On the record
“However I am now feeling a substantial change in AI, specifically in the past five years, where who you know is becoming more important, and what you can do is becoming less important.”
“To assign arbitrary numbers, I would say that five years ago, success was 75% your skills and 25% who you knew. Now, it is maybe 55% skills and 45% who you know.”
“When it's difficult to independently verify someone's technical contributions, people increasingly rely on indirect signals such as reputation, personal endorsements, and who will vouch for them.”
Opinions and predictions (not established facts)
- Who you know is becoming more important than what you can do in AI.
- The trend of connections mattering more than skills will continue due to increasing resource requirements and closed-source models.
- The best strategy for building connections is to be your best and truest self, rather than intentionally trying hard to get to know people.
What changed
AI researcher Jason Wei argues in his blog post that the balance of success factors in the AI field has shifted significantly over the past five years. He posits that "who you know" is becoming more important than "what you can do," moving from an estimated 75% skills/25% connections split five years ago to a current 55% skills/45% connections split. Wei attributes this trend to three technical reasons: the commoditization of intelligence through AI agents, the dramatically increased resource requirements for cutting-edge AI research, and the move towards secretive, closed-source models which reduces observability and makes independent verification of technical contributions harder. He uses Sam Altman as an example of a well-connected individual who effectively leverages personal relationships for professional gain, citing Altman's social events for employees and recruiting purposes as beneficial for OpenAI's success.
Why it matters for builders
Wei's argument suggests that builders in the AI space may need to re-evaluate their strategies for career advancement and project success. If networking and reputation are becoming more critical than purely technical execution, it implies a need to cultivate professional relationships and personal branding alongside technical skill development. This shift could particularly impact early-career researchers or those without access to significant resources.
Practical impact
Wei's analysis highlights a potential tension between technical merit and social capital in AI. While he doesn't dismiss technical skills, his observation that indirect signals like reputation and endorsements are gaining importance due to poor observability in closed-source AI development is a critical point for builders. It suggests that demonstrating technical prowess might increasingly rely on factors beyond code or published papers, such as vouching from established figures or strong professional networks. The example of Sam Altman illustrates how intentional relationship-building, even if stemming from genuine curiosity, can yield significant professional returns.
Caveats and source limits
Jason Wei states that technical skills still matter and are crucial for career success, and he does not claim they are entirely unimportant. His observation about the shift in importance is presented as a trend and uses approximate percentages. He also notes that intentionally trying to build connections might be suboptimal, suggesting that being one's
Sources
Claim check: 5/5 supported claims - 5 evidence links - 95% avg confidence
- Over the past five years, the importance of "who you know" in AI has increased relative to "what you can do."supported - jasonwei.net
- The commoditization of intelligence via coding agents has democratized research execution, making technical execution less of a differentiator.supported - jasonwei.net
- Increased resource requirements for AI research (hundreds/thousands of GPUs, substantial data/infra) make being resource-rich a key factor for success.supported - jasonwei.net
- The shift to secretive, closed-source models in AI has led to poor observability, making it harder to attribute credit and increasing reliance on indirect signals like reputation.supported - jasonwei.net
- Sam Altman exemplifies a well-connected individual whose social interactions and relationship-building have yielded significant professional returns for OpenAI.supported - jasonwei.net
Caveats
- The author provides approximate percentages for the shift in importance.
- Access to resources depends on factors beyond technical skill, such as reputation and professional relationships.
- The author suggests Altman's connections stem from genuine curiosity and generosity rather than purely strategic intent.
- Single-source caution: verify critical details at the linked source.
Radar score 76/100 - how it was calculated
- Reliability 77: Primary on-record statement
- Freshness 100: Fresh source date
- Novelty 64: New statement by a tracked AI voice
- Technical 48: Structured technical source signals
- Developer 49: Builder relevance source signals
- Ecosystem 70: Tracked AI voice
- Confidence 98: Claims have reliable evidence
Discussion
Loading comments...