Michał Stekla
ChatGPT is today something more than just a bot. In just two years, it has become one of the most widely used AI tools in the world. However, OpenAI has much greater ambitions for it — and specific plans have just leaked.
According to a document created in 2024, which emerged in the context of the Google antitrust case, we learn that OpenAI does not want ChatGPT to be just helpful. It is supposed to be Your main interface to the internet. An assistant that not only answers questions but understands you, anticipates needs, and acts on its own — like an intelligent, empathic person with a laptop at hand.
“We want ChatGPT to help you throughout your life — wherever you are,”
it states in the document.
This means one thing: ChatGPT is supposed to take notes from your meetings, manage presentations, check in on friends, and suggest where to eat along the way. And this is not a futuristic vision — the first such features are set to be launched as early as 2025.
T-shaped artificial intelligence?
OpenAI defines ChatGPT as a "T-shaped" model. What does that mean? Broad competencies in everyday matters and very deep knowledge in difficult topics — those that are beyond the reach of an ordinary person. Learning programming? Game development? Financial analysis? Certainly.
Plans for 2025? First development, then monetisation
The first half of 2025 will be about building a "super assistant". The second half? Creating new business models that will enable it to be maintained and developed. OpenAI does not hide the fact that one of the biggest challenges will be infrastructure — as the growing number of users = immense demand for computing power. That's why Altman is putting everything into building its own data centres.
Competition? Meta in the crosshairs
The document contains an interesting passage: the greatest threat to OpenAI is said to be an entity that "can integrate assistant features across its entire ecosystem without the risk of cannibalising profits." Google has a problem with this, whereas Meta – perhaps not so much. And although the name has been redacted, everything points to Meta.
And finally, an important piece of information: OpenAI supports regulations that will allow users to set ChatGPT as their default assistant – instead of, for example, the one built into the phone or system by the manufacturer.
It started with a chat with a bot. Now it seems that ChatGPT will soon manage our meetings, talk to people for us, and be… something like a digital alter ego.
