Here is why Sam Altman called OpenAI’s o1 model deeply flawed

Here is why Sam Altman called OpenAI’s o1 model deeply flawed
HIGHLIGHTS

OpenAI's recent launch of its o1-preview model has stirred up quite a debate in the AI community.

OpenAI CEO Sam Altman has opened up about the o1-preview model, calling it "deeply flawed."

“o1-preview is deeply flawed, but when o1 comes out, it will feel like a major leap forward," he said.

OpenAI’s recent launch of its o1-preview model, designed to closely mimic human thought processes, has stirred up quite a debate in the AI community. While this innovation sounds impressive, it has also raised significant alarms among experts. Now, OpenAI CEO Sam Altman himself opened up about the o1-preview model, calling it “deeply flawed.” In an interview, Sam Altman said “o1-preview is deeply flawed, but when o1 comes out, it will feel like a major leap forward.”

Reflecting on OpenAI’s journey, Altman revealed that it took nearly 4.5 years for the company to launch its first product, GPT-3. Despite the buzz surrounding its release, he admitted, “Our first product, GPT-3, was not something we were particularly proud of at the time.” The team quickly recognized the model’s limitations, leading to the development of GPT-3.5, ChatGPT, and ultimately GPT-4. This process taught the company that its vision needed to encompass more than just making incremental improvements.

Also read: OpenAI’s new model might be capable of deceiving and cheating, suggests godfather of AI

“We often focus too much on ‘less bad,’ on fixing immediate issues, rather than dreaming of the breakthroughs that could propel humanity to the next level,” Altman was quoted as saying in a report by Analytics India Magazine. This perspective underlines OpenAI’s mission: to seek long-term progress for humanity rather than settle for quick solutions. Altman believes that this commitment to innovation is vital for shaping how AI interacts with society.

In another interview, he described the o1 model as the ‘GPT-2 stage’ of reasoning development, explaining, “I think of this as like we’re at the GPT-2 stage of these new kinds of reasoning models.” He emphasised that while o1 is still in its early development phase, users can expect to see significant improvements soon. “Even in the coming months, you’ll see it get a lot better as we move from o1-preview to o1,” he stated.

Altman also highlighted the importance of startup energy in redefining success in the industry. He pointed out that countries like Italy should use their rich history as a foundation for future AI innovations.

Ayushi Jain

Ayushi Jain

Tech news writer by day, BGMI player by night. Combining my passion for tech and gaming to bring you the latest in both worlds. View Full Profile

Digit.in
Logo
Digit.in
Logo