newsfilter.io
Interview, Fireside Chat

Incubating Brain Co: Elad Gil’s AI Bet with Jared Kushner’s Affinity Partners

  • Elad Gil asserts that winning is the most effective method for building a strong company culture, citing it as superior to traditional perks like kombucha or motivational speeches.
  • Gil recalls early skepticism regarding cloud computing and SaaS, noting that enterprise leaders initially dismissed moving data to third-party clouds as insecure and "stupid."
  • Gil launched an internal placeholder project named "Branco" (later rebranded as "Brainco") which retained the name because no superior alternative was found by the team.
  • An "off the record" segment was flagged where Gil declined to discuss a personal aspiration he wished to keep confidential.
  • Jared Kushner was recruited to advise on AI adoption after Gil and Kushner compared notes on the difficulties Fortune 100 CEOs and government entities faced in implementing AI workflows.
  • The founding team of the new AI venture includes Eric Wu (Chairman), Dan Ashton, Mercia, and recently appointed CEO Clemens.
  • Gil explains that market structures vary from natural monopolies (e.g., Uber in the US) to oligopolies (e.g., payments with Stripe, PayPal, Adyen) and fragmented markets (e.g., restaurants).
  • Gil identifies four primary types of corporate moats: scale effects (lowering costs via volume), ecosystem effects (deep integration with third-party apps, like Salesforce), long-term contracts (locking in industries like pharma distribution), and network effects.
  • The new AI incubation strategy targets large-scale enterprises and governments directly, bypassing the traditional startup cycle of starting small, moving to mid-market, and then landing enterprise deals.
  • Gil's team consists of engineers typically found at foundation labs (e.g., OpenAI, Google) to build bespoke applications for specific verticals, such as healthcare, which can be resold to similar organizations.
  • Gil disputes the validity of a study claiming 95% of AI pilots fail, arguing that success rates depend on whether pilots are internal research attempts or vendor implementations and noting that technology adoption often takes a decade (citing mobile and the internet).
  • Gil predicts that the "redemption arc" for controversial founders like Parker Conrad (Zenefits/Rippling) and Palmer Luckey (Meta) involves emerging stronger and more focused after facing public scrutiny.
  • Gil warns against trusting media headlines due to the "Murray-Gell-Man effect," where experts see obvious falsehoods in familiar topics but accept unverified claims in unfamiliar ones.
  • Gil advises founders to prioritize "clarity of thought" over "passion," noting that most passions lead to dead ends and that employees require stable direction rather than the chaos founders often tolerate.
  • A $30 million Series A round for the AI venture was co-led by Gil's fund and Affinity (Jared Kushner's fund), with participation from Andrej Karpathy, Patrick Collison (Stripe), Reid Hoffman, and CEOs of Databricks, Perplexity, and Coinbase.
  • Gil confirmed that his fund will not pursue a take-private transaction or partnership with Opendoor, though he praised Eric Wu's operational capabilities.
  • Gil outlined a personal planning framework spanning 1-year, 3-year, and 10-year horizons across four spheres: personal life/family, societal impact, work/legacy goals, and global relevance.
  • Regarding AGI, Gil declined to provide a timeline, citing a lack of consensus on the definition of the term.
  • Gil estimated an 80% probability that a humanoid robot will walk on Mars before a human, reasoning that robots do not require life support systems for food or safety.
  • Gil predicted a low probability of the US government officially declaring the existence of aliens this year, citing historical precedents where such information remains undisclosed even if known.
  • Gil joked that he is certain he has encountered alien life in San Francisco, noting the city's unpredictable environment.
  • Gil suggested that if a humanoid robot were sent to Mars, it would likely be powered by an uploaded consciousness from xAI.