Elon’s Pay Package | The Brainstorm EP 105

Key Concepts

• Elon Musk's Pay Package: Discussion around Tesla's proposed compensation for Elon Musk, its ambitious targets, and its potential impact.
• OpenAI Developer Day: Announcements and implications of OpenAI's developer day, focusing on new tools and platform capabilities.
• AI as a New Operating System: The concept of AI chatbots evolving into a new layer of interaction and control for users.
• Agentic Behavior: The ability of AI systems to perform tasks autonomously and in a goal-oriented manner.
• Prediction Markets: The use of prediction markets for hedging and information elicitation, specifically regarding US-China tariffs.
• Creative Destruction: Economic theory related to innovation and its impact on growth, with a discussion on technological progress.

Tesla and Elon Musk's Pay Package

The discussion begins with a Reuters article concerning Elon Musk's upcoming pay package for Tesla, tied to the annual meeting in early November. Kathy, a guest, expresses strong disagreement with the article, calling it "ridiculous." She recalls a similar reaction to Musk's 2018 pay package, which she and her team had modeled as a potential outcome of ambitious goals. She notes that Musk achieved those goals two years earlier than expected, suggesting the package was a significant motivator, though she believes he is driven by more than just money, specifically the scaling of transformative technologies.

Key Points and Figures:

• EBITDA Goals: The proposed package includes extremely ambitious EBITDA targets. The midpoint of the schedule for EBITDA goals is $210 billion, a significant increase from the current $11 billion.
• Compound Annual Growth Rate (CAGR): Achieving the midpoint EBITDA target over seven and a half years would require a CAGR of 49%. To hit the entire pay package, the EBITDA would need to compound at 62%. Kathy states that her team has not found any companies that have accomplished such sustained growth rates over a similar period.
• Humanoid Robots (Optimus): Kathy mentions that while their models include the scaling of robo-taxis, they have "very little" factored in for humanoid robots like Optimus. She believes this is an even bigger market, and if the pay package motivates Musk to accelerate this area, it would be a "win-win for shareholders."
• Delaware Court Decision: Kathy references the Delaware court decision that voided Musk's March 2018 pay package, calling it "un-American, an assault on investor rights, and an insult to the board of directors." She advocates for the appeals court to "do the right thing" and reward Musk the 2018 package, noting he has not received any salary.

Arguments and Perspectives:

• Motivation Beyond Money: The consensus is that Musk is motivated by the challenge of scaling new technologies that transform lives and the environment, not solely financial gain.
• Achievability of Goals: While acknowledging the ambition, Kathy and Brett suggest the targets are "possible if everything goes right" based on learning curves. Brett specifically states that these goals are achievable and align with how they model the company, though he notes that unit growth can be unpredictable (e.g., due to events like COVID-19).
• Critiques of the Package:
  • "Too Much": The critique that the absolute number is too large is countered by viewing it as a "very small slice of a growing pie," where Musk is rewarded for growing the company.
  • "Too Big and Too Easy": This critique is dismissed as absurd, with the suggestion that if one believes this, they should own the company.
• Win-Win for Stakeholders: The package is framed as a "win-win-win" for shareholders, Elon Musk, and society, as hitting metrics on robo-taxis and humanoids benefits everyone.
• Employee Incentive: A key shareholder group often overlooked is the employees, who are incentivized with stock. Musk being fully incentivized alongside them is seen as a driver of "incredible wealth generation" for everyday employees.
• Protection Against Corporate Sloth: The pay package is also viewed as a mechanism to prevent "corporate sloth" in a large, rich company. Musk's proven ability to drive a culture of continuous improvement and innovation is highlighted as crucial for Tesla's future, preventing it from settling into a comfortable, less innovative state.
• Option Value on Humanoid Robots: The potential of the humanoid robot opportunity is considered significant enough to warrant extending Musk's control and incentive.

Notable Statements:

• Kathy: "I believe the Delaware court decision forcing Tesla to avoid the March 2018 vote on Elon's performance-based pay package is unamerican, an assault on investor rights, and an insult to the board of directors of one of the most stunningly successful companies in the world."
• Kathy: "I would add... led by the most productive human being on earth. and a human being who attracts incredible talent, people who want to solve the world's hardest problems."
• Brett: "It's the same here with Elon. There's like it's in Tesla's best interest to give him more control and power within the company."
• Kathy: "It's a winwin win, right? It's a win for shareholders. It's a win for Elon, but most importantly, it's a win for society because if he does hit these metrics on robo taxis, on humanoids, everyone benefits assuming you have access to this technology."
• Brett: "One of the really unique characteristics about Tesla LA as it currently exists is even in the small corners of the Tesla universe, there are clearly engineers who are doing really hard work to ship kind of amazing little updates to parts that don't like impact directly the bottom line."

OpenAI Developer Day

The discussion shifts to OpenAI's Developer Day, where a browser was not announced as some had predicted. Instead, OpenAI focused on integrating web applications into their platform.

Key Announcements and Concepts:

• Application Developer Kit (ADK): This kit allows developers to integrate their applications directly into chatbots like ChatGPT.
  • Example: A user can query ChatGPT to look at real estate, and the chatbot can then natively pull in an implementation of the Zillow app. The user can then use natural language to query this integrated Zillow functionality and instruct the agent to find specific houses.
  • Contextual Awareness: The integrated applications retain the user's entire context of previous interactions within ChatGPT.
• AI as a New Operating System/Abstraction Layer: This development is seen as further evidence of AI chatbots becoming a new operating system. Developers are provided with tools to create "agentic behavior" within chatbots, leveraging OpenAI's capabilities.
• Distribution: OpenAI boasts significant distribution with approximately 800 million weekly active users, which is expected to attract many developers to build apps for ChatGPT.
• Figma Example: Figma's stock reportedly traded up 10% after being featured on demo day, highlighting the potential lucrativeness of developing for this platform.
• Agent Creation Framework: OpenAI introduced a tool that allows developers to "wireframe" together agents.
  • Concept: Imagine creating a sequence of agents: one to categorize items (groceries vs. clothing), another to find the least expensive option, a third to check the work of the second, and a fourth to negotiate pricing.
  • Analogy: This is compared to laying out tasks for an employee or playing "manager mode" in a sports game.
• Underlying Capabilities: The ability to build these agents is enabled by OpenAI's internal work on advanced models like GPT-5, which possess agentic capabilities (e.g., knowing when to go to the web, checking relevance of information).
• Strategic Lock-in: Developing agents using OpenAI's framework makes it harder to switch to other platforms, thereby supporting their API business and making it "stickier." This is crucial in a competitive API market where churn is common due to price competition.

Arguments and Perspectives:

• Rapid Innovation: The OpenAI team is seen as rapidly delivering innovation and product advancements, not just better models but practical ways to enhance user productivity.
• Challenging Incumbents: This development is viewed as a direct challenge to Apple and Google, who currently own the operating systems. The concern is how these companies will react to platforms being built on top of their own.
• Apple's Missed Opportunity: The framework is seen as something Apple "should have" done with its "App Intents" for Siri and Apple Intelligence, but they have been slow to implement.
• Potential for Conflict with Apple/Google: There's anticipation of potential conflicts regarding app store monetization and "rent extraction" by Apple and Google. OpenAI's partnership with Apple suggests they have a strategy to navigate these guardrails, potentially leveraging legal precedents like the Epic Games lawsuit.
• Google's Response: Google is expected to develop competitive offerings, likely integrating similar agentic capabilities into Gemini and making it the default on Android devices.
• "App Store on Top of an App Store": This phrase is used to describe the new layer of application development being created by OpenAI.

Technical Terms:

• Agentic Behavior: The ability of an AI system to act autonomously and pursue goals.
• LLM Actions: Operations or computations performed by Large Language Models.
• Wireframe: A visual representation of the structure and layout of an application or system.
• Evaluation Functions: Mechanisms within an agent to assess the quality or health of its output.
• Fail Function: A mechanism for an agent to gracefully degrade or handle errors when things go wrong.

The Ledger: US-China Tariffs and Prediction Markets

The segment concludes with a discussion on prediction markets, specifically a bet on whether 100% tariffs will be in effect on China by November 1st.

Key Points and Figures:

• Market Odds: The prediction market currently sits at 13% chance of 100% tariffs being in effect by November 1st. The market has dropped 41% since its inception.
• Use Case for Prediction Markets:
  • Hedging: For businesses selling goods out of China, this market acts as a bespoke insurance contract to hedge against the risk of increased costs.
  • Information Elicitation: Prediction markets are seen as a valuable tool for surfacing information that might not otherwise enter the market, especially in complex geopolitical situations.
• Arguments for "No" (No 100% Tariffs):
  • Peace and Free Trade: A general preference for peace and free trade.
  • Back-Channel Communication: Speculation that recent tweeting/posting back and forth might have been a "sparring" match, with back-channel communications leading to a de-escalation.
  • Mutual Downside: Significant downsides for both the US and China if 100% tariffs are implemented.
  • "Taco Trade" Analogy: A reference to a previous instance where a seemingly aggressive stance was followed by a more moderate outcome.
  • Rare Earths: The analogy of blocking a river to stop water flow is used to suggest that rare earth supply chains are not as easily disrupted as perceived, with alternative sources and production capabilities existing.
• Consensus Bet: The hosts agree on the prediction of "no" for 100% tariffs, marking their first consensus bet.

Technical Terms:

• Tariffs: Taxes imposed on imported goods.
• Prediction Market: A market where participants trade contracts whose payoff depends on the outcome of future events.

Appendix: Nobel Prize in Economics and Technological Growth

The final part of the discussion touches upon the Nobel Prize in Economics awarded for research on "creative destruction" and its implications for economic growth.

Key Concepts:

• Creative Destruction: A concept from economics where innovation leads to the obsolescence of old technologies and industries, paving the way for new ones.
• Technological Progress and Productivity: The consensus is that technological advancement is essential for driving productivity and human uplift.
• Disagreement on Technological Impact: The core disagreement among economists lies in whether we are still creating technologies that are meaningfully impactful.
  • Diminishing Returns: One view is that technologies are randomly discovered, but each marginal technology is less impactful than the last, leading to a slowdown in growth.
  • Combinatorial Innovation: Another view is that technologies are randomly impactful, but they can be combined to create new, even more impactful technologies, leading to exponential growth.
• OpenAI's Role: The launch of an "app store on top of an app store" by OpenAI is seen as an example of technologies combining to yield more than the sum of their parts, supporting the idea of accelerating growth.

Arguments and Perspectives:

• Economics as a Challenging Science: Economics is described as a challenging field where fancy math roughly maps reality but is not always predictive.
• Technology as Puzzle Pieces: The idea that technologies are like puzzle pieces that combine to yield greater results is considered intrinsically accurate.

Conclusion:

The episode covers two major technological and business developments: Tesla's ambitious pay package for Elon Musk, framed as a driver of innovation and societal benefit, and OpenAI's Developer Day, which signals a shift towards AI chatbots becoming a new platform and operating system. The discussion also touches on the use of prediction markets for hedging geopolitical risks and the ongoing debate about the nature of technological progress and its impact on economic growth. The overarching theme is the rapid pace of innovation and its potential to reshape industries and society.