OpenAI has rolled out its ChatGPT Wallet via the Dots AI agent service, allowing users to streamline their purchases by linking a credit or debit card once. All transactions require biometric approval, ensuring security and user consent for each operation.
Biometric Approval at the Core
The AI agent manages the payment process, and users approve transactions through face recognition or fingerprint verification, creating a record of each transaction. This setup eliminates the need for re-entering card information for subsequent purchases, optimizing user experience while maintaining stringent security protocols.
OpenAI employs tokenization to safeguard payment data, substituting card numbers with digital tokens to prevent direct sharing of real card information with vendors or the AI agent. This technology is pivotal in enhancing transaction security and privacy.
Looking ahead, OpenAI is contemplating allowing users to predefine spending limits and categories, enabling the agent to execute transactions within these parameters autonomously. While the timeline for this feature remains uncertain, its implementation would mark a significant evolution in AI-driven financial management.
Sam Altman Discusses Human Supervision in AI Development
At the Semafor conference, OpenAI CEO Sam Altman addressed recursive self-improvement (RSI), wherein AI contributes to developing more advanced models. He acknowledged the difficulty in pinpointing a definitive starting point for this process, even in the context of artificial general intelligence (AGI). Altman’s statements were misinterpreted online as an announcement of AGI, a claim he refuted.
Altman expounded on OpenAI’s strategy to leverage current AI technology to accelerate advancements in next-generation models. When questioned about the possibility of AI independently designing and refining models without human oversight, his response was unequivocal:
“Certainly not.”
He emphasized that while technological capabilities could advance further, the company opts for a cautious approach due to lingering safety concerns.
Nvidia’s Ohio Guarantee and Huang’s Possible Return
During the event, Semafor’s host Reed Albergotti queried Altman about the potential return of Nvidia’s Jensen Huang. Drawing parallels to the open letter advocating Altman’s return to OpenAI in 2023, Altman expressed understanding for Huang’s situation, given his own experiences, and highlighted Nvidia’s early support for OpenAI. However, he abstained from speculating on Huang’s future, stating:
“I don’t think I’m sufficiently informed to answer that.”
Altman affirmed market confidence in Nvidia’s substantial guarantee for OpenAI’s data center project in Ohio, estimated at a ceiling of $105 billion. He conveyed optimism that the robust demand for AI will continue to support the industry despite leadership changes.
Future advancements in ChatGPT Wallet could pivot towards enabling predefined spending authority. Should OpenAI proceed with these enhancements, users can set purchase parameters, freeing them from individually approving each transaction. Furthermore, integrating stablecoin support could facilitate small-scale AI-initiated transactions, such as micro-payments for article access or API calls, rendering usage-based billing more feasible. Successful implementation depends on the willingness of service providers to accept stablecoins and maintain minimal transaction fees commensurate with these micro-payment amounts.



