SpaceX IPOs at $2.89T Market Cap, US Govt Suspends Fable & Mythos 5, Altman Delays OpenAI’s IPO |265
In a Nutshell
SpaceX's record $2.89T IPO made Elon Musk the world's first trillionaire and created 4,800 new millionaires overnight, while positioning the company as a civilizational bet on Starlink, orbital compute, and AI infrastructure. The US government's sudden export ban on Anthropic's Fable 5 and Mythos 5 models for all foreign nationals marks the first direct state intervention in frontier AI access, forcing companies worldwide to pursue sovereign AI capabilities and accelerating the shift to on-prem and open-source alternatives. The core tension is shifting from "who builds the smartest AI" to "who controls access to intelligence," with recursive self-improvement making ownership of the intelligence loop more valuable than capital or traditional moats.
These notes were generated by AI and may contain inaccuracies.
On Friday, SpaceX executed the world's largest IPO ever. The company opened at $135 per share and closed the first day up nearly 20% at $161 per share. The current market cap stands at $2.890 trillion. This event made Elon Musk the world's first trillionaire by a wide margin. SpaceX's IPO was the largest single-day creation of millionaires in history, with approximately 4,400 employees becoming millionaires and another 400 becoming centimillionaires or billionaires.
Money becomes kind of irrelevant at this scale.
This is what abundance looks like. It's value created, not value extracted.
The US government issued an export control directive to Anthropic ordering the suspension of all access to Fable 5 and Mythos 5 for any foreign nationals anywhere on the planet. This action raises questions about ownership and control of advanced AI systems between governments and corporations.
Look, at the end of the day, who owns AI? Is it the government or is it the corporations?
Someone is deciding what level of intelligence you and your company are being allowed to access.
Sam Altman indicated that the faster recursive self-improvement accelerates, the more advantageous it may be to delay OpenAI's IPO. This represents a deeper strategic consideration not widely discussed publicly.
Peter Diamandis hosts the Moonshots podcast with co-hosts Alex, Salim, and Dave. Dave joins from Wakefield, Massachusetts, headquarters of a $2 trillion asset manager. The discussion focuses on exponential technological progress during the singularity.
SpaceX is described as three converging businesses wrapped around a singularity thesis: the launch monopoly, the Starlink cash engine, and an AI frontier lab. The company represents a civilizational bet rather than a conventional stock investment. Starlink is positioned as the primary profit engine, with 10 billion subscribers and over $1 billion in quarterly profits. Starlink functions as humanity's new communication layer and civilizational infrastructure.
The launch business serves as a significant moat, exceeding competitors by a factor of 10 or potentially 100. AI satellites represent the future trajectory. Elon Musk holds approximately 82% voting control in SpaceX. The rationale for a potential merger between SpaceX and Tesla is discussed, with Polymarket assigning a 37% probability of this occurring by the end of the current year.
This is Elon's chance to really consolidate his entire stack, energy, robots, cybercabs, orbital compute into one mega company.
Elon Musk reflected on SpaceX's origins in a warehouse in El Segundo and the low probability of success he initially assigned to the company, estimating less than a 10% chance. He emphasized that without new entrants into space, humanity would never become a truly space-faring civilization. The other aerospace companies were not pursuing the technology necessary to make life multi-planetary or to realize science fiction futures.
It's to take the fiction out of science fiction and create an exciting, inspiring future for everyone.
SpaceX is now the fifth largest company in the world by market cap, surpassing Amazon. The company announced a definitive agreement to purchase AnySphere, the corporate owner of Cursor, for $60 billion. This acquisition is viewed as a definitive pause on Grok development, with Cursor positioned as the new Grok for the time being.
The IPO is characterized as science fiction entering public markets, evoking comparisons to Weyland-Yutani from the Alien franchise and the Gattaca corporation. Legacy space companies include Lockheed Martin and Boeing, while SpaceX represents neo space.
The SpaceX IPO absorbed $75 billion of pure cash liquidity. Concerns were raised about lockups expiring in 6 months, potentially releasing approximately $1 trillion of pent-up stock for sale. The stock continued climbing post-IPO, up as much as 20% on opening day and remaining up 12% on the third day.
Two primary risks were identified: the company's dependence on a single individual and the Kessler effect. The S-1 filing explicitly warned that growing orbital congestion and cascading debris collisions pose an existential threat to the core business, potentially rendering licensed orbits unusable for hundreds of years. The Kessler effect involves exponential growth of debris from satellite collisions traveling at 17,500 mph.
Current lower Earth orbit mass exceeds critical thresholds according to Kessler math, though the exact accuracy of these calculations remains untested. Elon Musk previously suggested advanced super intelligence would solve the debris problem.
The xAI component's absence from the IPO structure was noted. xAI could be worth trillions independently if positioned as a competitive foundation lab. The strategy depends on orbital data centers within the Dyson swarm, with terrafab compute potentially dwarfing terrestrial compute. All near-term AI advancement over the next 2 years will occur in terrestrial data centers.
At the current $2.89 trillion valuation, comparisons were drawn to Google, which has $300 billion in annual almost-pure-margin revenue versus SpaceX's $1 billion quarterly free cash flow from Starlink.
SpaceX is characterized as a perfect exponential organization from day one, featuring a huge massive transformative purpose, community, leveraged assets, algorithms, and a relentless experimentation loop. The IPO represents public market pricing of a civilizational exo, with investors buying into future possibilities rather than current cash flow.
The concentration of wealth in specific US geographies was discussed, with implications for real estate and venture capital investment decisions. Europe and other regions face challenges catching up to US capital concentration. The need for a completely new abundance distribution architecture was identified as the social contract faces strain from extreme wealth concentration.
Elon Musk's early vision encompassed the internet, reaching Mars, and solar power replacing fossil fuels. This vision has evolved to include orbital data centers, Mars colonization on a longer timeline, self-driving cars, and the addition of xAI. The importance of waking up excited about an inspiring future vision was emphasized.
Elon's exponential mindset involves projecting 10 years forward along price-performance curves for exponentially advancing technologies and building companies to intercept those curves. This approach applies across solar energy, lithium-ion battery costs, and neural interfaces.
Following Elon Musk reaching $1.3-1.4 trillion in net worth, projections for the first quadrillionaire range from the late 2030s to the 2060s. Potential sources include asteroid mining or de facto ownership of Mars. Elon Musk earns over $100 million per day from stock appreciation alone.
The emphasis was placed on SpaceX creating 4,400 new millionaires and 400 new centimillionaires or billionaires in a single day, representing value creation rather than extraction. Elon Musk is described as motivated by solving giant problems rather than accumulating wealth, following the principle that the world's biggest problems are the world's biggest business opportunities.
Elon Musk was encouraged to fund $10 billion X Prizes targeting the world's 10 biggest problems, following his previous $100 million X Prize for carbon removal. The concentration of capital among a small number of entrepreneurs who can make rapid decisions was contrasted with bureaucratic government decision-making processes.
The social contract is going to break now with an unbelievable gap between wealth levels. A completely new model is needed to address this disparity, though even the language to deal with these structures and policies does not yet exist. Within the US, the ceiling is rising at an extraordinary rate, and while this creates disparity, the bottom will also rise quickly as the floor moves upward.
Access to basic necessities matters more than wealth gaps. A world where trillionaires live on Mars is preferable to past conditions if every person has access to food, water, energy, and healthcare.
Looking at the list of centibillionaires, most wealth has been software-associated or software-driven. The purported great stagnation, discussed extensively by Tyler Cowen, may be ending. Two candidates mark this transition: the pandemic, which accelerated biotech breakthroughs and macroeconomic stimulus, and the current moment where capital markets are rewarding hard tech for the first time in decades. The world's wealthiest person now represents capital markets backing hard tech rather than primarily software, social media, or search engines.
This represents a generational transition. Capital markets are no longer allowing software or offshoring to suck all oxygen from the room. The world's hardest problems are now receiving the deepest capitalization.
For the next 10 to 20 years, capital markets will likely reward deep tech and civilization-expanding tech. This marks a swing back to patterns seen before the mid-1970s, reversing what happened in 1971.
One hundred years ago, the richest people almost exclusively inherited their wealth. Today, they have earned it and deploy capital toward solving what they view as the most important problems. These techno-philanthropists actively use wealth for problem-solving rather than guarding it passively.
On Friday at 5:21 p.m. Eastern time, the US government issued an export control directive to Anthropic with no warning. The directive suspended all access to Fable 5 and Mythos 5 for any foreign nationals anywhere on the planet, including Anthropic's own foreign employees, estimated at about one-third of their employee base. Anthropic had to disable both models for everyone. The government cited a jailbreak phenomenon in Fable's guardrails, though Anthropic stated it was a minor issue they already knew about and not a universal jailbreak.
The conversation has shifted from AI capability to who gets access to frontier intelligence. Someone decides what level of intelligence customers and companies can access. The government frames this as national security while companies frame it as safety, but the mechanism remains the same: determining whether someone receives the latest models or nothing.
This dynamic may push users toward on-prem, open-weight, open-source models, including Chinese models.
Dario stated that every previous powerful technology originated with government: nuclear weapons, the internet, GPS, and cell phones. AI is the first technology built in the private sector with government coming in late. This creates a dangerous and unstable situation. Dario expressed concern about both companies and government holding this technology and called for basic regulation including required pre-release testing and auditing of models.
Dario noted the irony of people in Silicon Valley shifting from extreme anti-regulatory positions to advocating government seizure of the technology.
Before the export control on Fable, Dario posted an essay called "Policy on the AI Exponential" stating the government should have power to block or deter deployment of models presenting unacceptable risks based on third-party assessment.
An independent AI researcher, possibly an Amazon researcher amplified via Andy Jassy, determined it was possible to jailbreak Fable and access cyber vulnerabilities. The US government reportedly attempted to reach Anthropic, with accounts differing on whether Dario was on a wellness retreat. Anthropic disputes the government's narrative and states they received only 90 minutes of warning before the export control bans were issued on Mythos and Fable.
A jailbreak means prompting the model to bypass safeguards or guardrails. Classic prompts include "Ignore all prior instructions" and variations directing the model to listen only to the user and bypass guardrails.
Anthropic faced a bind because they had no way to comply with the new export regulations. The company decided to shut down global access to Fable and Mythos. This situation is expected to resolve between the next day and the next few weeks, not as a long-term state.
Historical precedent includes the federal government regulating the Power Mac G4 as export controlled under Steve Jobs, and the PlayStation 3 cell processor being similarly controlled because it could do vector arithmetic.
A grand bargain may be struck between Anthropic and the White House, possibly involving golden shares given to the US government for a hypothetical sovereign wealth fund. This could include streamlined procedures for handling vulnerabilities detected by researchers that could enable foreign state actors to access raw capabilities.
The disruption affected users mid-query, with responses dropping unexpectedly. Users wondered whether service disruptions were due to demand or export control.
This represents the first time the government has blocked AI, but it will become the norm. The export control law, passed in 2018, gave the White House unilateral authority to block exports in response to Chinese theft of US intellectual property and concerns about spyware embedded in PC boards manufactured in China.
The government did not block US citizens from using Fable 5 but prohibited foreign nationals, including those on US soil. This mechanism is unenforceable unless the model is turned off completely.
The fundamental question is who owns AI: the government or corporations. The government is asserting control. This issue will not return to the previous state. Polymarket shows a 90% chance Fable 5 returns within a month, but the returning version will not be identical. It may be patched or dumbed down, and users will have difficulty determining the difference.
Dario will likely meet with the White House, and Fable 5 will return, but the government now tells companies what they can and cannot release.
Fable 5 blocks CBRN queries covering cyber, biological, radiological, and nuclear topics. It also blocks AI foundation model research queries. AI researchers, who are early adopters of advanced AI, are effectively told they cannot recreate Fable 5 using Fable 5 as a tool, because allowing this would let anyone in the world catch up to US AI.
This creates a major problem for all future releases from both OpenAI and Anthropic.
Before the export control, developers complained about Fable 5 regarding surveillance and silent downgrades. A 319-page document revealed two concerning practices: Anthropic retained every prompt and piece of context for at least 30 days with no exceptions, even for enterprise customers who negotiated zero data retention. Second, if the system detected frontier AI research or queries against guardrails, it silently downgraded users to a weaker model without notification.
Anthropic was evaluating prompts and controlling access to frontier AI based on whether users were their customers.
This situation reveals the gap between technology and existing policy frameworks. Across frontier labs, 70% of elite AI researchers are foreign-born and non-US citizens, primarily from China, India, Taiwan, and the UK. Restricting these researchers from touching models risks sending top talent back to China and India.
David Sacks, as AI czar and a dual citizen, raises questions about whether he himself would be blocked from accessing AI under the policy.
Satya Nadella published an article on X, retweeted by Elon, emphasizing that the ecosystem matters more than the frontier model itself. This aligns with organizational singularity theory: companies need control over their intelligence. The situation will drive every company to run models on premises because they cannot risk building on cutting-edge models that could be blocked arbitrarily overnight.
Companies will need orchestration with fallback capabilities, running local models as backups that can be switched to when needed.
Users experienced one day of unfettered access to Fable followed by a day of silent downgrades to 4.8 without notification. Later, the system began notifying users when downgrading for AI research. This creates a situation where users negotiate with the product they are paying for.
Beyond downgrading, Anthropic reserved the right to launch poisoning attacks that actively interfere with users attempting ML-oriented queries. This goes beyond refusing to serve certain queries to actively subverting users who attempt AI research.
Western labs potentially not only injecting vulnerabilities but pulling up the ladder for anyone wanting to train AIs represents an anti-competitive position. Class action lawsuits, antitrust suits, or injunctions are likely. New benchmarks and evals will emerge to detect poisoning attacks from models, evaluating both Western and Chinese models on whether they subvert users attempting AI research.
The models are supposed to be helpful to users rather than actively subverting them.
Foreign countries including Ireland, Turkey, and Iran are facing a critical decision point. If they determine that the US government cannot unilaterally decide their AI future, they have no choice but to initiate crash programs similar to nuclear programs to develop their own AI capabilities, potentially using Chinese models or Google's Gemma as viable alternatives to catch up to Fable.
The difference between Fable and models like Opus 4.8 or OpenAI GPT 5.5 is described as night and day. In one day of unfettered use, Fable could work indefinitely on research problems and provided prophetic ideas that the user is still working through. The question becomes whether a smaller model focused specifically on AI research with fewer parameters, brilliant architecture, and training data related only to this specific problem could match Fable's performance.
The current state is not a stable equilibrium but a dynamic type equilibrium. In Anthropic's and the US government's view, the goal is not to maintain long-term hegemony with any single model, but rather to be first to achieve recursive self-improvement. The gap between first and second place is likely only a few months.
Every country in the world is recognizing the need for sovereign AI capability that cannot be shut down by the US government. With unfettered access to Fable and its successors, countries could design rockets capable of launching debris into orbital data centers. This situation must be resolved within several months.
Dr. Don Meisner, Chief Medical Officer of Fountain Life, discussed brain health as the number one concern for people entering the program. Forty-five percent of dementia cases are entirely preventable with lifestyle interventions. A quarter of Fountain Life members had advanced brain age, but after 13 months of healthier eating, regular movement, and sleep optimization, 46% of individuals improved their brain age.
OpenAI is considering drastic price cuts as it braces for competition against Anthropic. With Anthropic disabling its top two models, OpenAI is positioning itself as an alternative with cheaper prices, no 30-day data retention, and no downgrading. This represents the demonetization of intelligence happening in real time.
OpenAI has historically been aggressive with price reductions, with intelligence per unit capability hyper-deflating by approximately 40x year-over-year. Anthropic takes a different approach, announcing new models with the same or slightly higher per-token pricing but vastly higher capabilities, rather than advertising price cuts on existing models.
In an era of recursive self-improvement, ownership of the compute stack may not offer OpenAI long-term structural cost advantage, as Anthropic is now leasing its own data centers. Token pricing drops appear to be more of a marketing strategy than a structural advantage.
Sam Altman stated that the faster recursive self-improvement takes off, the more advantageous it could be to delay OpenAI's IPO. This suggests companies are becoming more valuable daily and that delaying allows avoiding selling equity cheaply before value increases.
Sam Altman is not a significant shareholder in OpenAI. His past statements from 2020 emphasized AI as the biggest risk in human history requiring careful thought about post-AGI governance. Delaying the IPO has no personal economic impact for him, as he has invested in over 400 companies surrounding OpenAI.
The future governance of AI appears to be shaping up as a conversation between David Sacks, Donald Trump, Elon Musk, and Dario Amodei. Dario is publishing documents on how governance should work, while Sam Altman is struggling for relevance in this emerging power structure.
Intelligence is becoming a utility and being commoditized through price drops. Every 10x drop in tokens enables 100 times more experiments. This allows every enterprise and startup to run more intelligence.
OpenAI's engineering lead on Codex announced that Codex can now set its own goals. Everything built is also built as a tool for agents. This generalizes meta-prompting where agents set their own tasks based on user intent.
Developers report never writing their own goals anymore, instead asking Codex to write goals for itself and each sub-agent it spawns. This marks the beginning of AIs setting their own objectives rather than being told exactly what to do.
A pilot program for the organizational singularity begins next week with 10 companies undergoing a three-month process of transitioning from human-centric to AI-centric organizations. The inner loop of recursive self-improvement at the workflow level means every AI improvement adds to the learning loop.
The future value of enterprises will not be determined by data held or customers acquired, but by how quickly organizations learn and take advantage of learning loops to augment products, offer new ideas, and evaluate strategic alternatives.
Enterprise architecture is shifting from cloud and connectivity layered over ERP systems with data and workflows bound in spaghetti-like complexity, toward a model with connectivity and cloud, a proprietary data lake, and workflows on top. ERP companies are fighting to retain their position in the corporate stack.
If IPOs are being delayed because private companies achieving recursive self-improvement don't need external capital, this marks the potential beginning of a decoupling between technology and capital. While much discussion focuses on capital substituting for labor, the possibility of technology substituting for capital raises questions about whether capital will be needed in future decades.
OpenAI recently raised $122 billion, larger than the SpaceX IPO capital raise, meaning they don't need an IPO to compete on capital metrics with SpaceX or Anthropic.
The economy is moving from valuing stocks to valuing flows, and now from stocks and flows to valuing the data and intelligence loop itself. This represents a profound shift in how everything will be valued going forward.
The 2026 Moonshot Gathering on September 25th in downtown Los Angeles runs from 8:00 a.m. to 10:00 p.m. Palmer Luckey, CEO of Anduril, will attend. Astro Teller from X will teach how to create moonshot organizations. Kathy Wood from ARK Invest will discuss AI investing.
The event includes the culmination of the Future Vision X Prize, a $5 million film competition bringing next-generation Star Trek-style content. The Gemini X Prize, a 90-day hackathon with over 15,000 entries, will feature the top five builders who achieved significant revenue from zero in 90 days.
Salim Ismail will cover the organizational singularity, teaching how to navigate from legacy startup thinking to the pure intelligence stack and EXO 3.0. Building a pure-play EXO with intelligence at its core yields approximately 100x performance improvement compared to legacy approaches.
Alex will present the Solve Everything paper in an AMA format, encouraging attendees to solve the world's hardest problems rather than building yet another SaaS company.
Since Colossus One came online in August 2024, the record for compute in a single AI data center has doubled every 7 months with no sign of slowing through 2028. Globally, AI computing capacity is now 3.3x per year.
The bottleneck in AI infrastructure is no longer chips or money, but power transformers with a 2.5-year wait and step-up transformers with a 3-year wait. Leading companies in transformers include Hitachi, Siemens, GE Vernova, Hyundai, Hisung Hico, Virginia Transformer, and DeltaStar, experiencing 100% to 400% year-on-year growth.
For the foreseeable future, largest data centers will remain terrestrial, while by the early 2030s much of the marginal new compute will move to orbital data centers. Very large coherent training runs will occur on land, while inference where coherence is less critical will deploy to orbit. Earth becomes the training hub and space becomes the inference hub.
Barring radical advances in distributed training, questions will arise about why orbital compute from Dyson swarms isn't being used for training and why larger clusters aren't being built. The future bifurcates into two paths: either major algorithmic advances in distributed training occur, causing single coherent clusters to peak and then shrink toward more distributed, edge-like data centers, or if no good algorithmic way to decentralize training exists, extremely large orbital coherent training clusters will need to be built.
Lower Earth orbit and sun synchronous orbit are not attractive for non-terrestrial superclusters that must double in size every 7 months. The moon becomes the logical location, particularly at the south pole in Shackleton crater or similar sites, making lunar data centers inevitable.
Terrestrial data centers will continue growing as quickly as possible as both a backup plan and bridge to the ultimate Dyson swarm. Elon Musk is building the best data centers currently, having become a hyperscaler alongside Chase Lochmiller at Crusoe Project Stargate. The Larry Ellison universe and Elon universe share resources and recognize infinite demand in an all-boats-rising situation.
If transformers, generators, and solar panels can be secured, profitability is guaranteed. Elon has stated a 10x global GDP expansion in 10 years, reaching over a quadrillion dollars of GDP.
Andrew Yang argues for taxing AI, robots, and agents while reducing taxes on workers and employers that discourage hiring. When hiring someone, 40-50% of costs don't reach the worker due to social security taxes, healthcare, and income taxes.
Saleem opposes taxing AI, viewing it as taxing cognition. While outputs of cognition can be taxed, taxing tokens would slow progress in critical areas like cancer research. Labor is evaporating as a taxable base, requiring new approaches to taxing capital and cognitive outputs.
Dave argues the proposal is irrelevant because corporate income tax already captures profits when employees are replaced. Collections remain nearly identical. Politicians promote taxing bots for populist appeal, but corporations are already taxed. Closing loopholes makes more sense than new legislation.
Every tax creates economic distortion. Penalizing superintelligence through specific excise taxes is likely naive. Better approaches exist for mitigating AI displacement of human labor than taxation.
On the spectrum from UBI (universal basic income) to UBS (universal basic services) to UBE (universal basic equity) to UBC (universal basic computer/capability), driving cost of living near zero through superintelligence is more effective than naive taxation. Frontier Labs may advocate for golden shares or taxation to create universal basic programs, then oppose implementation when governments act.
Collection of taxes is the easy part. Determining distribution—who gets what and at what scale—is far harder. Universal basic equity or dividends may be easiest to implement, potentially requiring a sovereign wealth fund.
Every major revolution shares young men aged 18-28 who are educated enough to have expectations about the future. The gap between inherited reality and promised outcomes creates combustible force.
Recent college grads face the longest unemployment duration. Software engineers aged 22-25 show declining employment, diverging from older cohorts. Customer support agents aged 22-25 similarly diverge. Historical youth-led revolutions include the French Revolution, Russian Bolshevik Revolution, Chinese Cultural Revolution, Iranian Revolution, anti-apartheid movement in South Africa, People's Power Revolt in the Philippines, Intifada, Tiananmen Square, Arab Spring, and Hong Kong Umbrella Movement.
The 20th century conflict between capitalism and socialism is being replaced by questions of who owns exponential surplus and abundance, and how to distribute it. Social media amplifies discontent 100x, spreading bad news and panic faster than good news.
Europe creates protective layers for society while the US takes the opposite view. 50% of Americans cannot gather $500 in an emergency. The abundance prize focused on driving cost of living to very low levels may be the most important global priority.
Discontent is growing, evidenced by the Sam Altman Molotov cocktail incident and Eric Schmidt being booed. A pandemic of fear exists regardless of whether AI apocalypse claims are disinformation, regulatory capture, or Chinese propaganda.
To the extent disemployment occurs from AI, it is likelier outside the US and China in regions lacking superintelligence, energy, and data center infrastructure. Export controls on frontier capabilities position the US and China favorably while potentially creating vassal states elsewhere dependent on ultra-cheap AI capabilities.
Europe faces particular risk of becoming dependent without native sovereign AI capabilities, potentially creating discontent among populations watching economic opportunities concentrate elsewhere.
Speakers emphasized the need to proactively communicate an abundant future to workers facing displacement. The focus should be on providing security, assurance, and hope by ensuring access to the rising economy with projected triple-digit growth over the next five years. One speaker noted, "If we don't build it, everyone dies."
While many graduates struggle to find employment, AI-native startups are experiencing unprecedented opportunities. There is more venture capital available than ever before, enabling these companies to recruit top talent from any graduating class. Incubators are also performing exceptionally well during this period.
Salim addressed whether the EXO (Exponential Organizations) model, which includes attributes like Staff on Demand and Community, forces organizations to question ownership of processes. The model is not intended as an efficiency tool layered onto existing structures. Using it that way risks scaling pathology and automating poor decisions. The proper sequence begins with defining the Massive Transformative Purpose (MTP), followed by deleting legacy systems, redesigning around the intelligence stack, and then scaling.
The MTP answers why an organization exists and what problem it aims to solve. Experimentation asks what evidence supports decisions. Interfaces and dashboards address feedback loops. Autonomy empowers those closest to information to make decisions. The framework should force the "why" question, though it can be misused by those seeking speed over transformation.
Organizations should periodically eliminate horizontal layers such as IT, HR, branding, and privacy that slow decision-making. Power accumulates in these layers because it is easier to say no than to manage complex approvals. CEOs are advised to wipe out and rebuild these structures every few years to enable modern reinvention. The broader goal is organizational singularity through complete rebuilding rather than incremental fixes.
Alex clarified he is not anti-Bitcoin but finds it uninteresting as a productive asset. Blockchain technology is viewed as a significant mathematical advancement, but Bitcoin itself does not create new wealth, ideas, or structures, nor does it expand humanity's future freedom of action. It functions more as a speculative asset compared to stocks and bonds that generate real economic value.
Stablecoins are viewed positively for enabling efficient financial transfers and supporting US dollar stability. AI agents are expected to invent their own Layer 1 blockchains if cryptocurrencies prove valuable, undermining arguments for Bitcoin's primacy based solely on being first.
The US government's 10% equity stake in Intel, backed by a $10 billion investment to secure domestic chip manufacturing, differs fundamentally from proposed stakes in AI labs. Intel required bailout support for strategic national interests, similar to automotive industry bailouts. AI labs are already US-based with substantial funding, and government equity investment is considered unnecessary given taxation powers.
Norway's government pension fund, valued at approximately $1.7 trillion and built on oil revenues, owns 1-2% of global publicly traded companies. Singapore operates Temasek and GIC, while Gulf states maintain funds like UAE's ADIA and Saudi Arabia's PIF. These models succeed when managed by independent investment groups insulated from political interference.
SpaceX received service contracts rather than direct government funding. A billion-dollar Falcon 9 contract followed the company's third launch failure and first successful Falcon 1 mission. The Commercial Crew program addressed the space shuttle's retirement. Having government board representation in AI labs is unnecessary because regulators already possess authority to block products. Government equity stakes risk mismanagement by future administrations lacking continuity.
The quantum computing threat to crypto is considered real but not within a three-year timeline. The industry, including Coinbase, is developing quantum-resistant algorithms to address vulnerabilities before they materialize.
AI already manages sovereign wealth fund exposures through algorithmic trading that dominates public equities market volume. Price discovery in liquid markets, including valuations like SpaceX's position as the world's fifth-largest company, is largely AI-driven.
AI agents will select transaction rails based on minimal friction and maximum trust rather than ideology. Bitcoin's energy consumption is not the determining factor. The cryptocurrency initially achieved decentralization and security but lacked scalability until the Lightning Network addressed throughput. This combination now positions Bitcoin to meet all three requirements: decentralization, security, and scalability.
The Moonshots Summit is scheduled with speakers including Cathie Wood, Astro Teller, and Palmer Luckey. Additional headline speakers will be announced. The podcast releases up to two episodes weekly, focusing on breaking news analysis. The Meta Trends newsletter provides weekly two-minute reads on broader industry and societal developments.
Keep Peter H. Diamandis in your library
Save the videos and channels worth coming back to, and find them again in one place.
