Overview of Goldman Sachs CEO David Solomon on Running a Bank in the Age of AI
In this Bloomberg Odd Lots conversation, Goldman Sachs CEO David Solomon argues that AI will materially change banking and white-collar work, but not in the apocalyptic way some predict. He is broadly bullish on AI’s productivity gains, sees it as a force that will improve enterprise efficiency and economic growth, and believes human relationships, judgment, and emotional intelligence will remain central to finance. The discussion also covers Goldman’s use of AI, the future of junior banking jobs, capital markets trends, the rise of private markets, the state of equity valuations, and even AI’s impact on music creation.
Main Themes
AI will reshape work, but not eliminate the need for people
Solomon rejects the idea that AI will trigger mass unemployment or a “white-collar wipeout.” His core view:
- Technology always displaces some jobs, but it also creates new ones and changes how work is done.
- AI should raise productivity across large enterprises like Goldman Sachs.
- The most important long-term outcome is more economic growth and broader prosperity.
He emphasizes a long time horizon: the real changes will play out over five to 10 years, not overnight.
Human skills still matter a lot
A recurring point in the interview is that banking is still a relationship business.
- Solomon says cold calling, picking up the phone, and building trust remain powerful skills.
- He argues that a phone call has far more impact than a text or email.
- For client-facing roles, interpersonal judgment, EQ, and trust-building are hard to automate.
His view is that AI will amplify human capability, not replace the human layer that matters in high-trust businesses.
AI at Goldman Sachs
Where AI is already useful
Goldman is applying AI to internal workflows and operational processes, especially where there is structured data and repeatable work.
Examples discussed:
- Client onboarding
- Anti-money laundering and KYC processes
- Operational workflow automation
- Research and data analysis
Solomon says these are areas where productivity gains can be measured clearly.
Clean data matters
He stresses that AI is only as good as the data it is trained on:
- Clean, internal datasets are where models become extremely powerful.
- Open-web or social-media sourced data can create “garbage in, garbage out” problems.
- He gives a personal example of asking a model about Masters winners and getting an incorrect answer until challenged.
His broader point: AI is strong at synthesis, but human judgment is still needed to catch mistakes.
Goldman’s institutional advantage
Solomon suggests Goldman is well positioned because:
- The firm has a long-running culture of sharing and collaboration.
- It has unusually deep historical data, especially in trading.
- Its proprietary systems, like SecDB, give it a long data history others may not have.
Jobs, Hiring, and Junior Bankers
What changes for entry-level workers
Solomon acknowledges that AI will reduce some of the grunt work traditionally done by junior bankers and analysts, especially tasks like:
- Building presentations
- Compiling data
- Repetitive administrative work
But he does not think junior roles disappear. Instead, he expects:
- Slightly fewer hires over time
- More leverage per employee
- A need to rethink apprenticeship and training
Goldman still hires broadly
He notes Goldman brings in thousands of interns and new hires across:
- Client-facing roles
- Operations
- Engineering and technology
- Marketing and other functions
His concern is not job elimination, but how to teach people foundational skills when they no longer have to “work as hard to get the answer.”
Capital Markets, Private Companies, and IPOs
More companies may go public sooner or raise equity
Solomon says the market structure has changed:
- Private capital is more abundant than in past decades.
- Companies can stay private longer.
- But some of the biggest AI-era companies will still need public capital because their infrastructure needs are enormous.
He expects:
- More equity raises
- More large-scale IPOs
- Fewer public companies overall, but larger and more durable ones
This does not “break capitalism”
He rejects the idea that the rise of private markets undermines capitalism. His view:
- Public and private markets are just evolving.
- Companies go public when they need capital, currency, or liquidity.
- The current structure makes IPOs less attractive unless they are necessary.
Goldman’s role in mega deals
Solomon highlights Goldman’s role in large transactions, including major equity offerings and expected IPOs, noting that relationship-building over decades matters more than any one pitch or message.
Market View: Greed, FOMO, and Valuations
The market feels crowded, but not irrationally so
Solomon says the current market has a “greed” and “fear of missing out” dynamic, especially around AI-related names.
He compares the current environment to:
- The 1920s
- The 1960s
- The late 1990s
But he argues it is not identical to the dot-com bubble because:
- The top stocks have real earnings and cash flow
- Valuations are high but not as extreme as late-1990s internet stocks
- The broader market still trades at more reasonable multiples
His takeaway: the rally may continue because it is narrow and earnings are real, though macro risks remain.
Cybersecurity and Systemic Risk
Solomon also discusses AI and cyber risk:
- Powerful models create new vulnerabilities.
- Large financial institutions are investing heavily in cybersecurity.
- Smaller and midsize firms may be more exposed.
He warns that a cyber incident at a midsize bank could spread fear through the system, similar to how SVB’s collapse affected broader sentiment. He argues that both the private sector and government need to coordinate on resilience.
Music, Creativity, and AI
In a surprisingly detailed segment, Solomon talks about using AI in music production.
His core view on AI-generated creativity
- AI can democratize music production and lower barriers to entry.
- It enables people without traditional studio access to create music.
- But it raises major IP and compensation questions.
Why human voice still matters
He argues that:
- AI-generated output is not truly “your voice” unless it reflects your lived experience and judgment.
- Creativity still depends on human taste, emotion, and refinement.
- AI can accelerate the process, but human curation gives it meaning.
Notable Quotes and Ideas
- “Never different this time” — Solomon’s warning against assuming AI will overturn the basic structure of labor markets overnight.
- “The telephone is one of the greatest pieces of technology in the world, use it.”
- “Garbage in, garbage out” — his description of why clean data is essential for good AI output.
- He sees AI as a tool that will make human relationships, writing, speaking, and trust-building more valuable, not less.
Bottom Line
David Solomon’s message is optimistic but practical: AI will meaningfully improve productivity, change hiring patterns, and streamline operations at Goldman Sachs, but it will not erase the human side of finance. In his view, the future belongs to firms that combine strong data, strong technology, and strong people skills.