Musk Predicts AI Will Surpass Humans by 2025 – OpenAI, Microsoft, Google Updates

Musk: AI Could Exceed Human Intelligence Next Year

According to the Financial Times on April 8, Elon Musk said that if power supply and hardware can meet the growing demands of technology, new AI models will surpass human intelligence by the end of next year.

He speculated that around the end of next year we will have AI that is smarter than any individual human.

In an interview on X, Musk predicted that within five years AI could outmatch all humans on Earth.

Musk has long been optimistic about the development of artificial general intelligence (AGI), believing such tools will be extremely powerful and could defeat the most capable humans in any field.

However, his timeline for this breakthrough is earlier than his own previous forecasts.

In 2023 he predicted that by 2029 humanity would achieve “full” AGI, yet many of his bold predictions, such as fully autonomous Tesla driving and landing rockets on Mars, remain unfulfilled.

OpenAI Expands Custom Model Program for AI Solutions

OpenAI is extending its custom large‑model initiative, allowing enterprises to design their own tailored AI models. Launched in 2023, the program lets companies collaborate with OpenAI experts to train and develop large models for specific use cases.

The company is introducing new features such as assisted fine‑tuning and custom training to improve applications.

Assisted fine‑tuning goes beyond traditional methods, helping businesses build training pipelines and boost model performance.

Custom‑trained models are built using OpenAI’s base models and technologies (e.g., GPT‑4) to meet unique user needs, and OpenAI believes most companies will eventually adopt industry‑specific custom AI models.

This expansion comes as OpenAI aims for $2 billion in annual revenue and collaborates with Microsoft on new projects.

The new capabilities also make AI models more efficient and faster, reducing pressure on OpenAI’s infrastructure.

Developers using GPT‑3.5 now have additional fine‑tuning tools, including a quality‑comparison dashboard, third‑party platform support, and tool improvements.

Details on GPT‑4 fine‑tuning remain confidential.

Microsoft to Invest ¥4400 Billion in Japan to Strengthen Generative AI

U.S. tech giant Microsoft announced on the 9th that it will invest 4.4 trillion yen (about $29 billion) in Japan over two years to bolster the growing demand for generative‑AI infrastructure.

This represents Microsoft’s largest investment in Japan, aimed at building high‑speed data centers essential for generative AI and establishing the country’s first AI research hub.

Google Gemini Pro 1.5 Enters Public Preview on Vertex AI

Google announced at the Cloud Next conference in Las Vegas that Gemini 1.5 Pro, the most powerful model in its Gemini series, is now available in public preview on the Vertex AI platform.

Released in February, Gemini 1.5 Pro expands context length from 128 k tokens to up to 1 million tokens, where a token is a small piece of text (e.g., “fan”, “tas”, “tic”).

One million tokens correspond to roughly 700 k words or about 30 k lines of code—four times the input capacity of Anthropic’s Claude 3 and eight times that of OpenAI’s GPT‑4 Turbo.

The context window defines how much data the model considers before generating output, enabling it to answer questions like “Who won the 2020 U.S. presidential election?” using extensive background information.

US and Allies Collaborate on 6G Technology

The United States and major global partners have formally committed to establishing principles guiding the development of next‑generation 6G technology, a strategic move amid the race to shape future wireless standards.

This competition highlights the critical role of next‑gen wireless in economic growth and national security, especially between the U.S. and China, which recently launched the world’s first 6G test satellite.

An alliance of the United States, Australia, Canada, the Czech Republic, Finland, France, Japan, South Korea, Sweden, and the United Kingdom outlines a vision for open, reliable, and secure global 6G connectivity, emphasizing strong cybersecurity, privacy protection, and expanded access for developing nations.