Goldman Sachs Chairman and CEO David Solomon on AI: Implementation, Accountability, and Jobs

David Solomon · Goldman Sachs

Compute demand currently exceeds supply, but the size of the buildout is unknown and will be uneven.

Solomon says there is more demand for compute today than there is compute, and an expectation of a very large buildout. He argues nobody really knows whether the total spend will be $5 trillion, $10 trillion or more. He expects the buildout to be a journey with accelerations and pullbacks, not a straight line. He argues investors are trying to judge the pace of development and where the best opportunities lie, which drives volatility. He says markets will likely recalibrate at some point as participants reassess how much compute demand there will be, how fast it grows, what infrastructure is needed, and how faster, cheaper chips change data center requirements.

“there's more demand for compute right now than there is compute, and there is an expectation of a very significant build out. Whether or not the build out is gonna be as significant as some people are saying ... don't think anybody really knows. It will be a journey. It won't be a straight line.”at 14:01

David Solomon · Goldman Sachs

Rising capital costs and interest rates are pressuring AI business models and reshaping investor views.

Solomon ties recent market volatility around AI partly to financing conditions. He gives two mechanisms. First, lots of capital flowing into AI creates competition for that capital, which pushes up its cost. Second, interest rates rose quite a bit over the summer for other reasons. He says both changes shift investors' views and put pressure on some AI business models, so the financing environment is a factor in how the AI build-out is priced.

“Certainly as the cost of capital goes up, it changes people's views ... interest rates have gone up quite a bit this summer for other reasons, and those things push at some of these business models”at 14:35

David Solomon · Goldman Sachs

AI needs rules and clear accountability across government, model builders and deploying businesses, not a binary choice.

Solomon argues the AI debate is wrongly framed as either fast innovation with no guardrails or heavy restrictions that slow progress. He says innovation and guardrails can coexist, and that a structure of accountability is needed for deploying the technology. He splits responsibility three ways: government sets rules of the road; model and infrastructure developers must take responsibility that their systems work and do no harm; and businesses that embed AI in their products must answer for how it is deployed. He says the third layer gets too little attention. As evidence, he cites Goldman Sachs using AI-type models in trading platforms for almost twenty years, with a human accountable for every platform. He argues large financial institutions' risk, control, audit and testing practices can inform AI governance, while noting that other businesses must learn these controls as a new discipline.

“government needs to play a role and set some rules of the road. The people that are developing these models ... have to take responsibility for what they're building and that it works and that it doesn't do harm.”at 2:34

David Solomon · Goldman Sachs

AI automates tasks but mainly makes skilled workers more productive; new jobs should emerge over a decade or two.

Solomon accepts that AI will disrupt and automate work, and the open question is the pace. He argues the US economy is dynamic, creating and destroying about 30 million jobs a year and generating a net gain of about a million, with many jobs coming from small businesses. He says AI mostly makes smart people more productive rather than eliminating their roles, so white-collar work changes rather than disappears. He points to his own early career and to Goldman's analysts today, who do different work and are far more productive without being fewer in number. He expects new jobs to offset losses over the next decade or two. He flags a near-term risk: if the pace of change is fast, the gap between jobs lost and jobs created could widen, and policy action may be needed to soften the disruption.

“a lot of what this technology does is it makes smart people more productive. So when people talk about white collar jobs, just because the technology makes them more productive, it doesn't mean the job goes away.”at 7:20

David Solomon · Goldman Sachs

Falling AI costs make starting businesses cheaper, so small businesses need AI training to adopt the tools.

Solomon argues small businesses matter greatly to the economy. They are about 99% of US companies by count and drive 60% to 65% of annual US job creation. He says AI lowers the expense of building a business or platform, which gives more entrepreneurs the chance to start one. Those businesses still need coaching, education and experience deploying technology, even when small. To meet that need, Goldman committed $500 million to a Small Business AI Institute with leading AI labs. It extends a 17-year small business program that has served about 20,000 companies, adding AI training and more capital for participants.

“There's an expense to building a business, building a platform, setting it up. That expense is going down with this technology, so that gives more entrepreneurs the opportunity to start businesses.”at 12:28

David Solomon · Goldman Sachs

Not concerned about AI's long-term durability; the impact will take decades, with recent volatility being noise.

Asked whether recent market volatility around AI threatens its durability, Solomon says that with a long-term view he is not concerned. He expects AI to be a very important technology deployed across a large application layer in business, enterprise and government. He also says it will take ten, twenty or thirty years to fully appreciate its impact, so near-term market swings should be read against that timeline. He is confident the medium- and longer-term benefits will be substantial, while expecting the path to be uneven.

“With a long term view, I'm not concerned about its durability and sustainability. I think this is gonna be a very important technology that is going to be deployed in an enormous application layer in businesses and enterprise and government”at 13:39