Emilia Gette serves as a marketing associate at Workast, specializing in crafting innovative strategies to engage with the audience. With a focus on creativity and attention to detail, she contributes to driving impactful results across various marketing channels.
Spend a few minutes reading AI communities online and you'll notice something interesting. People aren't complaining that ChatGPT is bad. They're not complaining that Claude is bad, and developers certainly aren't complaining about Cursor.
The complaints are usually about something else. The tools themselves are impressive. Working together with them is where things start to break down.
A few years ago, the question companies asked was simple: "Should we use AI?" Today, many teams have already answered that question. People are using ChatGPT, Claude, Cursor, Gemini, Perplexity, coding agents, and specialized tools every day. AI adoption is no longer the challenge.
The problem is that adoption rarely happens in a standardized way. Everyone builds their own system, develops their own habits, and accumulates their own context. As a result, individuals become more productive, but teams often end up working with completely different workflows.
One Reddit user experimenting with different "team AI tools" described the experience this way:
"When I'm deep in a conversation with GPT about our project and my teammate joins, they have zero context. We're basically starting over."
Another person explained that using different AI tools feels like constantly talking to strangers.
"Every time I switch, it feels like I'm talking to a different person who knows nothing about what I was doing before."
Others describe manually sharing summaries, screenshots, and conversation threads with teammates because collaboration still feels unnatural. These aren't isolated complaints. Similar frustrations appear repeatedly across communities, suggesting that people are running into the same underlying issue regardless of which AI tools they prefer.
Most users aren't asking for smarter AI because they're already happy with the capabilities available today. ChatGPT is good at planning, Claude excels at research and writing, and Cursor has transformed software development for many engineers. AI is clearly helping individuals move faster.
But teams don't operate as isolated individuals. Collaboration still depends on shared context, and shared context becomes difficult when everyone works with different assistants, different prompts, and different systems. The challenge isn't that AI lacks intelligence. It's that intelligence becomes fragmented when every person has their own workflow.
Another interesting pattern appears repeatedly in these discussions. People describe creating their own systems to compensate for the lack of shared context. Some rely heavily on Slack and Notion. Developers maintain markdown files and project notes. Others share conversation summaries, screenshots, or exported chats to keep everyone aligned.
One Reddit user explained that they simply screen-share and "vibe code together" because existing workflows don't feel natural. Another pointed out that the AI itself isn't the bottleneck anymore. Inconsistent workflows are.
Teams are finding ways to make things work, but many of those solutions feel improvised rather than intentional.
Reading these discussions, one thing becomes clear. People aren't saying, "I need better AI." They're saying, "I need better coordination.
"AI tools are becoming incredibly capable, but as every employee develops their own workflows and assistants, collaboration can become harder rather than easier. That's perhaps the hidden problem nobody anticipated. AI made individuals more productive, but keeping teams aligned may be the next challenge.
When everyone has their own AI assistant, where does shared work actually live?
Workast
