The architecture of agentic communication will determine the extent to which generative AI democratizes access to economic opportunity.
https://cacm.acm.org/opinion/the-agentic-economy/
Generative AI has revolutionized the way we interact with technology, allowing people to express their intent in free-form natural language. It has paved the way for AI agents that not only converse with users but also perform actions on their behalf, flexibly and with minimal guidance. Delegation to AI has already begun to improve the efficiency of individual processes, making both consumers and businesses more productive in the set of tasks they had already been doing. However, we believe that the more disruptive—and yet to be realized—impact of generative AI is its potential to drastically reduce the communication frictions between and among consumers and businesses. This could lead to a reorganization of markets, shifts of market power, and the introduction of entirely new products and services.
Consumers have traditionally faced high communication costs when initiating relationships with businesses, reducing efficiency. For example, a consumer seeking a new tax preparer might hesitate to switch because she would have to explain her financial situation all over again to a new person or online service. These communication hurdles can prevent consumers from taking advantage of better products and services or lower prices. Businesses have tried to lower these costs with tools like online forms and voicemail menus, but these often just shift communication costs to the consumer and can make interactions more rigid.
Imagine instead a future where every consumer has an assistant agent to communicate their preferences and personal information to businesses, and every business has service agents to interact with consumers and other businesses. These agents could be designed to interface with each other seamlessly and flexibly, transforming the landscape of consumer-business interactions. Delegating interactions to such assistant and service agents lowers communication costs and makes markets more efficient by expanding the range of options available to both consumers and businesses.
To unlock the full economic potential of generative AI’s communication capabilities, two developments are necessary. First, consumers and businesses must widely adopt assistant and service agents. This is already under way. Second, these agents must be designed to interact seamlessly with each other to facilitate transactions. On the technical front, there has been significant progress in standardizing such agentic interaction, with frameworks such as Microsoft’s AutoGen, and protocols such as Anthropic’s Model Context Protocol and Google’s Agent2Agent Protocol. However, it remains to be seen how these advances will be adopted and implemented, or constrained, given their complex interplay with and dependence on market forces.
snip
The Financial vs. Technological Economy
My thoughts on the AI bubble
Sinéad Bovell
https://sineadbovell.substack.com/p/the-financial-vs-technological-economy
The Two Economies Inside the AI Boom
On Wednesday, Nvidia ( the company at the heart of the AI infrastructure buildout) reported earnings that beat expectations. Revenue was strong, data center demand remained robust, and the company’s position at the center of the AI buildout was reaffirmed. The stock initially rallied in after-hours trading. Then it reversed and sold off sharply. On the surface, this doesn’t make sense. Strong results should lead to optimism about the future. But that’s not what happened. Instead, good earnings seemed to make investors more nervous, not less. Here’s what I think is actually going on: we’re watching two economies operate inside one market, and they’re moving on completely different timeframes.
What the Market Is Pricing
Stock prices reflect belief about future earnings. Right now, current earnings are strong, but belief in future earnings is weakening. That’s unusual. Normally when a company beats expectations, the market becomes more confident about what comes next. But with Nvidia—and really, across the entire AI infrastructure stack—something else is happening. The market is rightfully questioning whether this pace of investment can continue without interruption. It’s pricing in the risk of a slowdown, a digestion period, maybe even a full correction.
What the Technology Is Doing
But then there is the other economy, the technological one. And it’s operating on a completely different clock. If AI is a general purpose technology—and I believe it is—then we’re not looking at a two-year story. We’re looking at a 10-to-20-year transformation. General purpose technologies reshape economies slowly, then all at once. Electricity took nearly two decades before factories were redesigned from the ground up. The internet took almost 15 years before it became indispensable.
Two Timeframes, One Market
The financial economy is focused on quarters and fiscal years. It’s asking: when does spending slow? When do returns materialize? When does the infrastructure buildout hit its natural limit? These are legitimate questions. The technological economy is focused much further into the future. Companies are asking: what happens when this infrastructure is fully utilized? What happens when adoption moves from 10 percent to 50 percent? What does the world look like when AI is embedded in every workflow, every product, every decision? Both can be true. The market might be right about the next 12 to 24 months. And the technology economy might be right about the next 15 years. But why does this nuance matter for us? In our latest episode of
I’ve Got Questions, I break down the bubble indicators I’m watching, where the risks are concentrated, and why—even if we see a major correction—this is exactly the moment to pay more attention to AI, not less.
snip
= new reply since forum marked as read
