Latest Interviews
Showing 1–11 of 11 transcripts.
Clear all filters- Goldman Sachs33 min
The Outlook for Data Center Power Demand as AI Token Use Grows
Brian Singer, Carly Davenport, Allison Nathan, George Lee
Goldman Sachs forecasts a surge in U.S. power demand reaching a 3.5% CAGR through 2030, driven by a 170% rise in global data center consumption that will equal Japan's total electricity usage. This growth is propelled by hyperscaler spending hitting $2.1 trillion by 2029, though projects face critical bottlenecks from skilled labor shortages, interconnect queues, and over 300 local moratoria. Consequently, the sector is increasingly relying on 30 gigawatts of behind-the-meter natural gas generation as a bridge while navigating regulatory uncertainty and community opposition regarding grid stability and environmental impacts.
- Goldman Sachs29 min
Building AI Systems for Capital Markets
Chris Churchman, Allison Nathan, George Lee
Marquee, Goldman Sachs' institutional digital platform, integrates Gen AI directly into investment workflows to transform complex data into auditable, natural language insights while preserving human reasoning capabilities. By transitioning through stages of engineering focused on environment and mandate definition, the platform addresses critical challenges like hallucinations and cognitive atrophy to ensure reliability under uncertainty. This approach enables a shift from automating legacy processes to reconceptualizing decision-making, ultimately safeguarding the firm's intellectual capital and trust with clients.
- Goldman Sachs26 min
The AI Investment Boom: When Will It Pay Off?
Jim Covello, Alison Nathan, George Lee
In a Goldman Sachs Exchanges episode recorded on May 26, 2026, Jim Cabello and George Lee analyze the divergence between surging AI technology and unproven economic returns. Cabello admits previous misjudgments regarding consumer adoption and hyperscaler spending patterns while warning that the current exclusive profitability of semiconductor firms creates an unsustainable market dynamic. The dialogue concludes that without evidence of enterprise-scale profitability within the next two years, the industry faces a potential correction where hyperscaler stocks are poised to outperform if capital expenditure moderates or profit distribution shifts upstream.
- Goldman Sachs21 min
AI Exchanges: Power Problems?
Brian Singer, Alison Nathan, George Lee
Goldman Sachs projects a 220% surge in global data center power demand by 2030 as hyperscalers commit over $300 billion to infrastructure, a trajectory comparable to adding the world's sixth-largest power consumer to the grid. This aggressive expansion faces severe bottlenecks in skilled labor and transmission supply, forcing a near-term reliance on natural gas simple-cycle generators while industry leaders navigate regulatory hurdles and "agentic" AI traffic growth. Despite the high capital intensity and a potential $40 per megawatt-hour premium for green energy, major technology firms maintain robust balance sheets to fund these transition costs while seeking to decouple their power expenses from consumer utility rates.
- Goldman Sachs26 min
AI Exchanges: AI’s Impact on Employment
Joseph Briggs, Allison Nathan, George Lee
While current U.S. corporate AI adoption hovers at 9% with higher penetration in large firms, the technology is already reshaping the labor market by increasing specialized engineering job postings by up to 50% while disproportionately displacing recent college graduates and junior roles. Although long-term forecasts suggest minimal structural unemployment due to historical technology-driven growth, rapid adoption within one to three years could trigger a 2% to 2.5% unemployment spike, whereas a gradual ten-year transition would keep impacts manageable. Strategic resilience depends on the sector's reliance on high-stakes decision-making or human interaction, yet the shift forces a fundamental challenge to traditional apprenticeship models and introduces hybrid workforce structures where managers oversee both humans and autonomous agents.
- Goldman Sachs23 min
AI Exchanges: Where Will the Power Come From?
Rebecca Kruger, Allison Nathan, George Lee
Hyperscalers and utilities are forging unprecedented partnerships to address the binding power constraint of AI infrastructure, as the sector's step-change demand outstrips the current grid's 40-year-old capacity and supply chain limitations. With nuclear and gas-fired assets facing severe construction bottlenecks and a critical mismatch between rapid data center deployment and decade-long power generation timelines, these entities are committing to billions in capital and ten-year contracts to ensure reliable baseload generation. This strategic realignment, supported by emergency permitting measures and advanced technologies like behind-the-meter generation, marks a decisive shift from two decades of flat energy growth to a new era of massive grid modernization and investment.
- Goldman Sachs22 min
AI Exchanges: How tech giants are navigating the AI landscape
Eric Sheridan, Allison Nathan, George Lee
Major technology giants are sustaining record capital expenditures through 2025 to solidify AI infrastructure dominance, with Goldman Sachs projecting this peak intensity to slow in 2026 despite tariff-related cost increases. Executives from Meta, Alphabet, and Amazon characterize the sector as being in the early innings of a third computing shift, prioritizing defensive positioning and infrastructure monetization over immediate application-layer profitability. Although investor sentiment remains bifurcated by macroeconomic risks, current valuations reflect a strategic reliance on these incumbents' unprecedented balance sheet scale to navigate the transition from raw training to commercial deployment.
- Goldman Sachs29 min
AI Exchanges: CIO Marco Argenti on the future of AI in the workplace
Marco Argenti, George Lee, Allison Nathan
A Goldman Sachs-led panel analyzes the rapid acceleration of AI reasoning models against the persistent "value gap" created by legacy systems and the behavioral friction of retraining human workflows. While "de novo" enterprises advance faster due to agile structures, the firm is actively cultivating "mindful disruptors" and embedding specific cultural judgment into agents to manage risks like hallucinations and data exfiltration. The dialogue concludes that future success relies on a hybrid workforce where AI agents provide elastic capacity and mechanical execution, leaving high-level creativity and professional oversight to human leaders.
- Goldman Sachs22 min
AI Exchanges: Will falling costs drive new opportunities?
Allison Nathan, George Lee, Kim Posnett
Host George Lee and guests analyze how DeepSeek's low-cost AI model challenges the economic viability of massive U.S. tech capital expenditures while triggering a shift in consensus from data scarcity to power availability as the primary industry constraint. Despite short-term market volatility, the panel highlights record CapEx commitments from hyperscalers and emerging enterprise adoption trends, including Goldman Sachs' internal AI tools and synthetic data markets, which collectively suggest a rapid transition toward autonomous AI agents and multi-gigawatt infrastructure needs. The discussion concludes with a bullish long-term outlook, predicting that declining computing costs will fuel new use cases and accelerate workforce productivity even as the sector navigates immediate supply chain bottlenecks.
- Goldman Sachs28 min
Navigating the new geopolitical and technology landscape
George Lee, Jared Cohen, Allison Nathan
High geopolitical uncertainty and the rise of "swing states" are forcing a shift from broad regional strategies to country-specific tactics as nations leverage distinct advantages in technology and supply chains. With geopolitical drivers now overpowering traditional economic interests, major powers are engaging in asymmetric competition over critical technologies like semiconductors and AI, creating an "impedance mismatch" between slow government regulation and rapid technological evolution. To navigate this landscape, leaders must integrate deep geopolitical analysis with commercial insights to mitigate risks and capitalize on the emerging "geocommercial" opportunities across these fragmented global markets.
- Goldman Sachs28 min
Jeff Lawson, Co-founder and CEO of Twilio
Twilio co-founder Jeff Lawson argues that reducing the cost and complexity of software experimentation accelerates innovation, citing the Wright Brothers' iterative success and the unique global reach of modern developers as evidence. Drawing from his experiences at Amazon Web Services, Lawson advocates for a strategic shift where executives treat developers as partners rather than coders to build "Minimum Lovable Products" within small, mission-aligned teams. This approach underpins Twilio's evolution from a communications API into a comprehensive Customer Engagement Platform, aiming to democratize the personalization capabilities of digital giants like Amazon and Apple for all businesses.