Notion Projects Aims To Revolutionize Workplace Collaboration, With The Help Of AI

Workplace collaboration tools are supposed to make things easier. I think they just make everything more complicated.

From my personal experience, these tools require quite a bit of manual input from users and are limited in their scopes of functionality. Strategy documents, proposals, roadmaps, meetings and notes may be found living in separate apps.

Notion hopes that AI can help cut down on the fragmentation within this space with its new product, Notion Projects. Projects aims to connect all aspects of collaboration in one place, making it easier for teams to plan, manage, and execute work, with the help of AI LLMs from OpenAI and Anthropic.

Notion Projects leverages AI to automate many tasks within a project lifecycle, reducing the need for a team member to manually provide updates, update action items or provide summaries within a workspace.

You may already be familiar with Notion as a project management tool, but the company is beefing up its project management suite with new project-specific features, including upgraded project templates and setup flows. Teams will be able to manage projects from beginning to end with more out-of-the-box functionality, comparable to its competitors Asana and Jira.

In a virtual interview, Notion COO and co-founder Akshay Kothari gave me a sneak peek of Projects. Having been a Notion power user for years, I was amazed at the AI Autofill feature, which is capable of filling out an entire workspace worth of information in seconds.

With the rollout of this release, Notion users can take advantage of AI Autofill, which automatically extracts information teams need and keeps it updated as projects progress. There are also new page-level and property notifications, improvements to subtasks and dependencies, and the ability to preview Google Drive files and Figma designs in project databases.

Additionally, Notion Projects comes with upgraded sprint plans, streamlined sprint templates, and a new sprint completion workflow, along with automations to help teams stay informed as things change, with more integrations soon to be announced.

For $10 per month, users can access unlimited AI functionality. Kothari acknowledges that the cost of running AI models is still high, but he hopes that as technology advances, AI will become more accessible and affordable for users.

According to Kothari, Notion has always been interested in AI, but it wasn’t until last year that they began to see its true potential. After gaining early access to advanced AI models, the company quickly realized that AI could significantly change how people do knowledge work. In response, Notion launched an alpha program in October, followed by a general release in February, which introduced AI-powered writing features to the platform.

One of the key findings from the alpha program was that AI could be used to augment human intellect rather than replace it. Kothari himself uses AI to quickly turn bullet points into product specifications, speeding up his workflow and reducing procrastination.

Every consumer and business app seems to be finding a way to integrate generative AI. I asked Kothari whether the company would have integrated AI tools this quickly into the platform if OpenAI’s ChatGPT hadn’t gone viral.

“I think we would. We saw the power of this technology months before OpenAI launched ChatGPT. I think we were one of the only companies to see it, and we set up a bet on it,” he said, adding the company was one of the few to have early access to certain advanced LLMs. However, ChatGPT did influence Notion to take a more conversational approach with its AI tool.

I was curious whether Kothari was worried about a potential backlash to this functionality. For example, Samsung banned its office workers from using generative AI after finding out sensitive code had been leaked via ChatGPT. Was he concerned at all about businesses rejecting these new features?

“Our business is built on trust and people trust us with very sensitive content,” he tells me, adding that privacy was a starting point for Notion’s initial discussions with OpenAI and Anthropic around using their respective LLMs.

“We mandated that anything we send them to create this content can’t be used for training [the LLMs] and that we couldn’t offer this functionality to our customers unless that was in place.”