Sam Altman Talks AI Evolution at TED2025

OpenAI CEO Sam Altman recently took the stage at TED2025 to provide a deep dive into the cutting-edge developments in artificial intelligence, specifically focusing on ChatGPT, AI agents, and the ethical dilemmas surrounding superintelligence. With OpenAI’s innovations impacting industries across the board, this conversation with TED curator Chris Anderson offered key insights into what lies on the horizon for AI and humanity.

---

From Image Generation to Intelligence and Consciousness

The discussion began with a live demonstration of Sora, OpenAI’s advanced image and video generation tool, which is integrated with GPT-4o, the company’s new AI model. Designed to combine generative art and advanced reasoning capabilities, Sora wowed the audience with its ability to visualize concepts like "intelligence versus consciousness." Altman explained that Sora’s power stems from its integration with GPT-4o, highlighting how the new generation of AI platforms merges visual media with advanced machine intelligence.

The conversation also touched on how AI raises questions about the rights of creators. The ability to generate art “in the style” of existing artists—both living and deceased—brings up ethical and economic concerns. Altman conceded the need for new business models, emphasizing the potential to create systems that allow artists to consent to having their styles used, with a share of the revenues generated.

"The creative spirit of humanity is an incredibly important thing, and we want to build tools that lift that up,” Altman said, expressing confidence that more powerful AI tools will lead to greater creative output rather than stifling human creativity.

---

Transforming Work Through AI Tools

The conversation pivoted toward AI’s potential to transform industries, particularly knowledge-based professions like consulting and software development. Altman acknowledged that the rapid advancements in AI are raising concerns about job security. However, he reiterated that past technological revolutions show that new tools often elevate human productivity, empowering workers to achieve more.

For example, Altman revealed how OpenAI's models are already reshaping professions. One of the latest features in ChatGPT, named Memory, allows the AI to retain details about a user’s preferences and past interactions across sessions. This functionality fosters a personalized experience, with the AI evolving into a seamless extension of the user’s needs over time.

"Over the course of your life, you’ll talk to ChatGPT, and some day, maybe if you want, it’ll be listening throughout the day," Altman said.

The AI tool’s ability to deliver tailored outputs based on the user’s activity has the potential to revolutionize personal productivity and decision-making.

---

Addressing Concerns in the Creative Industry

Anderson raised a critical point about intellectual property and the AI domain: Should creators have the right to control their work being used in AI-generated outputs? Altman acknowledged the divide within the creative community, noting that some artists view AI as a threat to their livelihood, while others embrace it as a groundbreaking tool for innovation.

Altman touted OpenAI's stance of not allowing images or text prompt-generated outputs to mimic living artists without their consent. He suggested developing new revenue-sharing models that accommodate scenarios where creators opt in, allowing AI to generate works in their style while ensuring they receive fair compensation. As Altman noted, the challenge lies in figuring out practical and ethical mechanisms to determine and distribute revenue when creative inspiration is involved.

---

The Potential and Risks of Open-Source AI

The rise of open-source AI also featured prominently in the dialogue. Anderson mentioned DeepSeek, an open-source competitor, and questioned how OpenAI views its impact on the AI landscape. Altman struck a conciliatory tone, noting that while open-source AI can sometimes lead to unpredictable applications, it also has an important role in the broader ecosystem of AI development.

Notably, Altman revealed that OpenAI held its first community session to discuss the forthcoming release of its own open-source model, which he claimed would be more powerful than existing open-source options. He acknowledged that OpenAI was initially late to adopt the open-source model but emphasized that the company is now committed to building an exceptional, near-frontier open-source offering.

---

Scaling Up: ChatGPT’s Staggering Growth

One of the most eye-catching revelations from the discussion was the exponential growth of ChatGPT. Altman disclosed that the platform has reached 500 million weekly active users and continues to expand rapidly, even doubling its traffic in just a span of weeks.

As the platform grows in popularity, OpenAI faces immense pressure to keep up with user demand. Altman quipped about the intense competition for computing resources, likening himself to someone spending all day "calling people and begging them to give us their GPUs."

Despite the challenges of scaling, Altman remains confident that OpenAI is dedicated to delivering not just the most advanced AI models but also the best user experience. This, he believes, will differentiate OpenAI’s products in a marketplace that is expected to see increasing competition from other intelligent models.

---

The Future of AI: Science and Agentic Systems

Altman expressed his excitement about the potential for AI to usher in groundbreaking scientific discoveries. Reflecting on emerging fields like AI-assisted research in medicine and physics, he suggested that AI could play a critical role in advancing solutions for diseases and other global challenges.

Agentic AI is also poised to revolutionize software development, according to Altman. These systems will enable engineers to undertake and complete projects at a pace once deemed impossible. "This has been one of my big ‘feel the AGI’ moments," he said, referring to seeing engineers leveraging new AI models to achieve in hours what once took years.

---

Addressing AI Safety and Risks

When asked about the risks of AI, Altman highlighted several areas of concern: the misuse of models for bioterrorism, cybersecurity vulnerabilities, and potential loss of control if models exceed human intelligence. While emphasizing the rigorous preparedness framework OpenAI uses to ensure safety, he recognized that the rapid growth and power of AI require constant vigilance.

"You don’t wake up one day and say, ‘We didn’t have any safety process in place. Now, we think the model is really smart, so now we have to care about safety.’ You have to care about it all along this exponential curve," Altman said.

---

Practical Takeaways from Sam Altman’s Discussion

• Image-Generation Models Are Transforming Creativity: Tools like Sora, powered by GPT-4o, are breaking down barriers for artists and creators, enabling new levels of creative output.
• AI Offers Unprecedented Productivity Boosts: Features like ChatGPT’s Memory could redefine how professionals work by creating personalized, seamless workflows.
• AI Innovation Faces Ethical Hurdles: OpenAI is actively seeking solutions to address ethical concerns in AI-driven creative processes, especially intellectual property misuse.
• Open-Source Competition Looks Inevitable: OpenAI is planning its own advanced open-source AI model to adapt to a growing demand for transparent and communal AI development.
• Scientific Discovery and Medicine Could See Breakthroughs: AI is transforming the way scientists conduct research, potentially unlocking major innovations in everything from physics to medicine.
• AI Safety Remains a Top Priority: Altman stressed the importance of proactively integrating safety measures into the development lifecycle to prevent misuse.

While the pace of AI advancements is staggering, Altman remains optimistic about the future, proposing that AI systems will amplify human creativity and productivity, provided ethical challenges are proactively addressed.

---

FAQs about AI, Creativity, and Ethics

What is GPT-4o, and how does it differ from earlier OpenAI models?
GPT-4o is OpenAI's latest generation AI model, which integrates image and video generation capabilities like Sora. It combines advanced reasoning with creativity, representing a leap forward in the capabilities of AI models.

How does OpenAI handle ethical concerns regarding creative theft?
OpenAI restricts its image generation tools from mimicking the work of living artists without consent and is exploring new revenue-sharing models to ensure fair compensation for creators if their work is used for inspiration.

What is OpenAI's approach to AI safety?
OpenAI has implemented a preparedness framework to identify and address risks before releasing new AI models. This framework evolves alongside advancements in AI capabilities to ensure safety at every stage of development.

What are agentic systems, and how will they impact work?
Agentic systems represent the next evolution of AI, where software models will autonomously perform tasks and significantly accelerate processes like software development.

What risks does Sam Altman consider most pressing regarding AI?
Altman identifies bioterrorism, cybersecurity challenges, and potential loss of control over self-improving AI models as the most critical dangers tied to advancing AI technology.

---