and Improvement: Despite a rough start, he was a "huge sponge for learning," constantly seeking mentors, soliciting feedback, and reading non-fiction on management best practices.
  • Acknowledged Weaknesses: He admits his biggest weakness as a manager is giving direct feedback; he tends to "pull punches" and avoid confrontation, which he has only improved to "passable" over the years.
  • Strengths: His strength lies in motivating action and inspiring teams through passion and clear vision for the work.

Generalist vs. Specialist Paths [1:10:25]

• Decision Framework:
  • Prodigies: For individuals with clear, "unnaturally unique talent" (e.g., a chess prodigy), specialization early is advisable.
  • Most People: He cautions against early specialization for the majority.
    • Exploration Needed: It's hard to know if you're truly a specialist without first exploring different areas (being a generalist).
    • "Idiot Savant" Risk: Over-specialization can lead to lacking fundamental life skills (e.g., mathematician Erdish who couldn't cut a grapefruit).
    • Irrelevance Risk: Committing too early to a niche specialty carries the danger of that specialty becoming irrelevant in a rapidly evolving tech landscape (especially with AI's rise).
• Value in Company Size:
  • Small Companies/Startups: Generalists are more valuable due to "quantization effects" – small companies often need individuals who can handle fractional roles across various domains (e.g., a PM handling comms part-time).
  • Large Companies: Specialists become more valuable as the company scales enough to warrant full-time employees for highly specific roles, making a generalist less efficient if only a small fraction of their skills are utilized.

Advice for His Younger Self [1:12:43]

• 1. Clarity of Values: "Decisions are easy when your values are clear to you." He advises spending more time reflecting on what truly matters to avoid chasing goals that don't align with long-term fulfillment.
• 2. Don't Be "The Dog That Caught The Car": He achieved his early career goal (becoming a dev manager at Microsoft) at a very young age [30]. This led to a period of depression as he lost his sense of purpose, akin to a child actor whose career ends early. He cautions against setting singular, finite career goals without considering long-term meaning.
• 3. Be Mindful of What Breaks: His extreme work hours early in his career (e.g., seeing his fiancée only for late-night dinners) could have easily broken his relationship. He advises balancing career ambition with personal well-being, recognizing that "some reeds bend and some reeds break." Achieving an E7 level at 30 vs. 38 makes little difference over a 30+ year career arc, compared to the importance of maintaining healthy relationships.
• The Nature of Advice: He acknowledges that his younger, highly focused self likely would not have accepted this advice, just as people often don't believe "money doesn't buy happiness" until they experience it themselves. One must be ready to receive advice for it to have an impact.