Slack plans to add genAI to its software

 Slack is already working on integrating new generative AI tools into its popular collaboration app. However, the company takes a methodical approach to introducing new features.

Slack plans to add genAI to its software

Slack, after announcing its plans in May to incorporate generative AI into its collaboration platform, is now set to introduce these new features to a select group of customers through a winter pilot program, with a full-scale rollout expected next year.

Over the past few months, Slack has been internally experimenting with generative AI to identify the most effective ways to integrate this technology for collaboration purposes. The initial focus is on three key use cases aimed at streamlining user workflows by automating specific processes.

The first use case involves summarizing extensive conversations that occur within Slack channels and threads. According to Ali Rayl, Senior Vice President of Product Management at Slack, the aim is to provide users with a concise summary of lengthy threads and discussions, saving them from having to read through the entire conversation.

Another feature in development is a channel "recap," designed to offer users quick overviews of discussions within specific team channels. This is particularly helpful for users who have been away from Slack for extended periods and want to catch up efficiently without sifting through a flood of messages.

Finally, Slack plans to leverage generative AI to enhance search capabilities within the application. Users will be able to ask specific questions, and the AI will search through all the information within the organization's Slack workspace to provide precise answers, eliminating the need to browse through various search results.

The integration of generative AI directly into a collaboration platform like Slack is expected to enhance productivity by saving users time, according to Kim Herrington, a senior analyst at Forrester. Leveraging generative AI within the context of existing workspaces can help users quickly transition from seeking information to gaining insights and taking action.

Slack is not the only company incorporating AI into its software; Competitors like Google and Microsoft are also implementing generative AI in their team collaboration applications. This marks the next phase in the evolution of group chat-centric workspaces, as stated by Mike Gotta, Research Vice President at Gartner.

However, there are challenges associated with the use of large language models (LLMs) like generative AI in business contexts. LLMs can produce incorrect information, commonly referred to as "hallucinations." Slack addresses this issue by providing transparency through citations or links to the source of information in conversation summaries. If the AI cannot attribute information to a source, it refrains from generating a response to prevent errors.

Slack is also selective about where and when generative AI is deployed within its application to minimize the risk of hallucinations. Short conversations in Slack's voice and video tool, known as huddles, are not considered suitable for AI-generated summaries due to the potential for inaccuracies in very brief conversations.

To address concerns about data privacy and security, Slack ensures that data remains under the control of customer organizations when using generative AI features. User queries are not used to train external models, and the AI only accesses data that users already have permission to view within their workspace.

In conclusion, Slack's introduction of generative AI promises to enhance collaboration and productivity within its platform, with careful consideration given to addressing challenges related to accuracy and data privacy. The upcoming pilot program will help identify and rectify any issues before a broader rollout.