Salesforce AI CEO Clara Shih speaks at the tech giant's Dreamforce event in October 2023. Photo courtesy Salesforce
While few people thought much of artificial intelligence until about a year ago, when ChatGPT exploded into the world’s consciousness, Clara Shih says she’s been thinking about it since college. As an undergrad at Stanford, she took an intro to AI class. As founder of Hearsay Systems, a social media-based software platform, she developed AI to identify sales leads and automate compliance.
Now, as CEO of Salesforce AI, the 41-year-old executive, part of 2023’s Future of Work 50, oversees the artificial intelligence strategy of one of the world’s largest tech companies, integrating it across products like Slack and Tableau. In the process, that’s already starting to redefine the jobs and day-to-day work of sales and customer service reps at the 150,000 companies that use Salesforce’s software.
“A salesperson today spends some ridiculous [amount] of their day on admin tasks or non-selling activities,” she says, such as researching customers, preparing presentations and writing emails to prospective clients. “These are very onerous, frustrating tasks.”
Salesforce wants to change that. As one white-collar job category after another stands to get reshaped by AI, the $31 billion tech giant is poised to edit job descriptions for some of the most common categories in the professional “knowledge worker” class—people who answer customers’ questions, manage sales accounts and come up with marketing content.
Spearheading such an effort requires a leader with technical know-how and managerial skills, says Robert Sutton, a professor of organizational psychology at Stanford’s Graduate School of Business, who first met Shih while looking for contenders for a fellowship program and has followed her trajectory ever since. Shih has long been ahead on industry-disrupting ideas, writing a best-selling book back in 2010 about the impact social media could have on business, Sutton says.
“Clara understands how to get things done at founder-led companies,” he says, referring to Marc Benioff, who founded Salesforce more than 20 years ago and remains CEO.
The creation of Shih’s role—Salesforce had not had a head of AI before—underscores generative AI’s importance in reshaping the future of work, certainly in hot, buzzy startups and mainstream businesses, but especially in big technology players, who must stay ahead as AI prompts an arms race of new features and tools.
Shih returned to Salesforce in 2020 after running Hearsay, which she founded in 2009; she’d done an earlier stint at the tech giant. Now at the helm of Salesforce AI, she’s creating new functions that make rote tasks simpler and easier like Sales GPT, which can schedule meetings and write sales emails from company-specific data, as well as Marketing GPT, which can conduct research on audience segments. The company’s Commerce GPT can craft product descriptions and make recommendations while chatting with customers.
Underlying it all: Salesforce’s all-encompassing Einstein GPT, which can generate everything from marketing copy to code. “Most of the AI [models] 18 months ago were very task-specific models,” Shih says. “These foundational models are really powerful at a broad range of capabilities.”
Although impressively capable, generative AI tools also pose security and privacy concerns, like sensitive company data leaking into the base AI model that may be accessible to competitors or the public. To try and assuage those fears, Einstein filters AI inputs and outputs for harmful content, requires a human to accept or reject the tool’s response before it’s shared, and anonymizes personally identifiable information and payment data, as well as other techniques.
Shih points to Gucci, which uses Salesforce’s customer relationship management software, as an example. By generating automatic replies, AI can free up service representatives at the fashion brand from tasks like using product catalogs, giving reps more time to take on creative jobs like helping with marketing campaigns.
Eventually, Shih says, AI will automate or simplify most multi-step tasks for workers. As a result, she doesn’t see a rosy outcome for all: “I think every time there's a disruptive new technology, there are some jobs that end up getting eliminated,” she said, adding that upskilling and embracing AI will be key to stay in the game.
Meanwhile, Shih also works closely with Salesforce Ventures’ $500 million generative AI fund, which she has called the company’s “secret weapon.” She helps identify and invest in key startups in the space—Anthropic, Cohere, Hugging Face and Runway to name a few—to stay embedded in the AI ecosystem. Within Salesforce, Shih oversees AI research and development, running experiments such as developing tools that can generate audio and video content.
For her part, Shih uses AI to do everything from write drafts of emails to create presentations or even help her come up with icebreakers when she led a recent offsite event for engineers and product designers. AI, she says, acts as “a thought starter. It's exactly what we call it. It's a copilot,” Shih says. “This is a pivotal time—not just for technology, but for society, economies and jobs.”
