Elon Musk, Tesla’s CEO, made a bold prediction in an interview on Monday, suggesting that artificial intelligence (AI) smarter than the smartest human could be developed as early as next year or by 2026. Commenting on the X-space despite multiple challenges of technological glitches , Musk discussed a variety of topics, including the limitations facing artificial intelligence due to the availability of electricity.
During an interview with Nicolai Tangen, CEO of the Norwegian wealth fund, Musk shared insights on the progress of Grok, an AI chatbot developed by his startup xAI. He mentioned that training of the next version of Grok is expected to take place in May, but emphasized challenges related to the shortage of advanced chips.
Musk, who also co-founded OpenAI, expressed concerns about OpenAI straying from its original mission. Last year, he founded xAI, positioning it as a competitor to OpenAI, which he sued for allegedly prioritizing profit over improving humanity through the development of artificial intelligence.
Tesla’s CEO revealed that training the Grok 2 model required approximately 20,000 Nvidia H100 GPUs, and future versions will require even more, estimating it to be as many as 100,000 Nvidia H100 chips. He emphasized, however, that apart from chip shortages, electricity supply will become a key factor in the development of artificial intelligence in the next few years.
Turning to the automotive industry, Musk praised Chinese automakers, describing them as the most competitive in the world. He repeated his earlier warnings that Chinese rivals will outperform global competitors unless trade barriers are introduced.
Touching on recent labor issues, Musk referred to the union strike against Tesla in Sweden, pointing out that the situation has improved. He mentioned that there have been discussions between the Norwegian sovereign wealth fund, Tesla’s main shareholder, and the EV company’s management about recent events.
Musk’s remarks shed lightweight on the landscape of artificial intelligence development and the challenges facing both the technology and automotive industries.
(with support from Reuters)
Posted: Apr 9, 2024 09:47 EST
