SemiAnalysis, a semiconductor research company, reports that Google’s upcoming generative AI tool is five times more powerful than the most advanced GPT-4 models currently on the market. Furthermore, they estimate that by the end of 2024, this AI tool could be 20 times more powerful than ChatGPT.
After companies like OpenAI and Midjourney jumped into the early lead with their AI tools, “the sleeping giant, Google, has awakened,” the company said in a report in late August.
However, Semianalysis speculates that the Alphabet-owned Internet search company may not want to make its new Gemini model publicly available because it could “neuter its creativity or existing business model.”
The Center for AI Safety said that although OpenAI has been ahead of Google for “many years,” the release of ChatGPT means Google has “significantly increased its AI investment.”
Pointing to Google’s “tremendous financial resources,” which dwarf OpenAI’s, the center said, “It’s no surprise that Google can quickly raise costs to compete with other leading AI labs.”
The proliferation of large language models (LLM) and generative artificial intelligence from late 2022 has warned that the technology could pose a threat to human life.
China has moved to regulate its AI businesses, while the British government is set to host an “AI Security Summit” in November at Bletchley Park, the site of World War II code-breaking operations.
Meanwhile, the United States Senate will host several meetings as part of the so-called AI Insight Forum. The first meeting is scheduled for September 13 and, Axios reports, will be joined by several industry leaders, including Google CEO Sundar Pichai, Meta’s Mark Zuckerberg, OpenAI’s Sam Altman, Microsoft’s Satya Nadella, Nvidia’s Jensen Huang. and Elon Musk of X, formerly known as Twitter.