By Hamudi Naanaa, CTO and co-founder at Portal.ai
That is the pitch, anyway. Productivity. Efficiency. Fewer humans in the loop. The assumption is that systems are slow because people create friction, so remove the people.
I work in AI. I am supposed to believe this. Here is the problem: I do not.
I grew up between Lebanon, Ukraine, and Germany. I have spent my whole life translating, not just languages, but contexts. The way one culture shows love through silence. The way another shows it through argument. The constant work of making yourself understood to people who see the world through completely different histories and mental models.
I see this everywhere now. A founder in a room full of investors who hear the words but miss the vision. Two people on a first date, both hoping to be seen, both performing instead. A partner trying to understand your intensity and not knowing how to ask. Cofounders who want the same future but cannot stop fighting about how to get there. Parents and children who love each other but have forgotten how to talk.
The friction is not people being difficult. The friction is the space between them.
I used to think this was just how it worked. The cost of being human.
Now I think it is a design failure we accepted because we had no alternative.
There is a clay tablet from ancient Mesopotamia, around 1750 BCE, where a customer named Nanni writes a furious complaint to a merchant named Ea-nasir about being sold bad copper. He is angry. He feels cheated. He has been wronged and no one is listening.
That tablet is nearly four thousand years old. And when people see it today, they laugh, because they recognize it instantly. The frustration. The miscommunication. The gap between what one person meant and what another person heard.
We have been here before. We have always been here.
If you read texts from five thousand years ago — Gilgamesh, ancient philosophy, religious scriptures — you find the same stories. Misunderstanding. Longing. The ache of being seen or unseen. People trying to reach each other and falling short.
Completely different world. Same emotional architecture.
This tells us something important. Navigating human connection is not a modern problem. It is human nature. We have carried it with us for millennia. And for all that time, we have used the same tools: language, intuition, memory, trial and error. Five thousand years. The same tech.
And what did we do since then? The internet connected 6 billion people. It did not teach any of them to understand each other. Until now.
Here is a sentence I keep returning to: “Every major human problem is a coordination problem between minds.”
Think about it. Business is coordination: between founders, employees, customers, investors. Relationships are coordination: between people with different needs, fears, and ways of showing love. Politics is coordination: between groups with different values trying to share a society. Even creativity is coordination: between collaborators, between artist and audience, between the idea in your head and the person who needs to understand it.
Once you see it this way, something shifts. The question is no longer “How do we make humans more productive?” The question becomes: “How do we help minds coordinate better?”
And suddenly, AI as a translation layer between people does not feel aspirational. It feels inevitable.
We have never been more connected. More than 6 billion people now hold active social profiles. People spend over two hours a day in feeds that promise connection. Yet when researchers ask users how they feel after those hours, a growing share answer that they feel more alone than before.
It is a strange kind of progress. We built a global architecture for communication and then discovered that communication is not the same as understanding.
And now the volume of content has reached a place where even the platforms struggle to filter it. More than half of public social content in 2025 is AI-generated. When language becomes abundant, meaning becomes scarce. And when meaning becomes scarce, people crave guidance toward moments of sincerity.
There are 8 billion people on Earth. That means 8 billion private languages, and no dictionary between them.
So here is the space everyone is overlooking.
Not AI that replaces human interaction. AI that sits between humans – like a translator. Because every person speaks from their own history. Their own context. Their own unspoken meanings. And most of the time, we have no way to bridge that gap except through patience and luck.
What if we had something better than luck?
This is the context in which a new category of technology is forming. Not social networks. Not messaging tools. Something quieter. I call it a “connector engine.” An intelligence that prepares us for other people rather than replacing them.
Portal is one example of this shift. It looks at you. It looks at the person you are about to meet. It reads what you both share publicly, the lines between your histories. Then it gives you a small private brief.
Not a script. Not a sales trick. A few questions you might ask that reflect the center of gravity between you. And a quiet note about subjects that might not land well, not because they are forbidden, but because they will not help you see each other clearly.
It feels almost old-fashioned. As if the machine is reminding you that the goal of meeting someone is not efficiency. The goal is recognition.
Imagine a future where every introduction begins this way. Two people meeting for the first time. Both entering the room with a small sense of alignment. Knowing why the encounter matters. Knowing where the overlap in their worlds might be.
It feels small. But small shifts accumulate. And they change how we experience each other.
There is another dimension to this: the network layer. Not introductions you already expect to have, introductions you never considered. Based on something deeper than common interests. Based on how you both think about risk. Or how you both talk about childhood. Or how you both write when you are tired and honest.
For years, our networks grew horizontally. More contacts. More acquaintances. More professional ties. Very few turned into real relationships. The connector engines of this new category work differently. They look for vertical depth. They try to find where two people might actually understand each other.
None of this is perfect. Machines can misread tone. They can misinterpret humor. They can overestimate overlap or underplay difference. The risks around privacy and inference are real.
But imperfection is not a reason to avoid the category. It is a reason to build it with more care. Because the alternative is the world we have now: endless feeds, endless messages, endless information – and very little that helps us see one another.
What these systems offer is not certainty. It is a small moment of clarity before a meeting. A way of entering a conversation without the usual blind spots. A way of preparing for someone’s humanity.
In the end, this is not about Portal. It is about us.
At the end of everything: every product, every company, every relationship, every idea, there is a human being. And between any two humans, there is a gap. We have known this for four thousand years. We just never had the tools to do anything about it, except talk, and hope, and try again.
Now we do.
Everyone is building AI to take humans out of the equation. We are building AI to recognize each other.
