Salesforce CEO Marc Benioff said his company will likely spend $300M on Anthropic tokens

  • Salesforce CEO Marc Benioff has high praise for AI coding agents and Anthropic.

  • The CEO said agents allow him to iterate on products more efficiently and faster.

  • He said Salesforce will also introduce technology to make coding within Slack easier.

Marc Benioff is all in on the AI ​​coding agent craze and is willing to invest heavily in his company.

In an episode of the All-In podcast released on Friday, the Salesforce CEO said his company expected to spend $300 million on Anthropic’s tokens.

Tokens are bits of information, such as parts of words, that are input by the user and processed by the model. The AI ​​company tracks token usage in order to bill users and large organizations.

“These coding agents are awesome. Humans are awesome,” he said. “I’ll probably be coding at Salesforce this year with $300 million in Anthropic (tokens). Everything will be cheaper.”

Benioff said in the podcast that AI agents have brought “unprecedented” levels of efficiency to his company, pointing to areas such as service, support, distribution and marketing.

Last August, the CEO announced that artificial intelligence agents would allow Salesforce to cut its support staff from 9,000 to 5,000 people.

Benioff revealed on the podcast that as coding agent productivity increases, Salesforce is also looking into how to make coding easier in Slack, the productivity platform Salesforce has owned since 2021.

See also  German FA chief calls for humility before World Cup draw

“We’re even working on technology inside Slack to make coding easier for everyone,” he said. “You’re going to see some cool stuff coming from Slack and code that I’m not ready to talk about yet. But there’s no question that we’re in a new moment in coding.”

Benioff said that while he expects to spend $300 million on Anthropic this year, he doesn’t believe every token used by the company’s employees will necessarily go to cutting-edge models like Anthropic’s Claude.

The CEO called for a future “middle layer” that could determine which token inputs should go into Anthropic and which might be suitable for smaller models.

Read the original article on Business Insider

Spread the love

Leave a Reply

Your email address will not be published. Required fields are marked *