Salesforce CEO: AI Agent is the New Engine for Enterprise Empowerment
Salesforce CEO Marc Benioff predicts that AI agents will become the mainstream of enterprise applications, completely changing the way enterprises interact with customers and triggering profound changes in the labor market. AI will expand rather than replace human work.
# AI Agents: A New Chapter in Enterprise Applications
Unlike personal assistants like Siri and Alexa, the future of AI agents lies in enterprise applications. Enterprises will build their own AI agents to represent them in intelligent interactions with customers, providing personalized services and experiences and deeply integrating business processes and customer data.
Salesforce has a deep accumulation in data clouds and automated touchpoints, providing a solid foundation for building enterprise-level AI agents. AI agents can be trained using data clouds to learn business rules and customer behaviors and integrate with automated tools to turn intelligent decisions into practical actions.
Taking Disneyland as an example, an AI-guided tour can integrate tourist information and park data to provide real-time personalized play suggestions. AI agents will also bring changes to industries such as finance and retail and can serve as intelligent investment advisors, virtual shopping guides, and intelligent customer service representatives to provide more accurate and efficient services.
# AI and the Workforce: Expansion Rather than Replacement
The development of AI has raised concerns about unemployment, but Benioff believes that AI will expand rather than replace human work.
Taking the healthcare industry as an example, AI agents can undertake repetitive and low-value tasks such as intelligent scheduling, intelligent consultations, and intelligent customer service, freeing up healthcare workers to focus on more important medical tasks such as diagnosing difficult and complicated cases, formulating complex treatment plans, and having in-depth communication with patients.
AI will not lead to large-scale unemployment but will trigger a rebalancing of the labor market. New AI-related jobs will emerge, such as AI engineers, data scientists, AI trainers, and AI ethics reviewers. New work forms will also appear, and human-machine collaboration will become the mainstream. Salesforce uses the AI agent Einstein to optimize employee allocation, transferring customer service personnel to other positions that require more creativity, strategic thinking, and complex problem-solving abilities, creating greater value.
# Pay-as-You-Go: Business Model Innovation for AI Agents
Salesforce’s AI agent product Einstein adopts an innovative pay-per-conversation model. The price per conversation is approximately between 50 cents and 2 dollars depending on usage. Compared to the traditional per-user pricing model, this model is more flexible, more in line with the characteristics of AI applications, and more economical. The use of AI is uncertain, and its value lies in interaction. The cost is related to usage, so pay-per-conversation is more reasonable.
AI agents can significantly reduce enterprise operating costs. For example, replacing some manual customer service with AI agents can save a lot of expenses and provide 24/7 service.
The nonprofit organization College Possible uses Salesforce’s AI agent platform to build an intelligent college application guidance system, greatly expanding the service scope and helping more low-income family students obtain college admission guidance, reflecting the social value of AI agents.
Currently, the pricing models of AI agent providers such as Salesforce, Microsoft, and Google are different. Salesforce charges per conversation, Microsoft Copilot adopts a subscription model, and Google Gemini’s pricing model has not been fully disclosed. Salesforce’s pay-as-you-go model is more flexible, can better adapt to the actual needs of different enterprises, and is more cost-effective.
# Enterprise Software in the AI Era: Changes and Challenges
AI agents drive changes in enterprise software. In the AI era, data, security, and AI will be deeply integrated. The core of enterprise software is still data storage, management, and security. AI agents are an intelligent layer built on enterprise software rather than replacing enterprise software itself.
AI agents transform the interaction mode from a graphical interface to natural language. Users can interact with AI agents through natural language, such as directly instructing AI agents to complete tasks. This will greatly improve usability and user experience, making enterprise software more intelligent and humanized. Behind this is the rapid development of natural language processing technology.
Benioff questions the radical remarks of Klarna’s CEO about a “software-free” future. He believes that even in the AI era, enterprises still need a secure and reliable database to store and manage various data and conduct strict permission management.
Benioff criticizes Microsoft’s Copilot for lacking deep integration and having security risks such as the risk of user data leakage. He emphasizes Salesforce’s advantages in data security and privacy protection and points out that Salesforce’s AI agent Einstein is built on a mature CRM platform and data cloud and has a more complete security mechanism.
Enterprises need to take stricter security measures to protect data, such as data encryption, access control, data masking, and security audits, to address the data security challenges in the AI era.
AI is also transforming software development methods. AI code generation and review tools can help developers write code more efficiently, detect errors and vulnerabilities, and
