Humans&, a new AI company, has a bold idea that could change the way we think about AI: the next step isn’t just smarter chatbots or bigger models; it’s coordination. The startup thinks that AI’s ability to write code, generate text, and answer questions is great, but the real breakthrough will come when AI can reliably coordinate people, tasks, tools, and decisions in real-world workflows. They say they’re making a model to show that it works.
The idea is simple but very strong. Most problems at work aren’t caused by not having enough information; they’re caused by not working together well. Teams aren’t on the same page, so projects fall behind. When ownership is unclear, work gets done twice. Approvals get stuck, which makes decisions stall. People have meetings when they don’t know what to do next. Coordination is the biggest problem in places where things move quickly.
Humans& says that AI is in a unique position to solve this problem, not by taking jobs away from people, but by acting as a coordination layer that makes it easier for teams to work together. AI that focuses on coordination would actively keep track of progress, find blockers, and connect the right people at the right time, instead of just responding when asked.
This idea goes against the current trend of AI being obsessed with raw intelligence. Many AI tools today are great at making outputs like drafts, summaries, and suggestions, but they still need people to carry out the work. Coordination is the missing link between “generating work” and “making work happen.”
Humans& says that the future AI system won’t just be able to answer the question, “What should we do?” It will help make sure that the plan goes through. That means knowing what the goals, priorities, timelines, dependencies, and human limits are, and then helping teams carry them out without making things worse.
In real life, coordination AI could change the way businesses work. Think about an AI that reads project discussions, finds unclear responsibilities, automatically gives out tasks, and checks in based on due dates. Or an AI that sees two departments working on similar projects and tells management before time and money are wasted. In customer service, it could tell when a request needs to be escalated and send it right away. In sales, it could make sure that marketing, SDRs, and account executives all got their hands on the same information without any delays.
It looks like Humans& is making this a model-level challenge, not just a feature. That matters because coordinating is harder than writing. It needs memory, an understanding of the context, the ability to reason through multiple steps, and the ability to work with multiple systems like calendars, task managers, CRMs, and communication platforms. Coordination AI has to work over time, not just in one chat window.
This method also shows a big problem with how AI is being used right now. A lot of businesses use AI tools on their own, but their teams still have trouble getting things done. AI might help you write emails faster, but it doesn’t cut down on the number of meetings. It can come up with plans, but it can’t make sure they are followed through on. It can sum up a plan, but it doesn’t make people responsible. Coordination is still a job that only people can do.
Humans think that figuring this out is the key to AI’s next wave of productivity. If AI can work together well, it could cut down on wasted work across organizations by a huge amount. This would make workflows smoother and decisions faster without needing more people.
But coordination AI makes us wonder a lot about trust and control. Coordination has to do with power, responsibility, and politics within an organization. People may not like it when AI gives them tasks or raises issues, especially if they think they’re being watched. For coordination AI to work, it needs to be able to help people without turning into a surveillance system.
This is why Humans&’s approach is probably more about adding to what people already do than taking over their jobs. The goal is not to get rid of managers or team leaders, but to get rid of repetitive coordination tasks like reminders, updates, follow-ups, and tracking ownership. This will free up people to focus on leadership, creativity, and decision-making.
Another problem is getting things right. Coordination systems can’t afford to see things that aren’t there or get the wrong idea about what is going on. A wrong assignment, a missed deadline, or a wrong escalation could all cause real harm. This means that the model needs to be very reliable, easy to check, and in line with what people want. It might need more strict rules than general-purpose AI.
The company’s focus is also a sign of how AI competition is changing. In the early days of AI, people fought over who could make the smartest model. Now, the race is on to see who can make AI that works with real-life tasks. Coordination is one of the most obvious “real world” problems where intelligence isn’t enough; you have to do things right.
If Humans& can show that coordination AI works on a large scale, it could create a new type of enterprise AI. Coordination AI could become infrastructure for teams instead of tools that help individual employees. This would change the way organizations plan and run their businesses.
This could have big effects on the economy. A lot of time is wasted by companies because of inefficiency, misalignment, and meetings that aren’t needed. Coordination AI could make people more productive not by making them work faster, but by making it easier for them to work together. That’s a completely different way to boost productivity.
In the long run, coordination AI might become the invisible layer in workplaces, keeping track of context, moving information between teams, and pushing decisions forward. It would work like a “project nervous system,” keeping everything linked without needing to be managed by hand all the time.
Humans& thinks that this is the next step in the growth of AI. AI became popular because of chatbots. Automation made AI work. But they say that coordination is what will make AI truly revolutionary.
We don’t know if the company can make this vision a reality, but we do know that the next AI frontier isn’t just about getting smarter answers. It’s about making things go more smoothly.
