In the frothy business of selling artificial intelligence (AI) service, Salesforce Inc has been punching above its own weight.
“Salesforce?” I hear you wonder. The folks in the dull business of selling customer relationship management software?
Yes, and they are getting aggressive about AI too. The company has just become the first major tech firm to launch a new breed of generative AI tools known as “agents,” which others on the cutting edge of the field have long talked about without delivering anything. Unlike its peers, Salesforce is explicit about how its novel tools would displace jobs. The approach is bold, but could be what is needed to push the firm ahead in the AI race as younger companies such as OpenAI and Anthropic encroach on its territory.
Illustration: Yusha
You can attribute this feistiness to larger-than-life Salesforce CEO Marc Benioff, whose maverick strategies helped him spearhead the Software as a Service (SaaS) revolution, and grabbed him a client base that includes 90 percent of Fortune 500 companies, including Walt Disney Co and Ford Motor Co. Salesforce makes money by selling subscriptions to applications, such as Sales Cloud and Service Cloud, which companies use to keep track of their sales and customer service processes. Salesforce’s big pitch at Dreamforce this week — its lavish conference that all but takes over downtown San Francisco each year — was Agentforce, a new service that lets customers deploy autonomous AI-powered agents.
Bots are out and agents are in, Salesforce said. The former are the chatbots, powered by technology from firms such as OpenAI, Google and Anthropic, that sit in the corner of a Web site and can answer a customer’s questions. Now, instead of just dispensing information, agents can take action. They can file a complaint, book an appointment or change a shipping address on file.
“Taking action” might sound like a huge risk for businesses to make with AI, given that generative models are known to hallucinate, and chatbots have occasionally given erroneous information to customers. Imagine an AI model mishandling a booking.
Salesforce says that would not be a problem.
“Hallucinations go down to zero, because [Agentforce] is only allowed to generate content from the sources you’ve trained it on,” Salesforce corporate strategy director Bill Patterson said.
That makes it more reliable than OpenAI, which scrapes the public Internet and all its false content, he added.
That is a daring premise. What is also daring: Salesforce has become quite comfortable addressing an issue that AI companies until now have mostly tiptoed around — that success for their business customers is measured by cutting costs, AKA fewer jobs.
One executive from a separate AI company recently said that their enterprise customers were hesitant to go public as a case study, because they feared reputational blowback. Slashing jobs is not a great look.
However, Salesforce had several case studies to point to on Tuesday last week, such as education publisher John Wiley & Sons Inc, which used Agentforce over the past three months to nearly halve the time spent answering customer inquiries.
August and September are when Wiley normally hires many more people to handle the spike in demand from students going back to school, but with Agentforce it did not have to shell out on new recruits, Salesforce said.
“What if you could surge your service organization and you could surge your sales organization without hiring more people?” Benioff asked in his keynote. “That’s Agentforce.”
Benioff reportedly rewrote his keynote days before the conference, because he believed he had to address the “deep frustration around AI.” One result seems to be that he did not hold back. Just to underscore Salesforce’s belief that Agentforce would replace humans, the company has flipped its business model for the service. Instead of charging per head, which would mean its revenue only grows if its customers hired more people, Agentforce charges US$2 “per conversation.” So if a client such as Wiley hires fewer workers, it can still pay Salesforce more. It is a shrewd pricing strategy.
Time will tell if businesses record a return on investment from Salesforce’s newfangled tools, particularly when many firms are finding it difficult to measure the success of generative AI. However, it marks a clear threat to newer firms such as OpenAI and Anthropic — which have admitted behind closed doors that they use Salesforce themselves — and that what they are offering to customers is an additional service to Salesforce’s CRM software. For many chief innovation officers, it is far easier to tell their sales staff that they can just keep using Salesforce, which is already plugged into their data and processes, rather than get used to yet another thing on their computers. Nobody wants to have to look at yet another thing.
Salesforce’s entrenched, wide distribution puts it in a strong position at a time when larger companies are generally slow about adopting new tech, including generative AI. Its undaunted approach to job displacement suggests it intends to make lots of money from this venture, too. When it comes to selling to businesses, AI firms might face a formidable competitor.
Parmy Olson is a Bloomberg Opinion columnist covering technology. A former reporter for the Wall Street Journal and Forbes, she is author of We Are Anonymous.
Chinese state-owned companies COSCO Shipping Corporation and China Merchants have a 30 percent stake in Kaohsiung Port’s Kao Ming Container Terminal (Terminal No. 6) and COSCO leases Berths 65 and 66. It is extremely dangerous to allow Chinese companies or state-owned companies to operate critical infrastructure. Deterrence theorists are familiar with the concepts of deterrence “by punishment” and “by denial.” Deterrence by punishment threatens an aggressor with prohibitive costs (like retaliation or sanctions) that outweigh the benefits of their action, while deterrence by denial aims to make an attack so difficult that it becomes pointless. Elbridge Colby, currently serving as the Under
The Ministry of the Interior on Thursday last week said it ordered Internet service providers to block access to Chinese social media platform Xiaohongshu (小紅書, also known as RedNote in English) for a year, citing security risks and more than 1,700 alleged fraud cases on the platform since last year. The order took effect immediately, abruptly affecting more than 3 million users in Taiwan, and sparked discussions among politicians, online influencers and the public. The platform is often described as China’s version of Instagram or Pinterest, combining visual social media with e-commerce, and its users are predominantly young urban women,
Most Hong Kongers ignored the elections for its Legislative Council (LegCo) in 2021 and did so once again on Sunday. Unlike in 2021, moderate democrats who pledged their allegiance to Beijing were absent from the ballots this year. The electoral system overhaul is apparent revenge by Beijing for the democracy movement. On Sunday, the Hong Kong “patriots-only” election of the LegCo had a record-low turnout in the five geographical constituencies, with only 1.3 million people casting their ballots on the only seats that most Hong Kongers are eligible to vote for. Blank and invalid votes were up 50 percent from the previous
Japanese Prime Minister Sanae Takaichi lit a fuse the moment she declared that trouble for Taiwan means trouble for Japan. Beijing roared, Tokyo braced and like a plot twist nobody expected that early in the story, US President Donald Trump suddenly picked up the phone to talk to her. For a man who normally prefers to keep Asia guessing, the move itself was striking. What followed was even more intriguing. No one outside the room knows the exact phrasing, the tone or the diplomatic eyebrow raises exchanged, but the broad takeaway circulating among people familiar with the call was this: Trump did