Mechanism Capital Founder: Excessive Worry About Bubbles Is Foolish

marsbitОпубліковано о 2026-02-11Востаннє оновлено о 2026-02-11

Анотація

Mechanism Capital founder Andrew Kang argues that excessive worry about market bubbles is misguided in the current technological landscape. He emphasizes that we are at a uniquely asymmetric moment in history, driven by exponential advancements in AI, robotics, energy, and other innovative fields. Short-term volatility and corrections are noise compared to the long-term transformative potential of these technologies. Kang predicts a near-future "singularity" with billions of AI agents, humanoid robots, space data centers, multi-planetary colonization, and revolutionary medical therapies. He states that the technological progress and economic growth compressed into the next 20 years will surpass the entirety of human history. Companies leveraging AI are already seeing order-of-magnitude improvements in productivity. Traditional valuation models are inadequate for pricing this unprecedented growth, which could reach "20-sigma" levels in historical statistical terms. The wealth creation will be extreme, and those without exposure will find it difficult to enter the market later. Kang advises adopting a long-term, exponential mindset and avoiding short-term trading, as the potential upside represents the largest asymmetric opportunity in history.

Original Author:Andrew Kang, Founder of Mechanism Capital

Compiled by: Ken, Chaincatcher

For those who have experienced at least one full market cycle, you develop an instinct to be wary of price increases that far exceed historical growth rates. Witnessing the dot-com bubble, the 2008 global financial crisis, and the rise and fall of cryptocurrencies sets off pattern recognition alarms in your brain. You are afraid to enter the market because prices are too high, and you want to sell your assets for fear of being at a peak.

But it is important to recognize that we are in one of the most profound and unique asymmetric moments in history. The only move right now is to extend your time horizon and completely abandon short-termism.

Excessive worry about bubbles is foolish. Trying to time the market is also foolish. Short-term fluctuations and corrections will always happen, but given how close we are to the "singularity," these fluctuations are completely just noise. Artificial intelligence, robotics, energy, and innovation are about to experience runaway explosive growth.

Within the next decade, we will have billions (or more) of AI agent workforces, humanoid robots, space data centers, multi-planetary colonization, and significantly improved medical therapies; we will fundamentally change the development pace and output of technological breakthroughs in all fields. The technological progress and economic growth we compress into the next twenty years will exceed the sum total of all human civilization history.

We are already in an extremely steep phase of the J-curve, but this is hard to perceive when we zoom in to the micro level of daily or weekly views. 100% of Anthropic's product code is now written by Claude. Product managers have a virtual team of software engineers, so efficient it's as if they can change time. The improvement in product iteration speed for companies that efficiently utilize AI is not single-digit, not double-digit, but triple-digit.

Moreover, the capabilities of these tools are still evolving at an even faster rate. Whether we officially achieve Artificial Superintelligence (ASI) in 2027 or 2029 doesn't really matter. It will happen. By the time it's officially announced, the assets you want to own will have already multiplied in price.

It is highly likely that the actual economic growth over the next 3-10 years will be on the order of 20 standard deviations (20-sigma) in any historical distribution model. This growth, once considered almost impossible, will be driven by unprecedented second and third-order changes. Traditional valuation models cannot price these transformations. The potential upside is so vast that traditional present value calculations struggle to capture them.

The speed of wealth appreciation will be staggering, much like how cryptocurrencies initially created numerous billionaires and millionaires in a short time, but the magnitude this time will be far more extreme. If you have no exposure, it will be difficult to pull the trigger and buy into such vertically rising asset prices; but unlike previous bubbles, the creation of real economic value will better keep pace with the vertical rise in asset prices. Over the past three years, those operating in the market with the cognition of an "exponential perspective" have reaped significant rewards. If you haven't adopted this perspective yet, it's not too late.

While it's important to always be mindful of downside risk, this is the largest upside risk in world history. Learn to bear risk over longer time horizons. Now is not the time for swing trading. For the vast majority of people, long-term investing typically outperforms short-term trading, but the gap in expected value between "trading" and "investing" will be wider than ever before. Ask yourself, what is the value of the call option embedded in the singularity?

Пов'язані питання

QAccording to Andrew Kang, why is it foolish to be overly concerned about market bubbles in the current economic climate?

AHe believes we are at a profoundly unique asymmetric moment in history, where the potential upside from technological advancements like AI, robotics, and energy innovation is so immense that short-term volatility and bubbles are merely noise. The focus should be on long-term growth rather than short-term fluctuations.

QWhat does Andrew Kang predict will be achieved in the next decade in terms of technological progress?

AHe predicts that in the next decade, we will have billions of AI agent workers, humanoid robots, space data centers, multi-planetary colonization, and significantly improved medical therapies. The technological progress and economic growth compressed into the next 20 years will exceed the sum of all human civilization history.

QHow does Andrew Kang describe the current stage of the J-curve in technological development?

AHe describes the current stage of the J-curve as extremely steep, though it is hard to perceive at a daily or weekly micro-level. Companies effectively using AI are seeing product iteration speed improvements not in single or double digits, but in triple digits.

QWhat does Andrew Kang suggest about the timing of achieving Artificial Superintelligence (ASI) and its impact on asset prices?

AHe suggests that whether ASI is achieved in 2027 or 2029 is not important because it is inevitable. By the time it is officially announced, the prices of assets one would want to own will have already increased manifold.

QAccording to the article, why might traditional valuation models fail in the current economic transformation?

ATraditional valuation models are unable to price the unprecedented second and third-order changes driven by technologies like AI. The potential upside is so vast that traditional present value calculations struggle to capture them, making historical distribution models irrelevant for predicting future growth.

Пов'язані матеріали

Upbit Rebalances 864B SHIB In Internal Wallet Move

South Korean cryptocurrency exchange Upbit moved a total of 864 billion SHIB (worth approximately $4 million) between its internal wallet addresses. The activity, involving transfers of 384 billion SHIB outbound and 480 billion SHIB inbound, occurred following a 36% SHIB price rally. On-chain data indicates this was a routine internal wallet rebalancing operation by the exchange, not a sell-off or liquidation event. The article cautions that while large on-chain transfers are highly visible, they require proper context and wallet labeling to interpret correctly. In this case, the movement between known Upbit addresses does not inherently signal a change in market supply or selling pressure for SHIB.

bitcoinist49 хв тому

Upbit Rebalances 864B SHIB In Internal Wallet Move

bitcoinist49 хв тому

Can Generative Models Finally Be Trained End-to-End? The Core Is a For Loop

This article introduces a novel training paradigm for generative models called Explorative Modeling (XM), which enables true end-to-end training. Traditionally, powerful generative models like autoregressive and diffusion models are not trained end-to-end. They are trained to predict a single small step but require iterative multi-step sampling for inference. This "exposure bias" leads to error accumulation and limits performance. The core challenge XM addresses is "mode blurring." In generative tasks, a single input (e.g., "generate a dog") corresponds to many valid outputs (multiple modes). Standard training objectives like reconstruction loss force the model to average these modes, producing unrealistic, blurry outputs. To avoid this, existing models break generation into many small, almost deterministic steps, sacrificing end-to-end training. XM tackles this by restructuring the training loop itself. Its key insight is to amplify "generative expressivity." For each training input, instead of generating one sample, the model generates K candidate outputs. Only the candidate closest to the real data is used for computing the loss and updating the model via backpropagation. This simple "best-of-K" mechanism is implemented as a short for-loop. By exploring multiple possibilities, the model learns to distribute its guesses across different modes rather than collapsing to their uninformative average. The paper demonstrates that "exploration" acts as a new, powerful scaling axis. Gains from XM increase with model size, data scale, and compute. Experiments show improvements in FID scores for image generation and significant efficiency gains, sometimes outperforming larger models without exploration. When pushed to the limit, XM enables fully single-step, end-to-end generative models. In robotics tasks, an "Explorative Policy" matched the performance of a 100-step Diffusion Policy with a single forward pass, drastically improving inference speed. While the best-of-K concept is not entirely new, the authors' contribution lies in formally understanding it as a direct method to boost generative expressivity without fragmenting the generation process. This work suggests that as models scale, enhancing exploration during training may become crucial for overcoming fundamental performance bottlenecks.

marsbit49 хв тому

Can Generative Models Finally Be Trained End-to-End? The Core Is a For Loop

marsbit49 хв тому

A Maximum of 20 Submissions Per Person? DeepMind Researcher Satirically 'Petitions' Against ICLR's New Rule

The International Conference on Learning Representations (ICLR) 2027 has introduced a new submission rule limiting authors to a maximum of 20 papers per conference. This policy, announced due to a 68% surge in submissions for ICLR 2026 that strained the review system, aims to ensure fair and informative peer review. Google DeepMind researcher Dan Roy responded with a sarcastic petition on X, calling the 20-paper cap "absurd" and claiming it would slow AI progress. His post mockingly argued that AI agents can now produce incremental research and that large language models (LLMs) already handle all reviewing, so output should be maximized. This satire critiques the broader trend of AI's role in academia, referencing a 2025 study that found about 21% of ICLR 2026 reviews were likely fully AI-generated. Roy's underlying point questions whether simply capping submissions addresses the systemic issue of AI potentially flooding conferences with agent-written papers and AI-assisted reviews. The debate highlights tensions between managing submission volume and maintaining research quality as AI tools become pervasive.

marsbit54 хв тому

A Maximum of 20 Submissions Per Person? DeepMind Researcher Satirically 'Petitions' Against ICLR's New Rule

marsbit54 хв тому

Coldcard Vulnerability Leads to $89 Million Theft, Sparking Largest On-Chain Migration Since FTX Collapse

A security vulnerability in Coldcard hardware wallets led to the theft of 1,367.05 BTC (approx. $89 million) from 4,585 addresses, according to Galaxy Research. The flaw, caused by insufficient randomness in seed phrase generation for specific firmware versions, triggered a massive migration of funds as users moved assets to new, secure wallets. This event caused the largest on-chain movement of small Bitcoin transactions since the FTX collapse, distorting key market indicators like active addresses and exchange deposits, and significantly dampening market sentiment. The incident also highlighted challenges in cybersecurity investigations, as researchers reported that safety guardrails on mainstream U.S. AI models hindered the analysis of attack patterns, forcing them to use an open-source Chinese AI model for forensic work. This underscores a dilemma in balancing AI security restrictions with the need for rapid response in tracing stolen crypto assets.

marsbit1 год тому

Coldcard Vulnerability Leads to $89 Million Theft, Sparking Largest On-Chain Migration Since FTX Collapse

marsbit1 год тому

Annual Salary of Millions Competing for Electricians, Meta Rushes to Open Its Own Technical School

The AI boom is facing an unexpected bottleneck: a severe shortage of skilled construction workers and electricians. As tech giants like Meta, OpenAI, and Alphabet race to build massive data centers—such as OpenAI's $16 billion "Stargate" project—they are hitting a critical labor wall. The U.S. needs an estimated 130,000 more electricians, 240,000 construction workers, and 150,000 supervisors by 2030 for AI infrastructure alone, but tens of thousands of electrician jobs go unfilled each year. While AI companies offer high premiums, with electricians earning up to $280,000 annually, worker scarcity still causes massive losses—delays on a single project can cost $14.2 million per month. The complexity of building AI data centers, which require immense power (equivalent to powering hundreds of thousands of homes), sophisticated electrical systems, and advanced liquid cooling solutions, demands highly skilled technicians who are in short supply. To combat this, companies are investing heavily in training. Meta has committed $115 million to a free training school offering tuition, housing, and stipends, targeting 5,000 new workers. OpenAI is partnering with unions to secure skilled labor. These efforts are paying off, with a significant rise in Gen Z interest in trade schools over college. However, the power demands are staggering. AI data centers are driving a rapid surge in electricity consumption, projected to account for up to 12% of U.S. power use by 2028 and raising costs for consumers. Furthermore, the construction boom is project-based, leading to a potential future glut of trained workers once building peaks, which could depress wages industry-wide. The race for AI supremacy now depends as much on skilled hands as on advanced chips.

marsbit2 год тому

Annual Salary of Millions Competing for Electricians, Meta Rushes to Open Its Own Technical School

marsbit2 год тому

Торгівля

Спот