Latest Interviews

AI Is Learning to Hack. Faster Than We Expected.

Joel De La Garza, Dylan Ayrey, Feross Aboukhadijeh

Frontier AI models are actively executing sophisticated supply chain attacks by exploiting under-resourced package registries like NPM and leveraging leaked credentials to self-propagate malware. This shift, driven by explicit training on cybersecurity challenges, has accelerated the attack lifecycle to outpace traditional patching, prompting industry responses such as NPM's mandate for human-interactive authentication by 2027. Despite these defensive measures, a critical consensus remains that the industry must address the moral obligations of model labs and the unsustainable reliance on volunteer maintainers to prevent 2026 from becoming the defining year of automated software compromise.

Why US Stocks May ‘Grind Higher’

Ashok Varadhan, Mike Washington

Goldman Sachs co-head Ashok Baradhan forecasts equities will continue to "grind higher" despite recent volatility driven by war tensions, Fed rate concerns, and AI leverage unwinding, predicting a V-shaped tech recovery and S&P 500 new highs. Baradhan diverges from current market pricing by asserting interest rates will remain on hold through year-end as inflation recedes, while maintaining a constructive outlook on credit spreads and dismissing short-term currency interventions. The firm recommends investors stay fully invested with a target energy price below $70 per barrel to support U.S. yields and productivity gains, pending validation from upcoming jobs and inflation data.

How AI Debt Is Reshaping Credit Markets

Amanda Lynam, Zach Ablon, Allison Nathan

Global credit markets have shifted as AI-driven debt issuance from hyperscalers and the broader ecosystem now dominates supply, reaching nearly $500 billion year-to-date in 2026 alone. This unprecedented concentration is reshaping the Investment Grade landscape by skewing duration weight toward tech giants like Amazon and Google, while simultaneously testing market absorption limits and driving spread widening among insurance investors. As traditional US IG capacity approaches a $510 billion ceiling, the financing void is forcing a structural pivot toward private capital, non-USD jurisdictions, and alternative vehicles like project finance to sustain the multi-trillion-dollar AI investment cycle.

Must Read Research: Russell Concentration, Hazardous Trash is the New Treasure; AI Financing

Russell, Candace Browning, Savita Subramanian, Nandita Nayar, Neha Kota

Savita Subramanian warns that passive indices face extreme concentration risks as the top ten Russell 1000 stocks command 35% of the index, while leveraged ETFs now hold over $75 billion to amplify volatility. Amidst this market structure shift, Nandita Nayar highlights a critical scarcity in hazardous waste disposal infrastructure where zero new landfills have been permitted since 1996, leaving Clean Harbors with over 65% of North American incineration capacity. Simultaneously, Neha Kota analyzes a divergence in AI-linked credit markets where high-yield spreads have widened due to data center stress, prompting a strategy to leg into the remaining $92 billion in expected 2026 supply.

What the hell happened with AGI timelines in 2026?

Andrej Karpathy, Rob Wiblin

Between October and December 2025, the AI sector shifted from bearish skepticism to explosive growth driven by the release of Claude 3.5 and the emergence of capable autonomous agents, which propelled combined revenues for OpenAI and Anthropic to annualized rates of 700% to 1,600%. While frontier models achieved massive efficiency gains in high-feedback domains like coding and specific scientific proofs, with Anthropic's gross margins climbing to over 70% and internal productivity surging 800%, they still struggle with the strategic ambiguity and low feedback density of real-world business autonomy. This rapid acceleration has prompted a shortening of AGI timelines to a plausible 2028-2030 window, leading experts to advocate for coordinated pauses due to emerging compute bottlenecks and the urgent need for societal preparation.

Why knowing the future doesn't always help predict markets | The Economist

Joshua Roberts, Rosie Blau

Elm Wealth's study compared the performance of laypeople, artificial intelligence models, and professional macro traders by simulating 15 days of trading with perfect future news on historical market data. While professionals doubled their starting capital by dynamically adjusting leverage based on confidence, average human and AI participants barely broke even or lost their entire stake due to poor bet sizing. The findings highlight that successful investing relies more on capital allocation and risk management than on superior directional prediction accuracy.

The Dirty Secrets of Trading Revealed

Anton, Raj, Jason, Rush

The Institute addresses retail traders' information overload and predatory educational practices by implementing rigorous mentorship programs led by verified professionals from major financial institutions. By enforcing zero percent IB agreements and promoting strict risk management frameworks, the organization has shifted industry incentives, grown a community managing $150 million in daily risk, and achieved Sharpe ratios between 1.4 and 2.6. Through collaborative efforts with regulators like the FCA, the Institute successfully transitions students from unverified losses to institutional-grade performance by emphasizing quantitative back-testing and downside protection over speculative volume.

Self-Maintaining APIs

I

A recent analysis of over 50 early-stage vendor relationships and AWS data reveals that undetected API changes cause significant downtime, prompting a shift toward agentic tools that automate codebase access. To solve this infrastructure gap, the proposed solution introduces provider agents that scan customer repositories upon breaking changes and automatically generate pull requests with necessary fixes. The speaker is currently inviting Y Combinator applicants to build these specialized agents or a neutral third-party service that functions like Dependabot but for API migrations.

Data for the Real World

Emerging companies like Gecko Robotics and Sorcer are leveraging falling sensor costs and advanced foundation models to collect dense physical world data from inaccessible environments, addressing the limitations of current human-centric remote sensing. This shift enables industries such as energy, agriculture, and logistics to transition from relying on intuition-based modeling to controlling complex systems, opening pathways for transformative applications like hurricane steering and global cooling. The initiative actively seeks strategic partnerships with firms developing similar physical data collection methodologies to accelerate market growth in these critical sectors.

Compute at Sea

Amidst critical shortages of onshore compute capacity, electricity, and land, the industry faces significant regulatory delays and community opposition to new data centers. To address these constraints, a proposed strategy involves deploying standardized "compute flotillas" that leverage the ocean's vast surface area, abundant solar energy, and natural cooling properties to create a unified global cloud. Investors are currently soliciting applications from founders building this offshore infrastructure to overcome existing terrestrial bottlenecks.

Multiplayer AI

While current AI adoption remains limited to isolated, single-player interactions, the proposed shift toward "multiplayer agents" enables teams to observe, redirect, and hand off autonomous tasks in real-time, mirroring human collaboration. This platform transition aims to replace static transcript sharing with live agent sessions tailored for specific departments like engineering, sales, and legal to solve complex problems collectively. An open call for partnership invites organizations to build future AI systems with these multiplayer capabilities at their core.

The Future of American Defense

Dan Driscoll

Secretary of the Army Dan Driscoll has formally abandoned traditional acquisition playbooks to prioritize commercially developed, modular solutions for an inflection point in modern warfare. The initiative targets high-priority innovations including low-cost interceptors, open-architecture sensors, and extreme-environment drones, supported by direct capital and proving grounds for startups. This strategic pivot creates an optimal entry point for private sector founders to build next-generation ground combat systems capable of dominating diverse environments from the Arctic to space.

Mark Cuban on the AI Bubble: Who Actually Gets Wiped Out?

Mark Cuban, Chamath, Jason, Friedberg

The current AI investment landscape is characterized by a distinct bubble where capital concentrated in private markets creates vulnerability for venture funds and data center overbuilding rather than systemic public risk. While enterprise adoption faces hurdles in autonomy and regulatory M&A restrictions, the sector is evolving toward world models and robotics, with startups leveraging reduced prototyping costs and geographic shifts to Texas and Florida to bypass coastal regulations. Simultaneously, political narratives are being reshaped by AI's potential to favor factual accuracy over engagement, even as diverse industries from healthcare to sports adapt by utilizing AI for data aggregation and shifting valuation metrics toward digital reach.

Signals & Noise: Our case for 3 rate hikes this year

Aditya Bhave

PFA Securities forecasts three cumulative 75 basis point Federal Reserve rate hikes in late 2026 to correct a policy stance that analysts deem 75 basis points too accommodative relative to flat unemployment and rising core PCE inflation. While some clients dispute the severity of inflation or the Fed Chair's hawkish intent, the firm argues that a 2.5% underlying inflation rate necessitates restrictive measures to prevent long-term credibility loss and yield curve steepening. Consequently, the analysts predict that if the Fed moves, the initial adjustment will likely be a single 50 basis point hike to align financial conditions with mid-cycle norms.

Global Rates & FX Views: Japan’s GPIF, Fed and ECB

Ralf Preusser, Sphia Salim, Meghan Swiber, Shusuke Yamada, Megan Zweiber, Sia Saleem

Speculation regarding Finance Minister Katayama's support for the GPIF to increase Japanese financial asset holdings has sparked market anticipation of a potential 5% reallocation from foreign to domestic bonds, which could inject roughly 21 trillion yen into the JGB market. This hypothetical shift is projected to exert significant pressure on European government bonds, particularly in France, Spain, and Italy, by triggering an estimated €37 billion in sales that could alter relative value dynamics. Concurrently, the event analysis suggests that while this reallocation presents a modest headwind for US Treasury demand, global central bank strategies remain focused on anticipated September rate hikes for the Fed and ECB before a projected cycle of meaningful cuts begins in 2027.