For One Person
For a TeamA year ago, the conversations I heard about AI centered on one question: which model is the smartest?
People compared ChatGPT, Claude, Gemini, and every new model release that came out. Which one writes better? Which one codes better? Which one reasons better?
That question made sense because the models were the product. I think that’s starting to change.
I think the next phase of AI will be defined by who can build the most useful relationship between a person and an AI, more than by who has the most intelligent model. Intelligence alone isn’t enough to make an AI useful to you.
The smartest AI system in the world is still limited if it doesn’t understand the person using it.
The Model Is Becoming Infrastructure
AI is moving faster than any technology I’ve watched before. Models that were considered the best available a year ago are quickly becoming outdated, and new systems keep improving at reasoning, writing, coding, research and problem-solving.
Those improvements matter, but they also create an interesting shift. When every major AI company has access to extremely capable models, the underlying intelligence starts becoming less of a differentiator.
A similar thing happened with computers. Most people I know chose their laptop for what it helps them do, without knowing the specs of every component inside it. The hardware matters, but the experience matters more.
AI is moving in the same direction.
The model is the engine, and I think the value will come from everything built around it.
The Missing Ingredient Is Context
The difference between a generic AI assistant and a useful one comes down to context.
A personal AI knows your goals, your preferences, your habits, your communication style, the tools you use, the people you deal with, and the decisions you make over and over. That changes what it can do for you.
Say you ask an AI to plan a trip. A basic assistant can find flights and hotels.
A personalized assistant knows that you prefer early flights, hate long layovers and usually stay at Marriott properties. It knows you have a golf round the day after you arrive, and that you like to keep your mornings open before important meetings.
The task itself is easy, and the value comes from knowing how you think.
This Is Why Agents Matter
When I hear AI agents discussed today, the focus is often the automation side of things:
- An AI that can browse the internet.
- An AI that can click buttons.
- An AI that can fill out forms.
Those capabilities are impressive, and the browser is a tool. The real breakthrough happens when an AI understands what needs to happen and why.
Finding a flight is easy. Knowing which one you would choose, and checking it against your calendar, is where the value comes from. The booking stays yours.
A useful agent does the legwork around your decisions, in the context of your life.
The Problem With Today’s AI
To me, the biggest limitation of AI today is that every interaction still starts too close to the beginning.
- You open a new chat.
- You explain what you need.
- You provide context.
- You remind it how you like things done.
- Then you repeat the process again later.
Relationships don’t work that way. If you hired an assistant and had to reintroduce yourself every morning, explain your priorities every afternoon, and remind them how you prefer things done every week, they would not be very useful.
The value of a great assistant comes from accumulated knowledge. They get better because they understand you better over time. AI should work the same way.
The Future Probably Runs on Many AI Models
One of the most interesting questions about the future of AI is whether one company will become the single winner, and whether that’s OpenAI, Anthropic, Google or someone else.
I think the future is probably more complicated. Different models will have different strengths. One may be better at writing, another at coding, another at research, and another at specialized tasks.
You don’t think about which search algorithm answers your Google query, and you shouldn’t have to think about which AI model handles a task either. The system should simply know.
I think the future may belong to the best system built around those models.
This Is Why We Built Otian AI
This is the idea behind Otian AI. We think the future is a personal AI assistant that understands how you live and work and helps you get more done with less friction.
Your schedule, goals, work, habits and everyday tools are yours alone, and a generic AI experience can’t fully account for that.
At Otian AI, we build Archie, which runs an agent on your own computer. It keeps what it learns about you, works with the accounts you connect, and does the legwork around the decisions that take your attention every day.
For one person, that might mean running a business, organizing projects and keeping communication in order. For another, it might mean staying on top of health, travel, bills or long-term goals.
The technology is the same, and what it does for each person is different.
The Future of AI Is Personal
The companies building the smartest models will continue pushing the boundaries of what AI can do. But intelligence alone does not create usefulness.
I think the most valuable AI will be the one that knows the most about you.
The goal is an assistant that understands what you are trying to accomplish, knows how you operate, and helps you get there without you managing every detail yourself.