"Not Even Wrong" Podcast
Investing in fundamentally new concepts and engineering practices with large impact.
Climate problem is bad problem because it is not clear what it actually is. Similar problems are pandemics, terrorism or equity. Bureaucrats are using these narratives to propel their own agenda and accelerate mission creep. Consequence is high government spending and societal scleroses. A good problem is a problem that 1/2 can be formulated and 2/2 measured such as decarbonization of transport or home energy. Solutions are 1/2 bottom up and 2/2 entrepreneurial (people with compulston to do something).
DAO. New way to organize work. Need to solve for three primitives of human behavior .1/3 complacency = building bad habits. 2/3 defying truth = not admitting mistakes and 3/3 Status seeking = risk of accentuated firehouse effect. GitHub solves for 2. DAO can help. But needs to emerge organically, cannot be imposed by things like lockdown.
Reaction to Marc Andreessen comments on remote work being a trend. Opposed to Elon Musk saying “come back to office.” Company = Popperian problem solving to build value with products and services. Physical presence required to overcome human behavioral primitives such as complacency, desire for status and confused relationship with truth. Problem with Andreessen argument is that remote work was coerced by authority. Andreessen law = when lots of smart people choose to work on idea it will eventually happen. Inverse Andreessen law = When ideas are imposed by authority they won’t happen.
“If much of the world is mystery, the limits are not.” Cormac McCarthy in “Blood Meridian.” Musk involved in politics is ok. Society is predictable at limits, both left and right. Nothing to loose or gain. But the larger part, the middle, eventually follows common sense. That’s why democracy works. Biggest risk to Tesla is US government recklessly spending and driving up inflation. More balance in DC can stop that. Good for economy and Tesla.
Why DAOs are exciting. Iterating around new ways to organize people with mission and purpose. Ultimately impact of Tech and software is how society morphs around the bits. Blockchain = dynamically validated decentralized database. New behaviors evolve around this particular feature.
Where do we stand? 1/2 China risk high due to Maoist resurrection. 2/2 Fed forced to increase terminal rate due to persistently tight labor market. Remote work and lax labor enforcement reduce productivity, driving up real wages and therefore fueling inflation. Stay put. It's time to build.
Discussing Andrej Karpathy appearance on Lex Fridman. (Human + Skill + Obsession ) ** Rate of Iteration = Wealth Creation. Eventually you will get to be the best at something and do it for the rest of the world. Zero to One in action. Wealth is created by humans for humans.
How will scarce money affect innovation? Price of credit in Bitcoin world? Zero? If you hoard you might as well lend. Marginal cost of lending = opportunity cost of consumption. If you don’t consume….you might as well lend. New field of research = mechanism design for crypto and how will people (and bots) behave in aggregate? So far only micro research. How about Macro Crypto? Investment opportunities in protocol design for social networks. Twitter, GitHub, LinkedIn.
ZKP workshop at Berkley. Internet of identity morphing into single word = Blockchain. Network of transactions with implicit verification. ZKP = prove validity of message without revealing message. Interesting area because many university departments involved. Mathematics, complexity theory, Computer Science (algorithms, programming languages, ISA, hardware) even EE is called upon (new substrates to deal with special gate architecture). Compute is evolving from mathematical expressions to circuits and networks of related mathematical expressions. Goal is to build network with trust less, scalable verification of identity of actors and messages.
Stack thinking. Segment the world and find solutions to problems. Example Compute Stack, Biotech Stack. Fosters wealth creation. Biotech stack is opening up. Creating new avenues to ask questions. Solve problems. Create products and services that people want = wealth creation = freedom. Enlightenment is an example of opening up political stack.
Our problem with snake oil capitalism. Participating in financial markets is a privilege. Needs to be protected. If you’re not smart enough, innovative enough, creative enough. hard working and humble enough to build products and services people want to buy at lower cost, then stay away. Wealth creation is not a game of chicken where getting out on time is the goal. Wealth creation is a long term and continuously positive vector. Snake oil companies are perpetually adjustable arguments. Examples: Uber, Wework, Lucid, Web3, QuantumScape.
Tesla price decrease in China is good thing. 1/2 Relief for macro due to lower commodity prices. 2/2 Tesla maximizing volume growth is right strategy to grow installed base. Higher gm even if ASP declines. Eventually sales growth will outpace volume growth.
Tesla EPS. 1/3 margins strong. 2/3 major macro risk. 3/3 competitive situation good. Discussing Bay Learn 2022. 1/3 Language models underpinning all AI. Robots can be trained with language models. 2/3 reasoning , Reinforcement Learning , build models with learning on little data. 3/3 academics inert and not genuine about what they do. Example most academics work on what Tesla is working on, robotics, camera depth and vision to replace LiDAR. Only place where Google is ahead of Tesla is in deploying language models. Tesla can leap ahead since they're tackling more interesting problems which will create new, more meaningful innovation stacks.
Book discussion “In The Distance” by Hernan Diaz. 1/3 Archetypical American narrative. Touches on Moby Dick, Jack London, Steinbeck. Kerouac. Cormack McCarthy. Explore what life can take and give when exposed to nature. There is such a thing as being, it's when you interact with nature. Can do mentality. 2/3 Purpose is a muscle that can be trained. Survival is the ultimate purpose. 3/3 Reality happens in your head and is when you interact with nature. Humans are universal problem solvers. Problem solving is a muscle, needs training like every other muscle.
RISC V summit at Andes Tech in SJ. 1/3 New paradigm of compute driven by ML software development. Software 2.0. 2/3 Data movement, dot product. 3/3 RISC V modular, flexible, lower power budget and higher performance. Strong in embedded and penetrating data center. Disruption.
The way we see the world . 1/3 Fed squeezing out wealth to bring people back to common sense. Namely government. Good thing. 2/3 Fundamentals of our businesses are strong. 3/3 Innovation in AI got boost by language models and advances in computer vision . Lots of potential in wealth creation. In volatile times stick to principles.
Discussing Deep Mind paper in Nature “AlphaTensor”. Apply AlphaGoZero to developing algorithms for faster matrix multiplication. Solving problems is a game. Game is exploration of counterfactuals. Humans love playing games because that is our nature. To be intelligent. AlphaTensor improves math which helps improve AlphaTensor. Escher recursion.
Reaction to Berkeley seminar. Physics inspired machine learning. Computational Optimization of Electric Fields for Better Catalysis Design. Protein Folding and function prediction. Potential energy surfaces. Using ML methods to develop functions describing behavior of particles. Science is often caught in pointless posture and sometimes even trapped in vanity. National grants and involvement by government not helpful. CompSci and Science don’t commute. You can go from compSci to Science but not reverse. Culture of open source.
Questions for Tesla on AI Day. 1/3 Self play and unsupervised RL. 2/3 Tech credit. What if you’re too early? 3/3 Zero to One. How to hire for Zero to One type projects and create culture?
Tesla AI Day reaction. 1/3 Humanoid Robot good. Learn from FSD, transfer and vice versa. 2/3 Innovation across the compute stack. Wealth creation happens across the stack because it accelerates iteration. 3/3 Flywheel of talent acquisition and culture creation.
Bitcoin, the gift that keeps on giving. Caitlin Long on Bitcoin Standard talks about MoIP, money over IP. Musk + Twitter= MoIP.
Who robots don't get cancer? What if robots repaired themselves? Michael Levin on where does intelligence stop and physics start? What is “just physics?” All organisms have some sort of will. Software 2.0 is a path towards flexible adaptive systems. Intelligence = solve problems with different inputs.
Book discussion “Bitcoin Standard” by Saifedean Ammous, 1/4 Bitcoin solves for hard money. 2/4 Economics is about time preference 3/4 hard money is a technology to allocate labor, i.e answers question: what should I do next? 4/4 No such thing as Blockchain technology. Bitcoin is solving for hard money, that’s it. Blockchain technology, like machine learning is Cargo Cult Culture.
Market collapse. US treasury yield up. UK guilds downs, BOE intervenes. ECB next? FED? Hand is forced. Further conversation on “Bitcoin Standard”. Rarity is function of time preference and knowledge. We can’t predict knowledge and thus can’t make predictions about scarcity. Bitcoin has absolute scarcity, massive impact on behavior. Hard money impacts behavior, time preference knowledge creation.
Stanford Graph Learning Workshop 2022. Data Science is incremental. GNN’s are innovation but incremental. Predictive models for internet companies like Amazon, Spotify or Facebook are incremental. Not much value added. Pharma and AI interesting. GNN can make a difference but not Zero to One. Reinforcement Learning more interesting for Life Sciences. Change in Compute Stack. Incumbents like Nvidia trying to fit new models like GNN into matrix to use GPU. Not interesting. Intel even worse. New compute stack around Data Flow versus Von Neumann very interesting. Main source of Wealth creation.
Jure Leskovec, Ass. Prof. Stanford
Book discussion “Bitcoin Standard” by Saifedean Ammous. Part 1. Importance of hard money. Hard money fosters Zero to One Investing. Our thoughts: Zero to One Investing is a Zero to One innovation ironically emerged after Bernanke original sin in 2008. Hard money and reckless Fiat both eventually lead to same result. Problem is, latter causes lots of collateral damage. Hard money is better source of prosperity. Our thoughts: US shifting to hard money regime driven by competition with Bitcoin.
State of the Union. Where do we stand? Performance bad, but good to know why. FED competes with Bitcoin towards hard money economy. Imagine US East based on current system of Northeastern elite running Fiat economy versus US West running on Bitcoin. How will that look 70 years from now? US has been fighting intellectual civil war between collectivism and freedom in past few years.. Freedom won. US will adopt hard money and freedom. Massive change in compute stack creating wealth. USD beneficiary. EUR to 0.5 USD. China can compete but will leadership open up? Rest of world getting poorer relative to USD. Freedom and change in compute stack main sources of wealth creation. Accelerating. Higher returns.
Book discussion “Sea State” by Tabitha Lasley. Memoir. Memoirs reflect through own lens. Go to new environment (Aberdeen). Rewire. Culture affects, makes you change. Market sell off due to Fed determined to stifle inflation. Good thing. Volatility is feature not bug. Leverage.
Big opportunity in Crispr based biotech. 1/2 Lower levels of abstraction in life sciences. Crispr allows view and manipulation down the stack of bio. 2/2 AI is better for life sciences. Deep Neural Nets and gradient learning are better ways to represent complex systems. Bio is one of them. Webinar on Sate of Biotech. https://webinars.liebertpub.com/e/the-state-of-biotech-2022/portal
What is a crisis anyway? Institutions like the Fed and its Northeastern cronies use “crisis” as an excuse to drive their agenda. Volatility is a feature not a bug of wealth creation. Rule based policy. Bitcoin.
Two big changes in compute paradigms. 1/2 End of Moore’s Law requires rethinking stack up and down. 2/2 Data centric compute. Inflection point for Wealth Creation. Faster iteration, lower cost of errors and lower cost of error correction.
AI Summit Part IV. Autonomous driving conversation. Deep Mind approach is “Northstar” for planning and control. More advances in animation needed. RL + Monte Carlo search tree. Better RL models. Who is working on this?
Episode September 14, 2022 III
Three questions for Tesla AI day inspired by AI summit. 1/3 merge of edge with data center through software defined hardware. Should Von Neumann architecture be rethought? 2/3 neuromorphic computing? 3/3 Deep Mind approach to control?
Episode September 14, 2022 II
AI summit take aways for portfolio. 1/3 Tesla has edge thanks to OS. 2/3 software defined HW driving composability and disaggregation. 3/3 Deep mind not involved in FSD due to Waymo conflict. Hostage.
AI Summit Santa Clara 2022. Impressions. What if John Von Neumann was around today? He’d say, “why are you still using my architecture?” Three levels of constraint in todays AI hardware. 1/3 compute 2/3 CPU-memory interconnect which is Von Neumann wall and 3/3 I/O wall. CLX protocol attaching Von Neumann wall. Software defined hardware solving for performance and power. Composability and disaggregation. Edge and cloud merging. Brainlike compute or neuromorphic 1/4 sparcity 2/4 low precision 3/4 quantization 4/4 model based like energy models and/or routing to use only specific nodes in model. Trillions of parameters but only use the most efficient. Edge training good for cars.
Economics seminar at Stanford. Climate Finance. 1/3 Climate X is a bad problem formulation. Call it decarbonization. 2/3 ECB buying stocks was not market neutral. But lowering rates moves private capital down risk curve. 3/3 Good thing about low rates is change through disruptive startups. Does hard money offer same upside for change?
Supply side economics turned upside down. Today it means very different things than in the 80s under Reagan. Today government is supposed to deliver desired products and services. In fact it’s crowning out private enterprise for the sake of political goals. For example helping US manufacturers develop battery production is a subsidy for unions and legacy industry. It’s a phrase since Tesla already successfully does all that in the US. Traditional supply side economics from the Reagan area is the exact opposite. It fosters private enterprise by getting government out of the way. The latter worked. The former doesn’t.
Changing paradigms in compute are inflection points where wealth creation happens. New and much larger markets are enabled through new forms of compute. Compute is best way to accelerate iteration which is key for disruption and wealth creation.
Why are SaaS companies failing? Problem is cost structure. Cloud based business models mostly benefit cloud providers. SaaS often no pricing power. A company is defined by its cost structure and its future is defined by how this cost structure will change.
Market design at the core of computational game theory which is in short economics + computation. Approach market design as an information problem. What do we know about the actors in the system and what should we know to get optimal outcomes? Who gets to live in the nice part of town? How do we decide? Who is a good player? What is good?
Research topics. 1/2 Blockchain design with focus on robustness, safety and preventing large scale attacks. 2/2 Algorithm design with focus on self supervised reinforcement learning. In both domains we are asking the same question. How can we nudge systems of independent self interested actors to converge towards desired outcomes?
Book discussion Hernan Diaz “Trust”. 1/3 winners write history. Be aware of narratives. 2/3 Wall Street wizardry is overrated and often wrong. 3/3 Anybody can make it on Walk Street. Skill is available to all. Question is. how do you turn it into returns?
Feeling heavy. Returns. Family. Student loan forgiveness is terrible signal. Sometimes I feel like we are the suckers at the poker table. Piling up debt is stupid. But government incentivizes debt through forgiveness. Similar problem with housing and mortgages.
Reaction to Neel Kashkari on Odd Lots Pod. “Don’t fight the Fed”. But how can the Fed fight fundamentals? Rules based monetary policy would be much better for prosperity as measured in units of freedom. Unfortunately Neel and his colleagues don’t think that way.
In “Trust” Hernan Diaz describes the financial market infrastructure and culture as deficient and fraud/bubble prone. We agree. Our way to navigate markets is to work within a wide range of possible events by managing leverage and applying an appropriate principal stack. Markets are the best way we know to allocate labor to projects. That’s why markets fundamentally provide a backbone for wealth creation.
MuZero is one of the most important innovations in AI. It asks a different question. How do complex systems behave when transforming from state to state. Self supervised reinforcement learning within Monte Carlo search trees. We’re particularly interested in applications of protein folding and material science, which is what MuZero seems to be targeting. We hope DeepMind collaborates with Tesla soon.
Silicon Valley wealth creation. Take big technological bets with fundamentally new concepts and engineering practices. Optimize for modular and iterative scalability. Example Tesla Dojo. Amazon AWS. Google algorithmic internet classifier. Deep Mind MuZero.
What should society optimize for? Disruption. Iteration, low cost of error and low cost of error correction. This applies to all levels of society, individual, family, state and globe. Instead of static hard to define goals like fairness or equality we prefer dynamic goals like optimizing for disruption.
Is there convergence in business? Does entrepreneurial activity converge to certain size and/or value? Is there maximum wealth that can be created? No. Human endeavors are driven be evolution of explanatory knowledge which is creative and whose path is neither constrained nor predictable. Wealth creation applies to “Beginning of Infinity.”
Musk vs. Twitter scenario. Judge could rule that this should be tried by the SEC since investor miss information is at the core of problem. SEC is supposed to protect investors. Twitter is not clear about how they report user numbers and possibility of overstating potential revenue is high. Regulators have to reassess their oversight of social media. Ironic outcomes are most likely.
Would Max Weber have invested in Tesla? Yes. Because Weber coined the concept and importance of Charisma in leaders. Charisma goes both ways. Who has it and who is influenced by it? The US is more receptive to charismatic leaders than Europe. US had good experiences with such leaders while Europe didn’t.
Web3 opportunity is exciting because it offers new ways to organize markets and society. Many branches of academia involved. What do we want? (Philosophy). How to do it? (Algorithm design, economics, computer science, math). Market design, hardware software layers, mathematical solutions to more efficient crypto layers. Whenever there is so much changing, opportunities open up.
Book discussion Benjamín Labatut “ When We Cease to Understand the World”. Dr. Faust, Zauberlehrling, Popper. Science is an art, a dangerous art. How far do we go? Where do we stop? How? Labatut asks questions Goethe epitomized in his poems. What Science to invest in? What do follow and what is too dangerous to pursue. To answer these fundamental questions we focus on natural Sharpe Ratio. Foster iteration, lower cost of errors and lower cost of error correction.
Question to Tesla shareholder meeting. Will Tesla apply self supervised reinforcement learning to FSD problem? Step needed is animation rooted in real world. What if Demis Hassabis worked with Elon Musk? What if Thomas Edison had worked with Ford, or Disney with HP?
Popper is a much better tool for investing than Bayes or induction. Explanation versus Induction in investing. We use Popperian approach through explanations for problems, not Induction or Bayesianism. New information is used to test explanations. Explanations are formulated through Principal Stack. Wall Street is caught in fallacy of Induction, ie. look at past observations and extrapolate into future.
Find your own thing. It’s a process, a feeling. Only way to succeed. Success = get paid for what you’d do for free. Empathy. Find way to be useful to others. Empathy is a muscle. Train and get better.
Book discussion Sally Rooney “Normal People”. 1/3 Adolescents deal with sexuality and we often neglect that. We shouldn’t. 2/3 Relationships are an accumulation of subtle communication, both verbal and non-verbal. 3/3 Social classes are defined by the people who want them, usually by those who think they are superior. It’s a completely arbitrary concept but it becomes real when enough people believe in it. Like religion. Dangerous.
Refute nuclear power. It’s not just a bad solution to decarbonization problem. It causes many more bad problems. Nuclear power is the inverse of a Popperian solution. Instead of creating new, better problems, it create worse problems such as high risk, waste and crowding out of new innovation. Nuclear power needs to be priced properly. Don’t believe it? Ask a private insurer to insure a nuke with no government backstop!
Is Tesla wrong with 4680? We are in the business of picking innovation stacks. We can’t predict the future because it’s hard to predict creative knowledge creation. But we can evaluate the validity of innovation stacks. Tesla is building the right innovation stack for decarbonizing the economy.
Book discussion “The Guide” by Peter Heller. 1/2 Civil liberties in the wake of force major. Governments have monopoly on violence and cannot be trusted. Similar to “The Unforgiven”. 2/2 SynBio is great but brings moral and ethical risk. Conversation is needed.
Workflow of SynBio. 1/3 Genome space thorough sequencing. 2/3 Protein space through Alpha Fold and 3/3 design circuits with living organisms. Example James Collins designs bacteria to read to toxic substances. Allogenic CarT is most important application of SynBio currently.
Reaction episode to Demis Hassabis conversation with Lex Fridman. 1/3 Self Supervised Reinforcement Learning for AI and AI to represent, understand and evolve physical systems. 2/3 AI best for life sciences like Alpha Fold. 3/3 Alpha Fold transforming gene space to protein space.
Book discussion “Wolf Hall” by Hilary Mantel. 1/3 King Henry VIII separates from Rome church in order to justify his divorce to Catherine of Aragon. Creates basis for constitutional monarchy, checks and balances and foundation of research driven industrial revolution which propels Anglo Saxon culture to top of world. Scratch your own itch. 2/3 Thomas Cromwell is COO of this operation. Pragmatic merchant runs English revolution. 3/3 Observation is theory laden. I see in this story parallels of today (new authority with blockchain). Others might not see that in Wolf Hall.
The hard things about the hard things. When it’s up or out say no to status quo. Either make the step up or go elsewhere and reboot.
Three big technologies shaping the future and our portfolio. 1/3 Intelligent Electric Cars, 2/3 Genomics and Crispr 3/3 Blockchain technology in particular solutions to Byzantine Generals problem.
Four questions to Onur Mutlu. 1/4 Design space for intelligent electric car, in particular power, reliability, latency and security. 2/4 Integration (Tesla) versus horizontal approach (Nio). 3/4 Application to compute to small (cell) or large ( car). 4/4 Dall.E question. Self driving solved by making car that sees and learns and navigates real world or in virtual animated world like Alpha Go Zero through self play and self supervised reinforcement learning.
Failure and rejection are part of Popperian process of discovery. Level, personality, capacity and Will. Iterate and learn, minimize error and cost of error correction. Learn from failure and success. Compound interest is a function of iteration.
Inflation is not a measure of money printing, it’s a measure of freedom or the lack thereof. Freedom and free markets are a moral imperative. Free markets allow everybody to participate voluntarily in the process of wealth creation. Markets like housing, health and education are not free and the key drivers of inflation in the US.
Regulatory bodies are often used by incumbents to prevent new market entry. Example NHTSA claims Tesla software less safe than others. Business, Market and Capitalism are related but not same. Markets need freedom. Freedom begets wealth creation and vice versa.
The problem with Biotech is there is no Elon Musk, Jeff Bezos, Walt Disney or Thomas Edison. Academics call themselves serial entrepreneurs when in fact they lack focus and dedication. See G. Church or J. Doudna and many others. In order to drive the field towards epochal transformative power with massive wealth generation we need less of G. Church and more of Elon Musk in Biotech.
Caribou Bio releases data on Car-T allogenic cancer therapy. Some patients have resurgence. Safety good. Now Increasing dose. Biotech is good investment because increasing natural sharpe ratio = higher rate of iteration, lower error rate and lower cost of error correction than legacy medical approaches. High X and Y. Market treats these stocks like everything else Bayesian.
Discussing Peter Thiel's thesis that wealth is created by monopolies. He divides wealth creation into x= you add lots of value to society and y= you’re able to capture a big part of that value. Tesla is such a company. X is big and Y is large. Same with SpaceX. Creating legal monopolies by rapid iteration. Building innovation stack. Hard to compete with.
Web3 is the thing which allows for a decentralized way of deciding what is and what isn’t (=solving Byzantine Generals Problem). It’s also the thing which allows for decentralized development of software that builds on itself (=composability). The biggest opportunity is the ability to organize society in new ways similar to Renaissance which introduced law and accountability to previously absolutist societies. More options, more possible outcomes.
How should we govern ourselves. Monarchy? Jubilee? No. Family run organizations are efficient but lack scalability and not enough iteration. My favorite is George Washington, a would be King that chose to build an enduring governance structure with iteration and error correction embedded. Constitutional Monarchy is like hybrid electric cars. Not as good as pure play.
Book discussion “The Plot” by Jean Hanff Korelitz. Both mystery novel and comic characterization of literary scene and writers. Is a story unique? Does it belong to somebody? Stories are like ideas, they emerge from people and belong to people who tell them, not live them. Stories can only be told, not owned. Writing is like any artistic endeavor, hard to succeed but worth it.
Travel blog part 3. Web3 protocol = many to many + decentralized (no authority). New way of organizing society. New way of living. Old professional services like lawyer not important. New functions such as token design, Tokenomics, protocol design.
Travel Blog part 2. Visit Tesla Giga and Gruenheide. Tesla Giga very large. About 1/3 done. Lots of construction, gravel, roadwork etc. people mostly foreign. Either English speaking or Polish. Positive feeling. Gruenheide very high end tourist spot with lakes around. Understanding opposition by locals. Not much to gain from industry.
Travel blog Berlin part 1. Every city has its soul which is the soul of all the people before. Berlin certainly has its share of souls. Tesla chose to be here for two key reasons. 1/2 Germany important for EU EV market. 2/2 Berlin no legacy automotive industry.
Chiara Marletto “Science of Can and Can’t” Chapter 6 Work and Heat. are Energy and information the same thing. Physics is the discipline that tries to explain how the world works. Can it be described in pure information theoretic ways. 2nd law is about stored information and reversibility. Can Ising Model help explain physics more generally? Is 2nd law of thermodynamics subjective to observer? Are entanglement and superpositions reasons for irreversibility of energy transfers, i.e. 2nd law?
What would Karl Popper think about Elon Musk? He’d be happy. Book discussion “The Open Society and its Enemies” by Karl Popper. Create knowledge that makes society a little bit better off. Better in practice means increase freedom. Musk is laser focused on driving freedom by increasing flexibility and options for customers, countries and the whole globe. Musk and Popper are buddies in spirit.
The fight for freedom is “The Fight”. We live in a period of time when this Fight is prevalent. Freedom begets wealth and wealth begets freedom.
TSLA key drivers and risk. Technology: 1/4 Battery good. 2/4 FSD monetizable but needs better (self supervised reinforcement learning) 3/4 Chip business good 4/4 Insurance Good. Person and legal risk. 1/2 more attacks on TSLA because more success. Tesla building bulldog legal team good.2/2 Musk risk high needs better Tesla bench. Market risk 1/3 China biggest risk due to lockdowns. 2/3 US interest rates not big risk due to plateauing. 3/3 Leverage in TSLA high due to Modigliani-Miller theorem necessary evil to live with.
The environment is a bad problem. Because it’s not clear and therefore can be used by anybody for their purpose such as keeping competitors away, raising prices and corrupting politicians. We focus on the problem of decarbonizing energy and transport. Digital discovery drives advances in material science. That’s a good problem.
“In Science read from right to left” Karl Popper. Same applies to investing. What is the thing that creates wealth? It’s a company that gives fellow humans something they desire at lower cost and with more value. That’s what we invest in. Tesla. Digitizing life sciences. Digitizing Material
Is Elon Musk a contemporary Prince Myshkin from Dostoyevski’s “The Idiot”? The novel examines the consequences of placing a positively good and beautiful individual at the centre of conflicts, desires, passions and egoism of hose with whom he becomes involved. His personality possesses the peculiar capacity to relativize everything that disunifies people and imparts a false seriousness to life.
Book discussion “The Open Society and its Enemies” Karl Popper. Part 1. 1/2 Going back to Greek thinking like Plato because they first transitioned from tribal to open society. 2/2 Open Society = you can change from within by piecemeal adjustment w/o violence. Goal is to make everybody slightly less unhappy or develop better solutions to problems. Innovation stack in society building. Same concept like in business. Key is small steps for change, test, error correction, repeat.
Business can be viewed as reinforcement learning problem. Reinforcement = iteration. Any institution can be viewed as iteration problem. Compound interest is function of iteration, which is driven by speed of iteration, cost of failure and cost of error correction.
Market turmoil. China risk. Leverage. People borrow in USD to buy assets. Carry trade reversal. But fundamentals are even better.
The purpose of science is to proliferate. Shaping norms so that more resources go towards science. Help develop weapons and other means of warfare, in particular communication (Turing, QuantumCrypto). Why is it proliferating more in totalitarian regimes? Because free societies make it more difficult to optimize for warfare. Risk is auto thief effect. How to make sure they don't kill us? Instead of optimizing for warfare Western science optimizes for status. Solution. Get away from status and optimize for problem solving. Create better incentives with tokenomics thourgh DeSci. Web3.
Web3 is when you open your computer you start making money. Tokenomics. MOIT = monetize ownership identity and trust. Twitter must apply those concepts to create an open society with freedom at its core. Freedom = act without authority under constraints of Tokenomics, ie rules of game are known to all.
Motivation to read Karl Popper “Open Society & its Enemies”. 1/2 How do organize society to optimize for progress and freedom. 2/2 Musk buying Twitter is motivated by same arguments as Popper uses to defuse extreme forces from left and right as well as petty bureaucrats. All these vectors are infringing on liberties and attempting to rewind the clock of progress, freedom and the achievements of enlightenment.
Drawdown due to market sell off. 1/2 Nuclear risk in Ukraine. 2/2 Higher US rates create stress. Book discussion “The Idiot” Elif Batuman. Language. Defining stuff is power. Defining stuff is freedom and transcends hierarchies. Math is abstraction layer that turned into power. Russian literature is something your supposed to like even though you don’t. Elif asks lots of questions but doesn’t offer answers. Makes for frustrating and often boring narrative. Journalists ask questions. Hedge fund managers ask questions and get paid to find answers.
Lecture “Intro Computer Architecture” by Onur Mutlu Part 2. Design principles: Look up, down, forward and backward. We pay particular attention to the “down” which is predict where technology is going. Focus on compute because compute is key to wealth generation. Paradigm change with Quantum computing.
Why we’re studying advances in Quantum Machine Learning. Accelerate BLAST (basic linear algebra subroutines). Accelerate data analysis and reduce memory bandwidth thanks to qRAM. Parallelism of quantum will reduce energy requirement per unit of analysis by orders of magnitude. One application is animation to enable self supervised reinforcement learning for cars and robots. Exactly what Tesla wants.
Lecture Computer Architecture 1, Part 1 by Onur Mutlu. Computation is driver of future products and services. Therefore understanding where computer architecture (CA) is today offers sneak peak into future. CA = design across stack ie. algorithm, program, microarchitecture, instruction set. Logic, device, electron. Today we solve problems by manipulating electrons. Tomorrow maybe photons, atoms etc. Feynman “plenty of room at bottom”.
It’s the product stupid! Will Musk-Twitter succeed? Will Musk succeed with Tesla. All these questions boil down to one thing: The Product.All questions about business are questions about product. Entrepreneurs are people who create products and services. The test is wether the product is something people want at a cost that allows for reinvestment. Bacteria goes to sunlight. Products are sunlight.
TSLA selling off. 1/2 Musk selling TSLA shares. 2/2 Bezos tweets that China might have leverage over Twitter through Tesla. This warrants fear but is not such a big concern. Musk wants to save free speech in US not make it happen in China. Proliferating trust, human identity and (maybe) ownership on Twitter. Mutual benefit between Musk and China. Twitter can surface truth about China which is different than Western narrative.
Web3 and publishing. Ownership, identity, property rights and monetization. Improve on current system of gate keepers and arbiters of prestige. Introduce composability. Musk needs to work on those problems with TWTR or elsewhere. Free speech = no authority decides, network decides. Smart contracts work with monetary incentive. Monetary = resources to increase freedom.
Biotech debacle. Down 50% ytd. Are these stocks fluff or real? They're real because the companies are building an innovation stack in solving the problem of digitizing life sciences. Transform the field from bespoke alchemy to engineering driven solutions. No threat of competition by incumbents like NFLX because startups are doing something radically different. Goal 5x in 5 years.
Book discussion “Grapes of Wrath” by John Steinbeck. 1/3 Great economic elaboration. Literature is best history and economics. 2/3 Defining novel of 20th century US. Technology displacing human labor. Strife. Avoid strife by increasing iteration in education. 3/3 Migration is iteration and brings fear and hope. People change when they migrate.
TSLA earrings discussion. Q1 2022. Execution. Growth through delivering fundamentally new concepts and engineering practices. 1/2 Roberto taxi. 2/2 Optimus. Both are 1 type events. At least 50% growth for 5-10 years. Optimus is the tractor of 21st century (Grapes of Wrath). High dexterity and hand eye coordination work will be replaced by machines.
NFLX collapse. NFLX is not disruptor and TSLA is. Impressions from “Grapes of Wrath”. Book of the 20th century America. 1/2 switch from brutal capitalism to more human, government induced social capitalism. Main reason for economic displacement is delta between speed of education adjustment to technological change. 2/2 migration is hard and hopeful.
Probability is sorcery. Not useful for analysis of real things. Only what is and what isn’t matters. Channel life decisions towards 0/1 decisions. Is or isn’t, not 50/50 stuff. Even gut feeling can have 0/1 decision power. High conviction means 0/1 decisions.
Twitter take over by Musk. Instill purpose. Free speech means no authority decides. Like Bitcoin. No throat to choke. Bring in bench to execute against new purpose.
Solution to China problem is key for near term success of TSLA. Find way to produce cars in suppressive lockdown. Bubble solution. Long term China higher risk because leadership shows hand. Maximizing for power not economic prosperity. Big change in post communist area since 1990. Not maximizing for economic prosperity but other goals. There is no other thing you can maximize for long term.
Two drivers of recent correction in markets and TSLA. Rates higher and China. Rates ok. China concerning. If you fight monster make sure you don’t become monster. China fighting zero COVID, risk of becoming monster. https://www.youtube.com/watch?v=--w5H...
“Higher taxes and more government are results not causes”. John Steinbeck “Grapes of Wrath”. Cause is basic human desire to work and create. Technology replaces humans and they can’t react fast enough. Education should be organized by VC because they have best insight into future. Key is to have agency = meaning + status. Anything which is predictable will be predicted. Some things are not predictable. Jobs that are hard to predict pay more. Stick with what you’re good at. Forget predicting stuff.
What would TWTR look like if Elon Musk had founded it? Grand master plan + engineering driven innovation stack + top talent. Grand vision= solve problem of deciding what matters with decentralized approach along lines of Byzantine generals problem and/or Bitcoin protocol. Protocol not platform. Reason he didn’t start TWTR is because tech talent wasn’t ready back then. Now available.
Talent development. Example MLS Next. US soccer is on the verge of becoming prime breeding ground for soccer talent. Why? Build elitist culture of performance and prestige around U15 and U17 age groups attracting best player and coaching talent. Like US academic system after WWII.
Evolution of biological systems. Analogy to investment process. Initial DNA, mutation, reaction to environment, how does the organism change with changing environment? Best indicator for wealth creation “ is the company interesting? Do I enjoy working on this investment?” Economics is theory of competitive response like Darwin is a theory of natural selection.
Nickel mining. Interesting because secular change downstream due to Tesla. High grade, NIMBY projects in US, Chile etc. Tesla is THE MARKET, that’s why they are ahead of curve in procuring materials. Decision to invest in mining is based on cost of capital which is TSLA.
Andrej Karpathy leaving Tesla on sabbatical. What should he be working on? Transformers for video to drive development of video animation for self supervised reinforcement learning. Go back to academia and create chair to develop talent for AI.2.0. This would help accelerate Teslas' mission to achieve autonomy.
Larry Fink letter to CEOs. New world order with selective globalization. Tesla at forefront. EV, solar, battery and NEW “Rethink NIMBY”. Tesla is pioneering manufacturing domestically to overcome NIMBY problem.
Fed influence on markets is game theoretic in nature. Signaling induces front running. No fundamental reason to react on Fed action. Fed induces more volatility. Solution is rule based monetary system.
When aspiring goals focus on two things. 1/2 get the job and 2/2 retain the job. Former is driven by actions with awe quality. Example in soccer GK prevented goals, spectacular saves etc. The former is long term like stay healthy, develop strong mindset and drill basics.
Book discussion “Circe” by Madeline Miller, a novel. Greek mythology is base for Western thought. Literature is base for human thought. Brain is like DNA and literature is RNA. Literature shapes thinking which shapes action. Three take aways. 1/3 Circe doesn’t follow prophecies she believes in her own will. 2/3 old thinking = prophecies , new thinking = own free will. 3/3 living means to love, fail, fight, win, loose. Goods don’t live, mortals do.
Privatizing sanctions. Companies are taking matters into own hands. Sign of deterioration of government institutions. It’s like mob rule which happens when governments don’t focus on safety. First time since WWII rogue state is challenging world order. Businesses are on top of global world order. “It’s the economy stupid.” Businesses are reacting to rogue state trying to push balance of power. Zeitgeist back to US post WWII with hope and purpose.
How will energy shock and de-russianizatuon affect Fed behavior and rates? No rate hike. Higher commodity prices need to flow through economy so supply chains adjust appropriately. China will have scalper role and mitigate impact of sanctions. Global economy a lot less affected by commodities flow through. Agile economies/companies will come out stronger.
Tesla is an energy company. The world has a problem right now which his how to de-russianize global energy. Tesla has been building infrastructure to solve exactly for this problem. Fundamentals of Tesla are extremely favorable.
Popper as investor = it’s about solving real problems. Wittgenstein as investor = just a game of words. Example of problem solving as investment is how to solve global energy when Russia is North Korea. Rouge states cannot be part of world markets. Apply new engineering practices to problem. Feels exciting. Solving problems feel s different than just applying methods.
Russia is turned into North Korea. Eventually every rouge state will go that way. Russian economy not impactful for global trade but energy markets will see adjustments through higher prices. Good for Tesla since Musk has been driving decarbonization of energy. Now it’s also de-Russianizitaion of energy = Sun + battery + EV. What started 20 years ago will now accelerate like Y2K or COVID effect on digitization.
Are we benefiting from serendipity or are we good performers? Every reader reads through his own prism. Introspection. We are not riding on serendipity. Principle stack. Investing in wealth creators means you have to think like them and adopt their lifestyle.
SpaceX modifies mission statement to focus on “Cyber attacks and signal jamming”. Follow new path on innovation stack. Good example of how Musk thinks about his companies.
Book discussion Julian Barnes “The Sense of an Ending”. Most people let life happen instead of making things happen in life. Action is what makes a person. Narratives are subjective. Little difference between thinking you have what it takes and actually having what it takes because your actions will be similar. Your actions define your person. Very Albert Camus.
Oil futures market pricing in non zero probability of Russian supply coming back in future. Also higher likelihood of US recession. US is hopeful and with purpose like in 1950s.
Western democracies are softening and what to do about it? Protect freedom must be priority. Rules of engagement. If you miss behave you get put in place. Dilemma. How to fight the monster without becoming a monster yourself. How to protect Popperian system of governance and error correction with potentially violent methods without jeopardizing the very system you fight to protect?
Is Elon Musk leaning towards right? No. He is fighting deterioration of government institutions. Creeping control of government institutions of how people behave in economy is bad. Musk like other entrepreneurs is a soldier of freedom and fights with innovation and rhetoric.
Book discussion Amor Towles “The Lincoln Highway”. Great narrative. Simple and powerful characters. Concept of road trip as a symbol for courage and change. Story line of hero character is diamond shaped. Start narrow then widen as life complexity increase then narrow again. Ending of book disappoints.
Ukraine Russia situation. Goal is to control gas assets. Fix institutional vacuum in Ukraine. Market impact not severe. Higher energy prices will accelerate recession that we’d have anyway due to post pandemic privileges effect. US must strengthen institutions.
Drawdown due to uncertainty around Russia. Recession risk up. Discussing two self driving car initiatives. Cruise full self drive taxi service. Wayve full self drive learning stack. Tesla has most advanced AI but can learn from competition.
Moneyball for sports, investing and war. Apply your resources where you get the best value when maximizing a win. What is a win? Invest in companies that try a lot and try where they get most efficient results. Tesla mobile app, supercharger payments, autopilot, mobile service.
Are you a giver or a taker? Invest in companies that are givers (Tesla) not takers (GM, ATT, Comcast etc). What job are you hired to do, or what can you give? We are hired to produce returns and help people with fear.
Solutions in the technology world require solutions in the human world. What do you do when a high performer gets held back by his boss? 1/3 make sure you’re on same page, same goals. 2/3 make sure he has no alter motives and 3/3 make sure you’re doing your best to work towards common goals.
Large draw down due to Russia risk. Ukraine invasion drives stocks down, oil up and an increase in recession risk. How to behave in situations when you set high goals and don’t get desired results and/or appreciation by superiors or customers? Scenario planning takes out anxiety. Manifesting bad and likely good scenarios. Fear is not the problem. It's the unknown. Reduce the unknown.
Call on interest rates and macro. We believe rates are plateauing and will decline in coming quarters. Inflation peak. Pandemic privileges retracting will push workers back to labor market, which will ease supply chains. Demand slowing down. Inventory build up. Tesla short term pain, long term gain. Will emerge even stronger from this because product and service better and lower cost.
Book discussion “Where the Crawdads sing” by Delia Owens. Three themes. 1/3 Nature 2/3 Love and 3/3 Justice. Good and bad in nature are just different shades of light. Nature grounds. Nature is efficient and only source of wealth. Love is necessary to achieve excellence. Love is life. Life is love. Justice is like nature, good and bad are different shades of light. Even moral philosophy and religion have different interpretations of justice dependent on circumstances. Justice is complex. Justice evolves and is subject to error correction.