Slack update aims to put AI at the center of the user experience

Image credit: Ken Stocker/Getty Images

Slack has evolved from a pure communication platform to one that allows businesses to link directly to enterprise applications without resorting to dreaded task switching. At the Salesforce World Tour event in New York today, the company announced the next step in the evolution of its platform. It puts AI at the forefront of user experience, making it easier to get information and build workflows.

It’s important to note that these are announcements and many of these features are not yet available.

Rather than slap the cover on AI, Rob Seaman says he’s working on incorporating it in different ways throughout the platform. This started with a small step last month, in partnership with OpenAI, bringing the ChatGPT app to his Slack. This is the first part of a broader vision for AI on platforms. That part is currently in beta.

Today’s announcement includes several new integrations, including SlackGPT. SlackGPT is the company’s proprietary generative AI built on the Slack platform, enabling users and developers to build AI-driven experiences. Slack’s content provides a starting point for building relevant models for the platform.

“We believe Slack has its own advantages when it comes to generative AI. Much of an organization’s knowledge on every topic, team, work item and project is already stored in Slack through the messages, files and clips shared daily. It is done,” he said.

Combined with Slack’s partner ecosystem and platform, customers have many options for integrating AI into their workflows. He says Slack is currently thinking about this in three ways.

“First, Slack will use SlackGPT to bring AI natively into the user experience, helping customers work faster, communicate better, and learn faster. We have conversation summaries and writing assistance, which will be available directly in Slack,” he said.

The former, as an example, helps employees catch up on long threads without having to read every message to get the gist of the discussion. The latter helps generate her Slack messages or content for the linked Slack application. This is a slightly obscure use case. Unless it’s an automated message that’s part of your workflow, it’s probably easier to create Slack messages yourself, but if you’re creating content for Slack Canvas, generative AI can help. .

Developers can get in on the action too, embedding AI steps into their workflows and giving them the option to build generative AI experiences themselves, leveraging external apps and large language models. Last week, the company rolled out its updated developer experience to the general public. This should make it easier to incorporate generative AI into the platform in customized ways, Seaman says.

“So, this gives us a foundation to give users the choice and flexibility to incorporate AI into their business work whenever they want. There are currently 2,600 apps in the ecosystem, which includes many leading LLMs, and we see many customers already choosing to integrate generative AI into Slack.”

Finally, EinsteinGPT, the Salesforce-flavoured generative AI announced in March, will also be built into Slack to help employees understand Salesforce content (such as which users are most likely to churn and which accounts are most likely to purchase). You will be able to ask questions directly. This is actually a more direct integration of Salesforce content into Slack, the company he paid $27 billion for a few years ago.

“Slack is really becoming a conversational interface for Salesforce, so we can bring EinsteinGPT-powered insights from real-time customer data that lives in Salesforce into Slack to give every team a better understanding of the customer,” he says. said.

Like most generative AI tools added to enterprise software, Slack announced these features long before they were released, but they give customers an idea of ​​what’s to come and how AI will shape Slack in the future. You should be able to understand how it transforms into The SlackGPT and EinsteinGPT integration is still in development, but developers can now build custom integrations with various LLMs. A workflow builder with a SlackGPT AI connector (which allows customers to instantly connect his ChatGPT or Claude to their workflows or build their own custom his connector that plugs in his LLM) It will be available this summer.