According to BlockBeats, OpenAI CEO Sam Altman has outlined ambitious internal goals for the company, aiming to develop an automated AI research intern operating on hundreds of thousands of GPUs by September 2026, and a fully automated AI researcher by March 2028. While acknowledging the challenges in achieving these targets, Altman emphasized the importance of transparency due to the potential significant impact.
OpenAI's safety strategy is built on five layers: value alignment, goal alignment, reliability, adversarial robustness, and system security. Altman expressed excitement about the fidelity of Chain-of-Thought processes, though he noted its fragility and the need for clear boundaries and abstractions.
In terms of product development, OpenAI is focused on creating a platform that allows individuals and companies to capture substantial value. Currently, users can develop through OpenAI's API and ChatGPT application, with plans to eventually offer an AI cloud to support large enterprises. The company has committed approximately 30 gigawatts of computing power, with a total ownership cost over several years estimated at $1.4 trillion. OpenAI is satisfied with the growth in model capabilities and revenue.
OpenAI's mission remains unchanged for both its nonprofit and PBC arms: to ensure that artificial general intelligence (AGI) benefits all of humanity. Initially, the nonprofit pledged $25 billion towards health and disease treatment, as well as AI resilience, which encompasses technology safety, economic impact, cybersecurity, and more. The nonprofit is now able to deploy funds relatively quickly.
By 2026, OpenAI anticipates that AI systems may make minor new discoveries, with significant breakthroughs expected by 2028. Altman stressed the importance of scientific advancement and institutions that disseminate scientific achievements widely as key to improving quality of life.
Alex
