The GPT-6 roadmap, shown with the ChatGPT logo, highlights memory, personalization, neutrality, and privacy as OpenAI’s next priorities. Image Source: ChatGPT-5
Sam Altman Teases GPT-6, Says “People Want Memory”
Key Takeaways:
Meta has launched its fourth AI restructuring in six months, signaling ongoing strategy shifts.
The new unit, Meta Superintelligence Labs (MSL), will be split into four groups, including a new TBD Labs (name to be determined later) focused on foundation models.
Scale AI founder Alexandr Wang, now Meta’s chief AI officer, will lead TBD Labs.
CEO Mark Zuckerberg is investing heavily in artificial general intelligence (AGI) and large-scale AI data centers.
Meta recently raised its capital expenditures forecast to $66–72 billion, with data infrastructure driving costs higher.
GPT-6 on the Horizon
Speaking to reporters in San Francisco, OpenAI CEO Sam Altman offered a rare look at the company’s product roadmap. He confirmed that GPT-6 is in development, though without a release date, and emphasized it will arrive faster than the gap between GPT-4 and GPT-5.
The biggest change, Altman said, will be memory. Future models will not only respond to users but adapt to them — remembering preferences, routines, and quirks to make interactions more personal.
“People want memory,” Altman said. “People want product features that require us to be able to understand them.”
Personalization and Neutrality
Altman stressed that OpenAI is working with psychologists to refine how memory features impact user well-being. While data from these studies has not been released, Altman suggested it may be shared in the future.
He also said upcoming versions of ChatGPT will comply with a Trump administration executive order requiring AI systems used by the federal government to be both ideologically neutral and customizable.
“I think our product should have a fairly center-of-the-road, middle stance, and then you should be able to push it pretty far,” Altman said. “If you’re like, ‘I want you to be super woke’ — it should be super woke.”
He added that if a user wanted the model to be conservative, it should reflect that as well.
Rocky Rollout of GPT-5
Altman’s comments follow a difficult launch of GPT-5, which drew criticism from users who described it as colder, less helpful, and less connected than GPT-4.
“I like the new one much better,” Altman said, acknowledging the rollout was mishandled. He confirmed OpenAI has since pushed a tone update to GPT-5, making it “much warmer.”
Privacy Concerns with Memory
Altman described memory as his favorite feature launched this year but admitted that temporary memory is not encrypted, raising privacy concerns. Without stronger safeguards, sensitive information could be at risk.
He confirmed that encryption “very well could be” added in the future but offered no timeline.
Altman highlighted the importance of privacy protections for queries involving medical or legal information.
“It’s in society’s interest for people to get good medical advice … good legal advice,” he said. “And if you can get better versions of those from AI, you ought to be able to have the same protection for the same reason we decided you could get them from a doctor or a lawyer.”
Beyond Chatbots: Neural Interfaces and Robotics
Altman also spoke about OpenAI’s interest in brain-computer interfaces, saying he imagines a future where users can “think something and have ChatGPT respond.”
“There are a few areas adjacent to AI that I think are worth us doing something, and this is one of them,” he said, naming energy, novel substrates, robots, and faster ways to build data centers as other areas of focus.
For now, Altman said OpenAI’s core consumer product remains ChatGPT, which he personally uses for work and parenting. But he noted limits to the chat format:
“The models have already saturated the chat use case,” Altman said. “They’re not going to get much better. ... And maybe they’re going to get worse.”
Q&A: GPT-6 and OpenAI’s Roadmap
Q: What is GPT-6?
A: GPT-6 is the next major AI model from OpenAI, expected to arrive faster than the gap between GPT-4 and GPT-5.
Q: What will be the key feature of GPT-6?
A: Memory, enabling ChatGPT to remember preferences, routines, and quirks to make interactions more personal.
Q: How will GPT-6 handle political or ideological neutrality?
A: OpenAI will comply with a federal executive order, keeping the model neutral by default but allowing users to customize its stance.
Q: What happened with GPT-5’s rollout?
A: Users criticized GPT-5 for feeling colder and less helpful. OpenAI quietly issued a tone update to make it warmer.
Q: What are the privacy concerns with memory?
A: Temporary memory is not encrypted, raising risks for sensitive data like medical or legal queries. Altman said encryption may be added later.
What This Means
Altman’s remarks signal that GPT-6 will be more than just another upgrade — it will mark a shift toward personalized, memory-driven AI. That direction could redefine user trust and daily reliance on AI assistants.
At the same time, Altman’s candid acknowledgment of GPT-5’s flaws and the lack of encryption for memory highlights the fragility of user trust. Without stronger privacy safeguards, memory may be a breakthrough feature with significant risks.
By tying model development to government regulations, user customization, and long-term visions like neural interfaces, OpenAI is positioning GPT-6 as not just a product but a platform for personalization and adaptability in the AI ecosystem.
For the AI industry, this is a reminder that the race isn’t just about raw capability anymore — it’s about memory, trust, and the ability to adapt to people’s lives.
Editor’s Note: This article was created by Alicia Shapiro, CMO of AiNews.com, with writing, image, and idea-generation support from ChatGPT, an AI assistant. However, the final perspective and editorial choices are solely Alicia Shapiro’s. Special thanks to ChatGPT for assistance with research and editorial support in crafting this article.