OpenAI, the company that shook up the AI world with ChatGPT, is already working on its next model, GPT-6. CEO Sam Altman shared more information about GPT-6 at a press conference in San Francisco.
While GPT-5 has only just been rolled out with mixed success, this announcement signals the company’s desire to speed up the pace of its innovations. Although no release date has been specified, Sam Altman hinted that the gap between GPT-5 and GPT-6 will be shorter than between previous versions.
But what will make GPT-6 so different from other models?
The Yiaho team, which will make GPT-6 available for free as soon as it’s released, looks back at OpenAI’s ambitions for this future model.
GPT-6: An AI that learns to know you
One of the biggest promises of GPT-6 lies in its ability to deliver a truly personalized experience. Unlike its predecessors, this model won’t just answer user requests: it will adapt to their preferences, habits, and individual traits.
This personalization is built on a key feature: memory.
According to Sam Altman, an AI that can remember past interactions and each user’s specific tastes could transform how we interact with AI day to day. Whether it’s answering work questions, helping solve parenting dilemmas, or suggesting creative ideas, GPT-6 aims to become a tailor-made digital companion.
This improved memory, already seen as Sam Altman’s favorite innovation this year, will allow the AI to better understand users’ needs.
For example, a user could customize their own chatbot, shaped to their tastes or specific needs—whether that’s a humorous, professional, or even ideologically oriented tone.
Also read: OpenAI improves GPT 5 for a more “human” AI
Lessons learned from GPT-5
The rollout of GPT-5 hasn’t been smooth. Many users voiced their disappointment on social media, describing the model as less warm and less intuitive than GPT-4.
In response to these criticisms, OpenAI released an update that makes interactions with GPT-5 more natural and engaging. Sam Altman acknowledged that the initial launch was mishandled, but he remains confident about the model’s improvements and GPT-6’s future reception.
The privacy question
However, introducing advanced memory raises questions about data protection.
For now, temporary data stored by ChatGPT is not encrypted, which could expose sensitive information—especially in areas like health or law. Sam Altman admitted that encryption measures could be considered, although no specific timeline has been announced.
He also stressed the importance of ensuring protections similar to those provided by healthcare professionals or lawyers, especially for medical or legal advice delivered by the AI.
GPT-6 and the future of AI for OpenAI
Beyond GPT-6, OpenAI is exploring innovative areas like brain-computer interfaces—an idea Sam Altman calls “fascinating.” He imagines a future where you could simply think a request and ChatGPT would answer instantly.
The company is also looking into other sectors, such as:
- energy,
- robotics,
- new technology platforms,
- building data centers.
OpenAI aims to become one of the leading companies of the second half of the 21st century, a period that will likely be dominated by artificial intelligence and robotics.
A neutral and customizable AI
In 2026, debates around AI ethics and neutrality will intensify. Sam Altman also mentioned a recent directive from the Trump administration requiring AI systems used by the federal government to be ideologically neutral and customizable.
OpenAI plans to align with this requirement by offering a model that adopts a moderate stance by default, while allowing users to configure it to their preferences—whether progressive, conservative, or otherwise.
A collaboration with the humanities?
To refine its models, OpenAI is working with psychologists to assess the impact of its technologies on users’ well-being. Although the data from these studies has not yet been made public, Sam Altman suggested it could be released in the future.
OpenAI wants an AI that is not only high-performing, but also attentive to users’ emotional and psychological needs.
Toward an AI at the heart of everyday life
For Sam Altman, ChatGPT must become an even more flexible tool, integrated into everyday life. He says he relies on AI himself for a range of tasks, whether for work questions or advice on raising his children.
However, he acknowledges that current models have reached a certain plateau in conversational use.
The future, he says, lies in advanced features like memory and personalization, which will make AI more relevant and useful. As OpenAI accelerates its pace of innovation, the world is eagerly waiting to see how GPT-6 will redefine our relationship with technology—or whether this model will flop like GPT-5!
Source: CNBC
