Competition for Artificial Intelligence Models

The race to release cutting-edge artificial intelligence (AI) models is moving at a rapid pace, with major players like Amazon.com and Samsung Electronics joining forces to challenge market leaders Microsoft and Google.

Microsoft, known by its ticker MSFT, gained an early advantage in the AI race through its investment in ChatGPT developer OpenAI. Microsoft's latest OpenAI model, GPT-4, is playing a crucial role in powering the company's commercial AI applications.

Google-parent Alphabet (GOOGL) emerged as a chief rival to Microsoft's partnership with OpenAI. Alphabet has developed an in-house AI model called PaLM, which is now in its second version. The Wall Street Journal reported in July that Google is also working on an AI model called Gemini, expected to pose a significant challenge to GPT-4. Gemini is projected to be released in the coming months.

However, the competition might witness a major shakeup as Amazon (AMZN) doubles down on its own AI model. Codenamed 'Olympus', Amazon's AI model has the potential to surpass both OpenAI's GPT-4 and Alphabet's technology, according to sources familiar with the matter as reported by Reuters.

Amazon has yet to comment on these reports as of early Wednesday.

Comparing different AI models can be a complex task, as their performance can vary across various tasks. Nevertheless, the most commonly used metric for comparison is the number of parameters, which reflects the size and complexity of the model.

In this regard, Amazon's Olympus is set to dwarf its competitors, boasting an impressive two trillion parameters compared to the reported one trillion of GPT-4, as disclosed by Reuters. OpenAI has not made any public statements regarding the size or specifics of GPT-4's training.

The race for AI dominance continues unabated, with each major player striving to innovate and outperform their rivals. Only time will tell which company will emerge as the trailblazer in the fascinating realm of artificial intelligence.

Google's PaLM Models: A Closer Look

Google has been making waves in the AI field with its PaLM model, which originally boasted a staggering 540 billion parameters. While the current version, PaLM 2, is smaller in size, Google claims it is even more efficient. This serves as a gentle reminder that while parameter count is one way to gauge model complexity, performance is equally dependent on speed and cost efficiency.

The AI Race: Amazon and Beyond

In premarket trading, Amazon shares remained unchanged. Conversely, Microsoft experienced a slight increase of 0.1%, while Alphabet saw a decrease of 0.1%.

While Amazon has been a prominent force in the AI race within the United States, international competition has also been heating up. Case in point, South Korea's Samsung Electronics (005930.Korea) made waves when it unveiled its very own AI model, aptly named Gauss, at a recent company event. Although technical specifics were not disclosed, Samsung highlighted Gauss's ability to generate text, images, and assist with coding.

Initially designed for internal use, Samsung has plans to expand the application of Gauss to several of its products in the near future. This development raises the intriguing possibility of Galaxy smartphones harnessing AI capabilities, further fortifying Samsung's rivalry with Apple and its coveted iPhones. While Apple has remained rather tight-lipped about its own AI ambitions, market expectations suggest the tech giant will invest heavily in the development of this technology, including a revamped iteration of its virtual assistant, Siri.

In premarket trading, Apple shares experienced a slight dip of less than 0.1%, while Samsung stock faced a 1.4% decline during Wednesday's local trading session in South Korea.