Key Takeaways
- The stocks of software companies such as Salesforce have soared this week on hopes that Chinese start-up DeepSeek’s cost-efficient AI model could make AI applications significantly less expensive to develop and run.
- DeepSeek’s success could hasten the roll-out of AI agents, semi-autonomous digital assistants that Wall Street analysts expected to command greater investor attention this year.
- Microsoft’s Azure platform, Morgan Stanley analysts wrote on Tuesday, could benefit from a proliferation of AI models and consumer applications.
Chinese startup DeepSeek’s super-efficient open-source AI model may have sunk some of Wall Street’s favorite AI stocks on Monday, but it may have also crowned some new favorites.
Yesterday was the worst day for Nvidia (NVDA) and the Philadelphia Semiconductor Index (SOX) since March 2020. Yet some tech sector stocks have soared—Salesforce (CRM) stock has risen 8% this week and Apple (AAPL) has climbed 7%.
That’s because the AI trade has, up to now, been driven by the suppliers of the “picks and shovels” of artificial intelligence: semiconductors, networking equipment, and electricity. But DeepSeek’s success suggests AI’s picks and shovels may not need to be as plentiful or as fancy as investors once thought. That could be a boon to AI buyers, like software companies, who pay cloud service providers and AI companies for access to foundational models and computing power.
Could DeepSeek Hasten the Rise of AI Agents?
Several analysts predicted at the end of last year that AI agents—digital assistants capable of performing more tasks than AI chatbots—would draw more interest from investors.
Major tech companies and upstarts have already begun rolling out AI agents that have the potential to reshape consumer behavior, and Wall Street has taken notice. OpenAI last week launched its AI agent, Operator, which the company said could order users’ groceries or book their plane tickets. Salesforce’s Agentforce saw “healthy early interest” in its first months, according to Bank of America analysts. And shares of med-tech company Tempus AI (TEM) soared last week after it released olivia, its AI-power health concierge.
The pivot from infrastructure to application may have been hastened by DeepSeek’s model, the cost-efficiency of which can likely be replicated by U.S. companies.
“We believe that DeepSeek’s innovations at the model layer reinforce our thesis that the Application and Platform layers are set to benefit as revenue moves from the Infrastructure layer and enterprises allocate more budget to AI,” wrote Goldman Sachs analysts in a note on Tuesday. Lower computing costs, they say, “should help catalyze broader usage of AI workloads and encourage adoption across businesses and consumers.”
Enterprise Software Could Benefit from DeepSeek Ripple Effect
Goldman expects the benefits of lower computing costs to accrue first to major software companies with established AI products, including Microsoft (MSFT), Salesforce and Adobe (ADBE). But, they say, more efficient models should lower the barrier to entry for smaller competitors and encourage the proliferation of AI use cases.
Morgan Stanley analysts agreed that enterprise software companies were most likely to benefit from the savings that should follow from America’s DeepSeek reckoning. “More cost-efficient models are bringing down ‘GenAI input costs’ for the broader Software ecosystem,” the analysts wrote. They highlighted Microsoft as a major beneficiary, noting its Azure platform is an optimal location for application developers to access and build on foundational AI models.