Bengaluru: IBM Chairman and CEO ARVIND KRISHNA observed that AI racing focuses on whether a larger model means better results. However, AI systems do not need to consume essentially extensive resources and do not need to invest much.
“I had a lot of conversations following the announcement of today’s IBM revenue, but Deepseek always came out …. For a long time, the AI race has a better result. It was a game of the AI model, but it was one of the other technologies. Krishna wrote that the storage and processing capacity were very expensive in the beginning of computing.
However, Krishna stated that these costs have plummeted through technological advances and the economy of the scale. “AI follows the same path. In IBM we have been on this journey for years. Efficient and appropriate models can bring exceptional value, that is the key to innovation. He said that the cost of the company was reduced by up to 30 times.
Last week, China’s AI Startup Deepseek had a dramatic impact on high -tech stocks around the world, attracting great interest among users. DeepSeek was established by Liang Wenfeng, which previously established hedge funds. The focus on the results is that you can create a Chatgpt class language model in just a part of Chatgpt’s cost. It is said that DeepSeek’s R1 AI model matches or defeats something like Chatgpt or Google Gemini with multiple parameters.
Krishna’s Linkedin post came after the fourth quarter of IBM. High -tech companies reported $ 62.8 billion in earnings in the 2024 fiscal year, increasing by 3 % with a certain currency. Genai Business has donated more than $ 5 billion since its launch and increased $ 2 billion compared to the previous quarter. IBM continues in the calendar year.
“Our AI portfolio has been adjusted to meet the diverse needs of enterprise clients, so that you can mix your own open models from models, IBMS, hugs, meters, and Mistral. Krishna in recent revenue calls. Was described.
