曹卿云 Qingyun Cao × Ion Stoica

AI基础设施的创业切口正转向后训练、多模态数据、机器人及跨云训练与数据搬运，尤其受算力稀缺和数万亿级训练数据推动。开放模型（不少来自中国）经后训练可能在特定任务超过专有模型，而模型排名每两个月左右重排，后续竞争关键在任务阈值、真实人类评估与商业化收入。

Paul Tudor Jones × Adam Grant × Joe Kernen × Becky Quick × Andrew Ross Sorkin × Katie Kramer

Paul Tudor Jones sees a 1999-like, two-speed market: short-term weakness could give way to year-end strength, but 2027 may be “really, really challenging.”Jones bases the warning on IPOs, secondaries and unlocks—including SpaceX, Anthropic and OpenAI—reversing equity flows, while AI issuance and sovereign borrowing pressure debt markets; he sees 60/40 odds of a forced fiscal resolution.

Patrick O'Shaughnessy × David Vélez

Vélez sees fintech disruption in its first inning: roughly 90% of the $10 trillion consumer-finance market still runs through branches.Nubank serves 140 million customers with 10,000 employees, at about 3% of traditional incumbents' cost to serve.Its US launch and stablecoin-backed accounts in upwards of 35 countries test whether that efficiency can travel, with organic brand-building the key hurdle.

Shanghao Jin

周五数据显示，convexity hedging减少、swaption squeeze近尾声；但Shanghao Jin未核验更新数据，长端利率未降而波动率回落。油价仍是尾部风险，利率上升可能重新触发对冲。AI债权人承担近似股权的需求风险，融资与再融资压力下，CSP现金流是关键验证。

曲凯 × 莫子皓

子皓认为 personal agent 的消费级时机已到，OpenClaw 却显示产品与传播链路仍是瓶颈。Ambi押注线下 context 和持续交互，降低获取智能成本，并以交互沉淀迁移成本。模型仍在进步，但分发、电池与佩戴体验仍限制扩张。

John Collison × Mark Zuckerberg × Alexandr Wang

Meta is making a $100 billion-scale infrastructure wager to return to the frontier with Muse Wonder, a larger model optimized for personal and professional agents, active memory, long-horizon focus and multimodality.Paid access, APIs, enterprise distribution, commerce and existing-business growth must fund the buildout, while resale or internal use partly hedges downside if models or AI disappoint.

Martin Shkreli

Shkreli limits quantum’s edge to factoring and elliptic-curve logarithms; current machines lack quantum RAM.He calls industry spending speculative and expects analog optical computing to challenge Nvidia’s H100, B100, and B200 within 20 years.His pharmaceutical thesis favors mispriced discarded molecules; he calls AI drug discovery within five years “nonsense” and expects no AI-related drug in the next 10 years.

Martin Casado × Dhaval Adjodah × Owen He

RSI founders Dhaval Adjodah and Owen He report early signs of performance rising as resource use stays flat or falls, with the model itself as optimizee.They cite retraining GPT-2 in hours for about $100 on the same GPU and data, and target inference efficiency across the stack.Casado has not seen equivalent GPT-5 evidence, leaving scalability beyond fast-evaluation tasks unresolved.

Jason Yanowitz × Michael Weisz

Kamino is positioning itself as embedded lending infrastructure that turns tokenized assets into productive collateral while distributors retain customer ownership, with Binance Wallet users already depositing $100 million into its vaults.The model requires transparent, cash-generating assets and day-one liquidity; Kamino currently has only credit and liquidity, will avoid venture-fund lending for six months, and expects revenue from interest spread or net interest income sharing.

Molly O'Shea × Aravind Srinivas

AI-native SMBs could become a larger source of US GDP growth through greater capital and headcount efficiency.Revenue is up 5X year over year, while a reported one-person customer spends more than $1 million annually and makes $3 million through automation, supporting a recursive-usage thesis.The open question is whether hybrid local-cloud agents can make privacy, sovereignty, and long-context economics durable advantages without sacrificing feedback loops or frontier capability.