'I Was Extremely Sceptical': OpenAI CEO Sam Altman On DeepSeek's Low-Cost AI Training
OpenAI CEO Sam Altman has opened up about DeepSeek's claims of training its AI model at lower costs than ChatGPT.
DeepSeek, the Chinese artificial intelligence start-up, has been making headlines topping global download charts after releasing its latest model, DeepSeek R1, in January. The company claimed that the latest AI model rivals the technology developed by OpenAI in terms of its capabilities and requires a lesser cost to create.
DeepSeek reportedly spent only $6 million in the final training of its V3 model using Nvidia’s H800 chips, an achievement that left industry leaders divided on the actual development cost. Many argued that the $6 million figure could only refer to the final phase, whereas the earlier stages, which include conducting experimental runs and determining model architecture, require more investment.
related stories
- OpenAI's Sam Altman Makes A Big U-Turn On Creating Low-Cost AI Models: Know Why
- ‘Elon Musk Insecure, Not A Happy Person’: Sam Altman Jabs Tesla CEO Amid OpenAI Bid
- How Elon Musk And Sam Altman’s Feud Is Turning Into A Long Legal War | Explained
- India Important Market for AI & OpenAI, Should Be Among Leaders of AI Revolution: CEO Sam Altman
Sam Altman On DeepSeek’s Low-Cost Training
OpenAI CEO Sam Altman has now broken silence on claims, indirectly taking a jibe at his company’s Chinese competitor. Speaking to the Hindustan Times during his ongoing trip to India, Altman opened up on a wide range of issues focused on artificial intelligence, including the right price of DeepSeek.
In response to a question asking about his response to the news of the Chinese AI model and the lower training costs Altman said, “I was extremely skeptical of the cost number. It was like, there are some zeros missing. But, yeah, it’s a good model, and we’ll need to make better models, which we will do."
The statement came in contrast to his previous comments, where he was seen praising DeepSeek’s R1 model. Taking to his official X handle, the OpenAI CEO stated that he feels thrilled about the competition and looks forward to pulling up new releases to compete with the low-cost Chinese AI model.
“DeepSeek’s R1 is an impressive model, particularly around what they’re able to deliver for the price. We will obviously deliver much better models, and also it’s legit invigorating to have a new competitor! We will pull up some releases," he tweeted. Altman concluded by teasing about the next-generation AI models from his company.
- Location :
- First Published: