🏦 How a Wall Street megabank is thinking about AI

“This feels very different from previous tech bubbles.”

Quote of the Issue

“Cognitive enhancement could accelerate science and technology, including progress toward more potent forms of biological intelligence amplification and machine intelligence. Consider how the rate of progress in the field of artificial intelligence would change in a world where Average Joe is an intellectual peer of Alan Turing or John von Neumann, and where millions of people tower far above any intellectual giant of the past.”- Nick Bostrom, Superintelligence

---

The Essay

🏦 How a Wall Street megabank is thinking about AI

We would all love to know if generative artificial intelligence, or GenAI, will turn out to be mere bubbly fizzle, a historic general-purpose technology,p or somewhere in between. Is it more like the metaverse and Google Glass or more like the internal combustion engine? (I guess the middle ground would range from something like 3-D printing on the lower end to blockchain in the middle-middle to the PC/internet combo on the very high-end as a significant GPT that falls short of the ICE or electrification.) Indeed, given the tremendous buzz, even “somewhere in between” would mean we’re looking at pretty important new technology.

Just how important is the macroquestion explored in a new Goldman Sachs research compendium, “Generative AI: hype, or truly transformative?” (Some of the econ aspects I’ve written about before.) Lots of super interesting outside expert interviews and internal economic/financial analysis that I thought would make an interesting share. Here are some gleaned insights into various microquestions that get at that all-important macroquestion.

1/ Are we seeing a classic technology hype cycle? New technologies often generate extreme excitement followed by deep disillusionment. Boom, then bust. Aspects of this phenomenon have been captured by the Gartner Hype Cycle and the technological cycles scholarship of economist Carlota Perez. In the GS report, the banks’ software and internet research analysts doubt whether what we’re currently seeing with GenAI qualifies as part of a classic hype cycle. Some key points: 

• The current AI boom is being driven by established and influential tech companies rather than startups as was the case during the 1990s internet bubble. That difference reduces the risk of the boom fading away or taking a long time to gain momentum, according to GS. The role of big companies also reduces regulatory risk. Yes, regulation can slow innovation, and “regulated industries usually have lower profit margins, but the barriers to entry are higher given the cost of complying with regulation. So, while investor s may have to accept lower profits, they also don’t have to worry as much that the companies they’re investing in will be disrupted by new entrants.”

• Unlike past technological shifts, such as the transition from mainframe computers to distributed systems and from distributed to cloud computing, AI has gained widespread support from technology providers. (“When a unanimous verdict exists among the technology providers that a technological shift is actually happening, it’s real.”)

• Genuine interest from business customers, including global corporations, further supports the notion that AI isn’t just a fad. GS discussions with chief information officers reveal their fascination with the potential productivity benefits that AI can offer if implemented internally. ( “And all of this is occurring at a time when the market is rewarding productivity gains. So, this doesn’t feel like a hype cycle.”)

• While it's impossible to know for certain whether a bubble exists until it bursts, the majority of companies performing well in the AI sector have relatively reasonable valuations when considering key financial metrics, such as multiples to GAAP EPS. (“Bubbles are typically about enterprise value to eyeballs/clicks, addressable market dynamics, or sheer euphoria as a driver of valuations as opposed to what the right multiple on net income is to pay. So, this feels very different from previous tech bubbles.”)

2/ Assuming something real and important (or even amazing) is happening, how long before GenAI has a substantial impact on businesses, workers, and consumers?