Slack's Successor Won't Be a Better Slack

Deep thoughts on AI and aspirations —— ByteDance Deep Thinking Circle

Slack has dominated team collaboration software for years, with nearly every company using it. But recently someone raised $20 million claiming they’ll take it down—not by building a “better Slack,” but by rebuilding from the ground up.

Sounds like a joke, but this thinking has merit: Slack’s pain points are as numerous as its conveniences. And the way to solve that pain might not be optimizing chat, but changing the underlying logic of collaboration.

The Root of Pain: Relevant Things Never Find You

Anyone who’s used Slack knows the feeling: dozens of channels, hundreds of messages daily, and maybe only 10% are actually relevant to you. But you can’t ignore them because you’re afraid of missing something.

The problem lies in the channel structure itself. Channels dump all content at you indiscriminately, whether it relates to you or not. To avoid missing relevant content, you’re forced to endure massive amounts of irrelevant noise. This contradiction is a structural flaw of chat tools, not something optimization can solve.

One emerging trend: redesigning the relationship between “content” and “you”—not pulling everyone into one channel, but letting each topic find its relevant people. When you discuss a project, participants join in while unrelated people aren’t disturbed. This way everyone only sees what’s relevant to them, reducing noise while actually increasing attention.

The Real Change: AI Becomes a Team Member

If it’s just better message organization, that’s improvement, not revolution. The real variable is: AI begins participating in collaboration as a team member, not just a tool.

Previously, if you wanted to know “what the designer’s been working on,” you’d either ask in a meeting, scroll through chat history, or ask someone else. An AI truly integrated into the collaboration environment can answer you directly and tell you where it learned that—you don’t have to scroll through dozens of messages, it does it for you.

Going further, AI shouldn’t just answer questions but proactively participate at the right moments: when you’re discussing a problem it’s seen solved before, it pushes the solution to you. From “you ask, it answers” to “it sees your need and helps proactively.” This shift may arrive faster than imagined.

This also explains why AI belongs in collaboration software rather than separate from work—only where work happens does it have enough context to know when to help.

The Other Side of the Bet: Why Bet Old Giants Will Be Replaced

The people doing this know they’re going up against Slack and Teams, but they’re betting: when technological paradigms shift, established players are often the hardest to pivot.

Just as Slack itself replaced older communication methods, the history of collaboration tools is one of constant rewriting through paradigm shifts. Current software makes “people adapt to structure” (learning how to use channels, set permissions), while AI-era software should make “structure adapt to people” (you say what you need, it organizes for you).

Whether the bet holds depends on one thing: is AI collaboration merely nice-to-have, or does it fundamentally change work efficiency? If the former, Slack adding an AI button suffices; if the latter, a new paradigm truly emerges.

Assessment

My view is that the value of such products isn’t in the “replace Slack” slogan, but in the question they raise: When AI can participate in collaboration, help you find information, and proactively assist, what should team collaboration software look like?

The answer is almost certainly not “a better chat app.” Whether it’s threaded or AI-native forms, the market will decide. But for all teams building collaboration tools, this question deserves careful thought—it determines whether you’re refining the old paradigm or defining the new one.

Key points: Collaboration pain stems from information overload caused by channel structure; the solution is letting content find relevant people; AI as collaborator is the real variable; paradigm shifts make old giants hardest to pivot; value lies in answering “what should AI-era collaboration look like.”