Elon Musk Predicts AI Smarter Than Humans Next Year, Raises Tension in Tech Industry
Daniel Kim Views
OpenAI, Meta, GPT-5, and LAMA3 are about to be released
Musk xAI is also rumored to be pushing to attract $4 trillion in investment
Jensen Huang “AGI appears within 5 years.”
Professor Andrew Ng: “It’ll take 30 to 50 years.”
|
Recently, Elon Musk, CEO of Tesla, announced that artificial intelligence (AI), smarter than humans, could emerge as early as next year. He also suggested that an AI capable of solving problems on its own, like a human, could appear by 2025. The tension is mounting as OpenAI, which sparked the ChatGPT craze, and Meta, the operator of Facebook and Instagram, are expected to unveil next-generation AIs soon. However, some predict that developing an AI that surpasses human intelligence in a short period will not be possible.
According to the information technology (IT) industry on the 14th, Musk said in an interview with Nicolai Tangen, CEO of the Norwegian Pension Fund, on X (formerly Twitter) on the 8th (local time), “If we define general artificial intelligence (AGI) as an AI smarter than the smartest human, it could be possible, say, within two years, maybe next year.”
However, Musk did not provide concrete evidence, using expressions like probably and I think.
|
OpenAI and Meta are ramping up the tension as they prepare to launch next-generation AI models capable of inference and planning. Currently, AI only produces results optimized for human commands. However, if AI can infer and plan, it can provide appropriate answers through inference, even without learning-related data. This lays the groundwork for AI to evolve into AGI, capable of learning and growing on its own, solving various problems like a human. It brings us closer to the singularity, the point when AI surpasses human intelligence.
Musk is also accelerating AI advancement. It has been reported that Musk’s AI startup, xAI, is pursuing a capital raise of up to $4 billion. Recently, Bloomberg reported, citing documents sent to investors, that xAI is planning to raise between $3 billion and $4 billion, valuing the company at $18 billion. Some analysts suggest that if xAI successfully raises astronomical funds, it could speed up AGI development.
|
Some claim that AGI could emerge within five years. Dario Amodei, CEO of Anthropic, which developed Claude3.0, predicts that humans could use AI that performs at the human level in a quick 2-3 years. Jensen Huang, CEO of Nvidia, predicted last November that “AGI at the same level as humans will emerge within five years.” Sam Altman, CEO of OpenAI, predicts that AGI will be built within 4-5 years, while Shane Legg, co-founder and senior AGI scientist of Google DeepMind, sees a 50% chance of AGI development by 2028.
However, some predict that it will not be possible to develop in a short period. Andrew Ng, a professor at Stanford University in the U.S., known as one of the Four AI Kings, predicted in a roundtable discussion with Kakao employees last year that AI would take several more decades to advance to the level of AGI. Professor Ng emphasized, “The transformer is an excellent technology, but it cannot reach AGI. It will take 30-50 years from now to reach the AGI stage. We need another leap.” Yann LeCun, Vice President and Senior AI Scientist at Meta Platforms, Inc., said, “Human-level AI will come out someday, but not in 3 or 5 years,” adding, “It’s hard to predict when, and it may take longer than we expect.”
Most Commented