GPT-5 leaks: New interview between openAI founder Sam Altman and Lex Fridman! GPT-5, Sora, board farce, Musk, Ilya, AGI

On March 18th, 2024 North American time, Sam Altman was interviewed by Lex Fridman. In this conversation, Sam Altman shared his insights on the future development of artificial intelligence, sora, Ilya, AGI, and OpenAI, as well as his insights on humans and technology.

The entire discussion covered multiple areas, from internal challenges of OpenAI, legal disputes with Elon Musk, to thinking about Sora, GPT-4, GPT-5, and AGI (Artificial General Intelligence).

Introduction

At the beginning of the interview, Sam Altman put forward a viewpoint about the future. He believes that computing power will become the most valuable resource in the world and may even become a global "currency". He foresees that in the near future, we will have very advanced artificial intelligence systems, which will attract great public attention and discussion.

He also emphasized that in the pursuit of artificial general intelligence (AGI), we should expect a huge power struggle, which will be inevitable and part of the road to AGI.

OpenAI Board Legend

Sam Altman detailed the turbulent period of the OpenAI board of directors, which was the most painful experience of his career. He described the chaos, shame, and frustration during this period, while also mentioning the great support and love he felt from it.

He mentioned that despite experiencing difficulties, he also believes that all of this has helped OpenAI establish a stronger and more resilient organizational structure, preparing for more challenges in the future. He also shared the personal impact of this experience on him, including how it changed his trust in people and his vision for the future.

Ilya Sutskever

When talking about Ilya Sutskever, Sam Altman expressed deep respect for Ilya. He praised Ilya's calmness and firmness in the face of OpenAI's difficulties, as well as his serious attitude towards AGI and its potential social impact.

Sam emphasized that although Ilya did not see the implementation of AGI, his thoughtfulness in ensuring the correct handling of these issues is admirable. He also mentioned Ilya's deep insight into technology and his long-term contributions to OpenAI culture and mission.

Elon Musk lawsuit

Sam Altman responded to Elon Musk's lawsuit against OpenAI, explaining OpenAI's position and expressing confusion about Elon's possible motives. He believes that Elon's lawsuit is to some extent an attempt to influence OpenAI, but he also admits that he cannot fully understand Elon's true intentions.

Sam said that although Elon was once one of the founders of OpenAI, they had differences in the direction and control of the company, which ultimately led to Elon's departure. He also mentioned that Elon hoped OpenAI would become part of Tesla, but OpenAI chose to remain independent.

Sora

Sam Altman introduced the Sora project, a new type of AI model developed by OpenAI that can generate and simulate complex 3D worlds. He mentioned Sora's ability to handle obstructions and physical laws, as well as its potential in video generation.

Although Sora still has many limitations, Sam is full of expectations for its future development. He also discussed Sora's potential applications in artistic creation and entertainment industries, as well as how it can promote new forms of creative expression.

GPT-4

During the discussion of GPT-4, Sam Altman shared his deep impression of this model, especially its abilities in programming, writing, and translation. He also mentioned how GPT-4 serves as a creative brainstorming partner to help people solve problems more efficiently.

Although GPT-4 has made significant progress, Sam believes that there is still a lot of room for improvement compared to future models. He also discussed the potential impact of GPT-4 in education and research, as well as how it can change the way we acquire and process information.

Memory and privacy

Sam Altman discussed OpenAI's idea of introducing memory functionality into AI models, which allows AI to better understand users and become more useful over time.

He emphasized the importance of user choice, that is, users should be able to control the information remembered by the AI model and delete it when necessary. He also mentioned the latent risks of this feature, including the spread of misinformation and privacy leaks, and how OpenAI strives to ensure that these risks are properly managed.

Q*

When talking about the mysterious Q project, Sam Altman did not reveal too many details, but he said that OpenAI is studying how to improve the reasoning ability of AI.

He hinted that although the outside world knows little about the Q project, the project is part of OpenAI's continuous innovation and exploration of new areas of artificial intelligence. He also mentioned OpenAI's emphasis on AI safety and ethics, and how they ensure that the development of AI technology does not bring unforeseen negative consequences.

GPT-5

Regarding GPT-5, Sam Altman did not provide a specific release date, but he confirmed that OpenAI will launch a new model this year. He mentioned that although GPT-4 is already very powerful, they expect future models to have greater leaps in performance.

He also discussed the challenges OpenAI faces in developing new models, including how to balance technological innovation and social responsibility.

700 billion US Dollars

Sam Altman discussed the huge demand for computing power in the future, predicting that the world will require unimaginable amounts of computing resources.

He believes that nuclear fusion will be the key technology to meet this demand, as it can provide almost unlimited clean energy. He also mentioned challenges in data center construction, supply chain management, chip manufacturing, and how OpenAI plans to address these challenges.

Google and Gemini

When talking about the competition with Google and the challenges of the Gemini project, Sam Altman said that OpenAI not only wants to replicate Google's success, but also wants to explore how to help people obtain and process information in new ways.

He mentioned the possibility of combining chatbots with search engines and how to create a better User Experience. He also discussed OpenAI's innovation in advertising and business models, and how they can avoid moving towards traditional ad-driven models.

The leap of GPT-5

Sam Altman shared his expectations for GPT-5, believing that future models will make significant progress in understanding complex problems and providing solutions.

He expects that with the development of technology, people will be able to program in natural language, which will change the way programmers work. He also discussed the potential applications of AI in fields such as education, healthcare, and scientific research, as well as how it can help humans solve some of the most difficult problems.

AGI

When discussing artificial general intelligence (AGI), Sam Altman stated that he does not expect AGI to be achieved in the short term, but he believes that within the next decade, we will develop systems that can significantly accelerate scientific discovery.

He believes that the development of AGI will be a gradual process rather than a sudden transformation. He also emphasized the profound impact of AGI on society and the economy, and the need for us to be prepared for it.

Alien

Finally, Sam Altman and Lex Fridman discussed the possibility of extraterrestrial civilizations. Sam expressed his hope for the existence of extraterrestrial intelligent life, believing that this would be a major discovery in human history.

It also explores how artificial intelligence can help us better understand the universe and our place in it, and how it might reveal the universality of intelligent life in the universe.