OpenAI develops GPT-6 to personalize user experience

OpenAI CEO Sam Altman said the company is working on GPT-6 with a focus on faster release and greater personalization compared to previous models.

Altman discussed these plans with reporters in San Francisco last week, saying future versions of ChatGPT will allow users to define tone, personality, and create chatbots that match individual preferences.

He highlighted enhanced memory as a key feature, enabling ChatGPT to remember user routines and adapt responses, though he acknowledged ongoing privacy concerns due to unencrypted temporary memory.

Altman said that additional encryption could be added but did not provide a timeline.

He also said future ChatGPT versions will comply with a US executive order requiring AI used by the federal government to be ideologically neutral and customizable.

Altman addressed criticism of GPT-5’s launch, noting an update was made to improve its tone.

He added that OpenAI is exploring areas like brain-computer interfaces, energy, and robotics, but ChatGPT remains the company’s main product.

🔗 Source: CNBC

🧠 Food for thought

1️⃣ AI memory creates a privacy paradox that companies are still solving

Altman’s focus on memory as GPT-6’s key feature highlights a fundamental tension in AI development that users are actively discussing.

While personalized AI experiences require storing user data across sessions, this creates significant privacy risks that concern users2. OpenAI community members specifically worry about unauthorized access and potential data breaches when ChatGPT retains personal information2.

The company acknowledges this challenge directly. Altman confirmed that temporary memory isn’t encrypted, meaning sensitive information could be exposed without stronger safeguards1.

Privacy advocates recommend using accountless versions of chatbots and avoiding sharing sensitive information, showing that even privacy-conscious users recognize the trade-offs3.

This reflects broader industry discussions about whether AI memory enhances user experience enough to justify the privacy risks, with technical explanations showing how these systems analyze conversations to identify patterns worth remembering4.