Signage with logo at headquarters of venture capital investment firm Sequoia Capital, on Sand Hill … More
At Sequoia Capital’s recent AI Ascent 2025 conference, the atmosphere vibrated with a mix of excitement and urgency that is typical for transformative technological moments. Partners Pat Grady, Sonya Huang, and Konstantine Buhler presented a compelling case for AI as not just another innovation cycle, but potentially the largest market opportunity in modern business history.
“The AI opportunity is absolutely massive,” declared the Sequoia partners during the event. When comparing it to the cloud transition, they highlighted “a starting point for AI services that’s at least an order of magnitude bigger than the software market was at the beginning of cloud adoption.”
Quantifying The AI Opportunity
To grasp AI’s potential scale, Sequoia frequently referenced the cloud computing revolution. Today’s cloud market generates approximately four hundred billion dollars in annual revenue. However, according to their analysis, AI’s addressable market could dwarf these figures.
What makes the AI opportunity particularly distinctive is its dual-pronged attack on existing profit centers: “Both services and software are under attack,” noted the speakers. “Companies typically start with software, evolve into co-pilots, and ultimately transition to autopilots—moving from selling tools into software budgets to selling outcomes directly into labor budgets.”
This progression represents a fundamental shift in how technology creates and captures value. Both the total addressable markets for software and labor are described as “up for grabs” due to AI. The potential endpoint ten or twenty years from now has a chance to be “absolutely massive,” according to the conference speakers.
The Acceleration Factor
What distinguishes the AI revolution from previous technological shifts is its unprecedented speed of adoption. When cloud computing began its ascent, considerable marketing efforts were required to generate awareness and adoption. By contrast, as noted in the presentation, the release of ChatGPT on November 30, 2022, “instantly captured the attention of the entire world.”
“The physics of distribution have fundamentally changed,” emphasized the Sequoia partners. This accelerated adoption stems from multiple factors. Social media platforms like Reddit and “the artist formerly known as Twitter” provide immediate distribution channels to between one point two billion and one point eight billion monthly active users.
More fundamentally, internet penetration has expanded from approximately two hundred million users during the early cloud era to five point six billion today—effectively connecting “every household and every business in the world.” This means that “when the starting gun went off, there were no barriers to adoption.”
These structural advantages create what the speakers described as a “tremendous sucking sound” in the market—an irresistible gravitational pull toward AI adoption that “absolutely swamps any volatility you see in the markets” from macroeconomic factors. The message to founders was unambiguous: “ignore it.” Instead, focus on the fact that if you don’t seize the opportunity, “somebody else will because nature hates a vacuum.”
The Battle For The Application Layer
Despite significant attention paid to foundation models, Sequoia’s analysis suggests the greatest value will accrue at the application layer—the customer-facing software built atop these foundational technologies.
“Looking at companies that achieved one billion dollars plus of revenue during past transitions, most were positioned at the application layer,” observed the speakers. “This pattern is expected to hold true for AI. The value is at the application layer.”
The past year has witnessed the emergence of AI’s first “killer apps,” including ChatGPT for general use, Harvey for legal applications, Glean for enterprise search, and Cursor for coding. Additional emerging companies mentioned include Sierra, Abridge, Listen Labs, and Open Evidence. The conviction expressed was strong that “the application layer is where the value will ultimately accrue.”
However, the application layer is also where the competitive landscape is intensifying. Foundation models, powered by advancements like test time compute and reasoning with tool use and inter-agent communication, are becoming increasingly capable and are starting to penetrate deep into the application layer.
From Hype To Real Value
The evolution of AI adoption has followed a pattern described in the conference. In 2023, data showed that AI-native applications suffered from “terrible engagement ratios,” with initial curiosity failing to translate into sustained usage. This dynamic has shifted dramatically, with daily-to-monthly active user ratios for applications like ChatGPT now approaching levels seen on platforms like Reddit.
“This is extremely good news,” noted a conference speaker. “It indicates that more people are deriving value from AI and integrating it into their daily lives.”
Beyond simple engagement metrics, AI is enabling increasingly sophisticated applications. Examples mentioned include creating accurate ad copy, visualizing educational concepts, and improving patient diagnoses.
The technological capability threshold has advanced particularly rapidly in certain domains. Voice generation technology has progressed from “almost there” to “fully crossing the uncanny valley,” prompting comparisons to the movie “Her” and suggesting a “her moment for voice” has arrived.
Meanwhile, coding assistance tools have achieved what the speakers called “screaming product market fit.” The release of models like Anthropic’s Claude 3.5 Sonnet triggered a “rapid vibe shift” in the coding landscape. People are using AI coding for increasingly impressive tasks, fundamentally changing the accessibility, speed, and economics of software creation for both experienced and novice engineers.
Building Sustainable AI Businesses
While opportunity abounds, building sustainable AI businesses presents unique challenges. Sequoia partners emphasized that “ninety-five percent of building an AI company is just building a company”—solving important problems, attracting talent, and executing effectively. However, the AI-specific five percent introduces critical considerations:
Durable Revenue
A significant concern is what the speakers termed “vibe revenue”—initial adoption driven by curiosity rather than genuine value creation. “Founders must inspect adoption, engagement, and retention metrics to understand what users are truly doing with the product,” they cautioned. “Avoid the delusion of real revenue when it’s just hype.”
Earning customer trust was highlighted as “paramount, potentially even more important than the product itself at this stage,” as trust ensures customers will believe in the company’s ability to improve the product over time.
Viable Margins
While current AI applications often struggle with high computational costs, the path to healthy margins exists. “The cost component—like cost per token—is expected to continue decreasing,” noted the speakers. Simultaneously, companies that successfully move “up the value chain from selling a tool to selling an outcome” can potentially increase their price points, improving overall unit economics.
Data Flywheels
Perhaps the most important competitive moat for AI businesses lies in data advantages. “Data flywheels are one of the best moats you can build,” emphasized the Sequoia partners. “However, having one isn’t enough; it must clearly tie to a specific business metric.” A data flywheel that doesn’t move a business metric was dismissed as “bullshit”—either non-existent or simply not mattering.
The Next Frontier: The Agent Economy
Looking ahead, conference speakers outlined a vision for AI’s next major evolution: the agent economy. This represents a progression from individual AI assistants to “machine networks,” now commonly called “agent swarms” capable of collaborating on complex tasks.
“In this economy, agents will do more than just communicate information,” predicted the speakers. “They will transfer resources, make transactions, keep track of each other, and understand trust and reliability.” Crucially, this economy is “all about humans.” The agents will work with people, and the people will work with the agents.
Realizing this vision requires overcoming significant technical challenges in three key areas:
- Persistent Identity: Creating agents with consistent personalities and understanding of users
- Seamless Communication Protocols: Developing standards for information exchange between agents
- Security: Building trust mechanisms in a landscape of automated interactions
The potential implications are profound. While the prediction of the “first one-person unicorn” hasn’t yet materialized, companies are already scaling “faster than ever before with fewer people than ever before.”
The combined effect of these technological and mindset shifts is described as “way more leverage with significantly less certainty.” While you can achieve more, you must effectively manage the inherent uncertainty and risks.
The Race Is On
The insights from Sequoia’s AI Ascent paint a picture of a technology transition that is here, moving rapidly, and presenting an enormous economic opportunity, measured potentially in trillions of dollars.
While the value primarily resides at the application layer, the competition is fierce. Success requires a customer-back approach, deep vertical focus, building robust moats like data flywheels tied to business metrics, securing durable revenue beyond hype, and earning customer trust.
As the Sequoia partners concluded: “There is a tremendous sucking sound in the market, and standing still is not an option. It’s a run-like-heck business right now, demanding maximum velocity all of the time.”
For business leaders across industries, the implications are clear: the AI race is not just underway—it’s accelerating.
This post was created with our nice and easy submission form. Create your post!
