How is the AI revolution different from the cloud revolution for enterprises?

While cloud was primarily an efficiency play, AI is seen as a competitive imperative. Companies that don't adopt AI risk falling behind due to productivity gains, faster innovation, and the ability to attract AI-native talent.

What is the future of pure 'model' companies in the AI space?

It's unlikely that pure play model companies will thrive solely by selling tokens. The future likely belongs to AI companies that offer significant surrounding software, security, compliance, and enterprise integration, effectively becoming software companies powered by AI models.

Will AI commoditize intelligence to zero cost?

The trend points towards the cost of intelligence (like AI tokens) converging towards zero, similar to compute and storage. This means startups need to focus on building software that leverages this cheap intelligence to solve real-world problems and deliver outcomes, rather than on the intelligence itself.

What's the difference between 'core' and 'context' AI investments for businesses?

Core AI investments are those critical to a company's unique value proposition and competitive advantage (e.g., drug discovery in life sciences). Contextual AI investments are enabling functions that are necessary but not unique (e.g., HR systems, clinical trial automation), and these are typically better bought as external solutions.

Are enterprises comfortable using hosted AI models like OpenAI or Anthropic?

Comfort levels are increasing, especially in industries that take security and compliance seriously. While some enterprises will maintain on-premise solutions, the maturation of AI platforms in terms of security and reliability is building trust for hosted models.

How will AI expand the Total Addressable Market (TAM) for software?

AI is expected to significantly expand software TAM by enabling companies to perform tasks previously impossible or too costly, rather than just digitizing existing labor. This creates entirely new use cases and budgets, potentially increasing the overall software market size by several times.

What is the ultimate utopian outcome of AI automation?

The utopian vision is that AI-driven automation will lower the cost of goods and services, leading to abundance and lifting lifestyles for everyone. This could enable better education, healthcare, and opportunities, particularly for underserved communities.

How AI Is Changing Enterprise

Y Combinator

Science & Technology4 min read50 min video

Feb 19, 2025|127,555 views|2,252|81

YC Y Combinator

Save to Pod

Key Moments

TL;DR

AI is transforming enterprise workflows, with value shifting from models to integrated software solutions.

Key Insights

Enterprise AI adoption focuses on outcomes, not just models, integrating AI into existing workflows.

The value of AI will increasingly come from software layers that abstract and utilize foundational models, not the models themselves.

Intelligence is becoming a commodity, driving down costs and creating opportunities for startups focused on vertical or horizontal AI solutions.

Startups can build moats by developing software with significant proprietary logic and workflow integration, rather than just wrapping AI models.

Enterprise adoption of AI is accelerating, driven by competitive necessity and the emerging AI-native workforce.

The TAM for software is expanding significantly due to AI, enabling a broader range of companies to leverage technology for new use cases and efficiencies.

ENTERPRISE FOCUS ON OUTCOMES, NOT MODELS

The core of AI adoption in enterprise settings is not the underlying artificial intelligence model but the tangible outcomes it delivers. Businesses require software that integrates seamlessly with their existing systems to achieve specific goals, such as automating customer support, processing contracts, or streamlining healthcare transcriptions. While models are becoming more intelligent, the real value for enterprises lies in the comprehensive software solutions that wrap these models, providing essential workflow, proprietary business logic, and data integration. This focus ensures that AI serves as a tool to achieve business objectives rather than an end in itself.

INTELLIGENCE AS A COMMODITY AND STARTUP OPPORTUNITIES

As AI models become more powerful and accessible, intelligence is rapidly transforming into a commodity. This trend drives down the cost of core AI capabilities, encouraging startups to look beyond simple model wrappers. The play for startups lies in building robust software layers that abstract the underlying models, offering specialized vertical or horizontal AI solutions. The key to success is creating software that leverages AI to solve complex, real-world problems, integrating with existing enterprise systems and providing unique value propositions that proprietary business logic and customer data enhance.

THE EVOLVING ROLE OF MODEL PROVIDERS

The landscape of AI model providers is shifting, with few remaining as pure-play model companies. Many are evolving into software companies that leverage their models to power applications for consumers and businesses. Revenue generation is increasingly tied to providing not just API access, but also security, compliance, governance, scalability, and dedicated customer support. Companies like OpenAI and Google are primarily seen as software businesses with underlying AI models, while meta benefits from open-sourcing its models. The challenge for new entrants is to build a comprehensive value proposition beyond just the model itself.

ACCELERATING ENTERPRISE ADOPTION AND THE AI-NATIVE WORKFORCE

Large enterprises are increasingly embracing AI, recognizing it as a critical competitive imperative rather than just an efficiency gain. This shift is partly driven by the emergence of an AI-native workforce, where new entrants to the job market are already accustomed to using AI tools for information retrieval and problem-solving. Companies that fail to adopt AI risk being unable to attract talent and falling behind competitors who can leverage AI for increased productivity, faster onboarding, and better client services. This competitive pressure is forcing even highly regulated industries, like banking, to actively explore and implement AI initiatives.

EXPANDING TOTAL ADDRESSABLE MARKET (TAM) THROUGH AI

AI is significantly expanding the total addressable market for software, akin to the transformation brought by cloud computing. Previously, complex software deployments were limited to large enterprises due to high costs and implementation times. AI, by democratizing advanced capabilities, allows a much broader range of businesses to access and utilize powerful software solutions. This expansion occurs because AI enables companies to perform tasks that were previously unfeasible or too expensive, leading to increased software spend as new use cases become viable and cost-effective.

REDPLOYMENT OF EFFICIENCIES AND CONSUMER BENEFITS

Contrary to a zero-sum view, AI-driven efficiencies are primarily additive rather than simply reallocating existing labor spend. Companies reinvest these gains into building better products, expanding services, and serving customers more effectively, fueled by competitive market dynamics. This reinvestment leads to faster revenue growth and innovation. Ultimately, the benefits of AI are expected to trickle down to consumers through improved products and services, fostering an era of abundance. The key lies in regulatory environments that allow these generated surpluses to translate into tangible societal improvements and lower costs across the board.

CONTEXT VS. CORE: STRATEGIC PRIORITIES FOR ENTERPRISES

Enterprises adopting AI must strategically differentiate between 'core' and 'context' functions. Core functions represent the unique value proposition and competitive differentiator of a business (e.g., drug discovery in life sciences), demanding in-house AI development and proprietary solutions. Context functions, while essential, are not unique differentiators and are better served by adopting external, standardized solutions (e.g., HR systems, CRM). This strategic focus prevents wasted resources on reinventing commodity functionalities and allows companies to concentrate AI efforts on areas that drive genuine competitive advantage.

THE ROLE OF OPEN SOURCE AND SECURITY IN ENTERPRISE AI

Open-source AI models are gaining traction in the enterprise, offering cost benefits and fostering a symbiotic relationship between developers and large organizations seeking support and managed versions. While initial concerns about data security with hosted models were significant, comfort levels are increasing as providers demonstrate robust security, compliance, and privacy measures, mirroring the evolution of cloud adoption. A segment of the market will continue to favor on-premise or enclave deployments, but the trend points towards greater acceptance of secure, integrated hosted AI solutions, especially as their enterprise-readiness matures.

Mentioned in This Episode

●Software & Apps

●Companies

●Concepts

●People Referenced

Common Questions

The 'wrapper' debate questions whether building apps on top of foundation models like ChatGPT offers significant long-term value, or if these applications risk being easily incorporated into the core models themselves. The consensus is that true value lies in the proprietary business logic, workflows, and data an application integrates, not just the AI's output.

Topics

Market Expansion

Mentioned in this video

mediaThe Light Cone

The name of the podcast series.

personDavid Solomon

CEO of Goldman Sachs, whose anecdote about AI writing an S1 in 10 minutes, a task previously requiring six people, is used as a powerful example of AI's impact.

conceptJeevan's Paradox

Mentioned as a potential driver for a future focused on abundance for everyone, contrasting with a 'Black Mirror' dystopian future.

personJeff Moore

Mentioned as the creator of a framework regarding 'context' and 'core' in business strategy.

softwarePeopleSoft

Mentioned as an example of older, on-premise enterprise software that is being superseded by modern solutions.

softwareCRM

Used as a prime example of how SaaS dramatically expanded the TAM compared to on-premise solutions, making it accessible to a much wider range of businesses.

softwareEHR system

Cited as an example of a specific outcome-based application for AI in healthcare.

softwareRepet

Mentioned as an example of a company expanding use cases through AI, suggesting it's additive rather than solely replacing existing labor.

softwareERP system

Mentioned as a 'contextual' function that companies typically buy externally rather than build themselves.

toolServiceNow