đź”´ LIVE From TomorrowX | David Solomon, Brian Armstrong, Ron Biscardi, Gavin Baker, Alex Davis, Guy Oseary, Andrew Feldman, Keller Cliffton, Chantel Jeffries, Cornelius Schmitt, Chase Lochmiller
đź”´ LIVE From TomorrowX | David Solomon, Brian Armstrong, Ron Biscardi, Gavin Baker, Alex Davis, Guy Oseary, Andrew Feldman, Keller Cliffton, Chantel Jeffries, Cornelius Schmitt, Chase Lochmiller
2 hours ago•TBPN•John Coogan & Jordi Hays
Podcast2 hr 50 min
Listen to Episode
Note: AI-generated summary based on third-party content. Not financial advice. Read more.
Quick Insights
  • Monitor Goldman Sachs (GS) for evidence that its durable revenue can grow 8%–9% annually, particularly through asset and wealth management and lending.
  • Treat NVIDIA (NVDA) as a volatility watch: leadership and AI-investment policy remain uncertain, so track board decisions and continued support for AI startups.
  • For AI exposure, prioritize the broader themes of cybersecurity and AI applications, while checking whether customer demand translates into revenue; the discussion provides no specific stock targets.
  • Consider private AI offerings such as Anthropic only after reviewing IPO pricing and terms—the discussion gives no valuation or basis for judging an entry price.
Detailed Analysis

NVIDIA (NVDA)

  • The discussion centered on the uncertainty around CEO Jensen Huang’s possible return. Gavin Baker said he had not heard from a shareholder, employee, or corporate partner who opposed Huang returning, and argued that his relationships and deal-making role would be difficult to replace.
  • David Solomon urged the board to move carefully, noting that there was not enough public information to judge the situation. He also cautioned against assuming how NVIDIA’s interim leadership would handle the company’s investments.
  • Baker highlighted NVIDIA’s support for open-weight AI models, arguing that a broader model ecosystem could help avoid dependence on a small number of closed providers.
  • The speakers noted that NVIDIA has been an important source of capital for AI startups. A reduction in that support could affect companies that rely on very large funding rounds.

Takeaways

  • Treat leadership and investment-policy uncertainty as a potential source of volatility; the transcript does not establish what the board will decide.
  • For investors in AI suppliers and startups, track whether NVIDIA continues to fund companies and support open models.
  • Baker’s view was strongly bullish on Huang’s importance, but Solomon emphasized that the available facts were incomplete.

Goldman Sachs (GS)

  • Solomon described a strategic shift toward more predictable revenue. He said roughly 60% of Goldman’s revenue now comes from what the firm considers more durable sources, compared with a lower share eight years earlier.
  • Asset and wealth management is a major part of that effort. Goldman supervises a little over $4 trillion in assets, and Solomon said the durable-revenue base could grow by about 8%–9% annually.
  • Goldman has also expanded lending to clients, which Solomon characterized as a more durable revenue source than some of the firm’s traditionally transactional businesses.
  • He said Goldman expects to remain focused on institutions, governments, and very wealthy individuals rather than broadly expanding into direct retail wealth management.
  • Solomon identified cyber risk as a major concern, including the possibility of a significant incident that could undermine confidence or slow economic activity.

Takeaways

  • The discussion points to a strategy of balancing Goldman’s cyclical trading and investment-banking businesses with fee-based asset management and lending.
  • Investors can monitor whether the firm delivers on its stated durable-revenue growth goals, while keeping in mind Solomon’s explicit concern about cyber risk.

Anthropic (Private; IPO Discussed)

  • Guy Oseary said he invested in Anthropic at an approximately $5 billion valuation and described his early investment thesis as based on the potential of the technology and the talent involved.
  • He described an Anthropic IPO as imminent, saying it was expected “in two days.” The transcript provides no IPO price, valuation, or offering terms.
  • Gavin Baker said AI’s biggest long-term value creation may come from applications built over the next couple of decades, rather than only from infrastructure, foundation models, or hyperscalers.
  • The discussion also raised the possibility that a concentrated group of model providers could limit the bargaining power of companies that supply or build on top of them.

Takeaways

  • The speakers see strong long-term potential in leading AI labs and in applications built on their models, but the transcript gives no basis for assessing an IPO valuation.
  • For investors, distinguish the potential of the technology from the price and terms at which shares may become available.

OpenAI (Private)

  • Oseary said he invested in OpenAI at an approximately $27 billion valuation in early 2023 and described ChatGPT’s initial launch as an unusually compelling product moment.
  • He said many investors had concerns about the company and its structure at the time, but that he relied on conviction about the technology rather than those objections.
  • The discussion presented OpenAI and Anthropic as examples of companies that attracted early investment because investors saw the potential of large language models before the opportunity was widely accepted.

Takeaways

  • The account illustrates the potential rewards—and uncertainty—of investing early in transformative technology. It does not provide a current valuation or a specific recommendation to invest.

Cerebras (Private)

  • Gavin Baker said his firm is an investor in Cerebras and described its chips as particularly well suited to fast AI inference.
  • Cerebras CEO Andrew Feldman said the company’s system provides substantially faster inference than GPUs, claiming it can be 10–20 times faster. He said demand was very high but constrained by the availability of data centers.
  • Baker and Feldman discussed the potential use of fast inference and ensembles of specialized models in cybersecurity, where responding quickly can matter.
  • Feldman said Cerebras is pursuing a new hardware release about every year, with the goal of roughly doubling performance each time.
  • Feldman said about 75% of Cerebras’s data centers in 2027 are expected to be outside the United States, citing U.S. permitting constraints. He also said smaller sites can work well for inference, while very large campuses are more suited to training.

Takeaways

  • Cerebras represents an investment theme in specialized AI chips and inference capacity, but the company’s own comments suggest data-center availability is a constraint on growth.
  • Watch whether faster inference produces sustained customer demand and whether Cerebras can expand capacity without relying on the largest U.S. data-center projects.

Crusoe (Private)

  • Crusoe described itself as an AI infrastructure company spanning data centers, managed GPU clusters, and AI services such as inference and fine-tuning.
  • CEO Chase Lochmiller said customers are interested in fine-tuning models on proprietary data, particularly for specific tasks where a customized model may outperform a general-purpose one.
  • The company is pursuing projects in multiple regions, including Iceland, Paraguay, and Norway. Lochmiller cited a 280-megawatt project in Paraguay and discussed Iceland’s geothermal energy and cooling advantages.
  • Crusoe is also developing modular, manufactured data centers that can be deployed in smaller increments near available power. The company argues that this could reduce construction disruption and make use of power that might otherwise go unused.
  • Lochmiller said large training campuses and smaller inference facilities serve different needs, and argued that distributing AI infrastructure could improve resilience.

Takeaways

  • The discussion highlights opportunities in AI data-center construction, energy access, and private-data fine-tuning.
  • Important execution factors include securing power, permits, customers, and reliable construction capacity. The transcript also notes community resistance and permitting delays as real challenges.

Coinbase (COIN)

  • Brian Armstrong said Coinbase is focused on “agentic finance”—AI agents using financial accounts and payment rails to carry out tasks.
  • He said Coinbase’s product for connecting agents to accounts was growing about 20% week over week at the time of the interview. Examples included trading, managing portfolio allocations, tax-aware decisions, and placing limit orders.
  • Armstrong said Coinbase is adding controls and permissions so agents can operate within defined limits, rather than having unrestricted access to funds.
  • He also cited Coinbase’s security measures, including cold storage for more than 99% of funds, and said the company had received approval for a national trust charter from the OCC.
  • Armstrong said Coinbase had acquired roughly 10 companies the previous year and was continuing to evaluate acquisitions, particularly in agentic finance.

Takeaways

  • Coinbase is positioning itself at the intersection of crypto, financial services, and AI automation. Investors can monitor whether agent use translates into sustained customer activity and revenue.
  • Security and safeguards for autonomous financial activity remain central to the investment case, as Armstrong himself emphasized.

Bitcoin (BTC)

  • Armstrong said he was not yet concerned that AI-related mathematical research had made Bitcoin’s current cryptography vulnerable.
  • He did say that quantum computing could eventually be accelerated by AI and that the industry should prepare early. Coinbase is participating in a community effort to develop a path toward post-quantum cryptography, including discussion of Bitcoin improvement proposals.

Takeaways

  • The discussion frames quantum computing as a long-term preparedness issue, not an imminent threat to Bitcoin.
  • Investors can monitor progress on post-quantum upgrades and avoid treating speculative timelines as evidence of an immediate change to Bitcoin’s security.

CrowdStrike (CRWD) and Cognition (Private)

  • Baker said Cerebras has a partnership with CrowdStrike and that Atreides, his firm, is also an investor in Cognition.
  • Baker described Cognition’s security product as its fastest-growing product to date and said it uses a collection of models to address security tasks.
  • The discussion’s broader view was that AI can be used both to generate attacks and to strengthen defenses. Baker argued that fast inference and multiple specialized models could help defenders respond more quickly.

Takeaways

  • Cybersecurity is a prominent investment theme in the transcript, particularly products that use AI to detect and respond to threats.
  • The discussion does not give CrowdStrike’s valuation or a specific stock recommendation; the positive comments about Cognition are Baker’s view, not a performance forecast.

New Limit (Private; Longevity Biotech)

  • Armstrong described New Limit, the longevity biotech he co-founded, as using AI to generate and evaluate biological hypotheses, followed by testing in wet labs and other experimental settings.
  • He said the company’s approach is to narrow down promising candidates through successive stages of testing before moving to human trials.
  • Armstrong said New Limit expected its first human clinical trials to begin the following year. He also emphasized that understanding and simulating biology remains difficult, noting that accurately modeling a whole organism is far more complex than modeling a single protein.

Takeaways

  • The opportunity discussed is the use of AI to speed up biological research, particularly in longevity and cell reprogramming.
  • The transcript also makes clear that experimental validation and clinical trials remain major steps; the company’s claims should be judged against future human-trial results.

Zipline (Private)

  • Zipline was described as a drone-delivery company expanding in the United States after receiving FAA approval to operate beyond visual line of sight across all 50 states.
  • CEO Keller Cliffton said the company had recently launched in Austin and was scaling in Dallas. He described Zipline as supply-constrained, with some locations receiving more demand than available aircraft capacity.
  • Cliffton said Zipline had three strategic partners each seeking more than one million deliveries a day of capacity. He also said the company was working toward a manufacturing scale of one million deliveries per day.
  • Individual aircraft may make roughly 20–50 deliveries per day, depending on operating conditions. Zipline plans to extend operations toward 24-hour service and serves healthcare, food, and quick-commerce customers.
  • Cliffton discussed Zipline’s manufacturing complexity and said that hardware delays and reliability problems had been significant challenges. He credited investor Valor with providing manufacturing and engineering support.

Takeaways

  • Zipline offers exposure to drone delivery, logistics automation, and domestic hardware manufacturing, with the company describing demand as strong.
  • The main points to monitor are whether Zipline can scale manufacturing, maintain reliability, and convert partner demand into actual deliveries. The transcript emphasizes that hardware execution is difficult and capacity is not yet sufficient.

Zellerfeld (Private)

  • Zellerfeld is developing 3D-printed footwear and other products. Founder Cornelius Schmitt said the company had produced tens of thousands of pairs of a single design and was working toward capacity for about one million pairs annually.
  • Schmitt said conventional footwear product cycles can take 12–18 months and cost hundreds of thousands of dollars to develop, while Zellerfeld’s process can produce a design in roughly a day for about $50. These are the founder’s comparisons.
  • He said Zellerfeld had built its own printers because suitable equipment was difficult to source, and was operating around 1,700 printers in Austin.
  • The company is exploring a broader vision in which customers can download and print digital product designs, extending beyond shoes to other consumer goods.

Takeaways

  • The investment theme is additive manufacturing: shorter product-development cycles and more customized production.
  • The key question is whether Zellerfeld can scale production and demand beyond footwear while maintaining product quality and competitive economics.

AI Infrastructure, Cybersecurity, and Applications

  • Solomon and Baker both emphasized that AI may create opportunities well beyond chips and data centers. Solomon pointed to future applications in areas such as health, food, supply chains, and wellness; Baker argued that applications could become a major source of value creation over time.
  • Speakers also discussed cybersecurity as an area of rising importance. Solomon warned of the possibility of a serious cyber incident, while Baker and Feldman argued that AI-enabled defense could help organizations detect and respond to attacks more quickly.
  • A recurring concern was the possibility of concentrated buyers or a small number of dominant model providers gaining leverage over suppliers and companies building on their platforms.

Takeaways

  • The transcript suggests looking beyond the most visible AI infrastructure companies to application software, security, and domain-specific models.
  • At the same time, investors should consider concentration risk, dependence on external platforms, cyber resilience, and the practical limits of data-center capacity.

Lululemon (LULU)

  • Solomon recounted Goldman Sachs’s role in Lululemon’s IPO pitch, describing how the team wore the company’s apparel to show that it understood the brand.
  • He used the story to illustrate the importance of client relationships and trust in investment banking. He offered no view on Lululemon’s stock, valuation, or business outlook.

Takeaways

  • This was a story about Goldman’s deal-making approach, not an investment thesis on Lululemon. The transcript contains no price target or specific recommendation.
Ask about this postAnswers are grounded in this post's content.
Episode Description
(00:00) - Ron Biscardi discusses his role as CEO of iConnections and the launch of Tomorrow X, a major AI and technology summit in Austin. He explains how the event brings leading CEOs and institutional-grade investment insights to everyday investors, online creators, and attendees from around the world. (06:46) - David Solomon discusses his leadership of Goldman Sachs, emphasizing human relationships, trust, and judgment alongside technological innovation. He explains the firm’s shift toward durable revenues, institutional focus, AI adoption, and cybersecurity, while also sharing perspectives on IPOs, private markets, reindustrialization, and America’s economic strengths. (37:03) - Gavin Baker, founder and managing partner of technology investment firm Atreides Management, discusses Jensen Huang’s irreplaceable leadership at Nvidia, the strategic importance of open-weight AI models, and AI’s role in cybersecurity. He also highlights faster inference, defensive model ensembles, and AI-driven biotechnology as promising areas for investment and societal progress. (52:25) - Alex Davis discusses his work as an investor, CEO of Disruptive, and executive chairman of Grok. He outlines a new $10 billion fund targeting 10 capital-intensive companies and explores how shifting AI investment, potential down rounds, and reduced corporate funding could create attractive opportunities for venture investors. (01:01:03) - Guy Oseary, an entertainment executive and venture investor known for identifying talent early, discusses his instinct-driven approach to backing companies such as OpenAI and Anthropic. He also explains Get Access, a membership platform that connects promising companies with influential, value-adding investors from entertainment and other industries. (01:10:49) - Andrew Feldman discusses Cerebras’ wafer-scale AI chips, which deliver exceptionally fast inference for cybersecurity, financial markets, and other time-sensitive applications. The Cerebras co-founder and CEO also explores emerging AI architectures, competition with Nvidia, rapid hardware development, surging global demand, and the challenges of expanding data-center infrastructure. (01:32:11) - Keller Cliffton, co-founder and CEO of autonomous drone-delivery company Zipline, discusses the company’s expansion, FAA approval, and full-stack approach to building aircraft, software, and delivery infrastructure. He describes soaring demand, manufacturing challenges, investor support, and Zipline’s potential to strengthen American industrial and technological leadership. (02:01:17) - Chantel Jeffries is a DJ, investor, and startup advisor who discusses her approach to angel investing, supporting founders, and using her network and cultural influence to help brands grow. She also shares her thoughts on celebrity partnerships, founder-led media, handling public criticism, and creating accessible investing content. (02:09:55) - Cornelius Schmitt discusses founding Zellerfeld, a 3D-printing company, in his German student dorm eight years ago. He outlines his vision of using software-driven printing to rapidly and affordably manufacture shoes, clothing, electronics, and eventually almost any product—from home printers or facilities worldwide. (02:21:23) - Brian Armstrong, co-founder and CEO of Coinbase, discusses the convergence of cryptocurrency and AI through agentic finance, specialized AI models, and secure payment infrastructure. He also covers Coinbase’s cybersecurity strategy, quantum-computing risks to Bitcoin, acquisitions, and his longevity biotech company NewLimit’s efforts to restore youthful cellular function. (02:36:36) - Chase Lochmiller discusses Crusoe’s global expansion and its vertically integrated AI infrastructure platform, spanning data centers, GPU clusters, and managed AI services. As Crusoe’s co-founder and CEO, he highlights renewable-powered projects, modular data centers, cybersecurity, and the benefits of distributing compute near low-cost, underused energy sources. TBPN is made possible by: Ramp - https://ramp.com Public - https://public.com Cisco - https://www.cisco.com Console - https://www.console.com CrowdStrike - https://www.crowdstrike.com Figma - https://www.figma.com MongoDB - https://www.mongodb.com NYSE - https://www.nyse.com Railway - https://railway.com Shopify - https://www.shopify.com Codex - http://openAI.com/codex Lambda - https://lambda.com Follow TBPN:  https://TBPN.com https://x.com/tbpn https://open.spotify.com/show/2L6WMqY3GUPCGBD0dX6p00?si=674252d53acf4231 https://podcasts.apple.com/us/podcast/tbpn/id1772360235 https://www.youtube.com/@TBPNLive
About TBPN
TBPN

TBPN

By John Coogan & Jordi Hays

Technology's daily show (formerly the Technology Brothers Podcast). Streaming live on X and YouTube from 11 - 2 PM PST Monday - Friday. Available on X, Apple, Spotify, and YouTube.