The future of artificial intelligence (AI) promises substantial rewards for tech giants like Microsoft and Alphabet in due course. However, both companies emphasize that significant investments must precede any notable gains, as they stated on Tuesday.

Microsoft reported a notable surge in costs as it constructed new data centers to bolster its AI capabilities. This trend is projected to continue as the company acquires chips from prominent suppliers like Nvidia Corp to empower these data centers.

The expenses incurred by Microsoft in AI are twofold, according to analysts. First, the company aims to power its own products, including the forthcoming $30-a-month Copilot AI assistant. Second, Microsoft aims to cater to businesses seeking to utilize its Azure cloud computing services to develop AI-based products.

The executives at Microsoft disclosed that the service is expected to contribute the majority of its revenue in the latter half of fiscal year 2024, culminating on June 30.

Ben Bajarin, the CEO and principal analyst of Creative Strategies, explained the scale of their chip acquisition: “They’re procuring a plethora of H100s,” referring to Nvidia’s flagship chips designed for AI purposes. He also predicted a similar trend for Amazon in the near future, as both companies dominate the cloud market for AI system training.

Alphabet, on the other hand, has managed to maintain cost containment, although not indefinitely. The company’s Chief Financial Officer, Ruth Porat, soon to become president and chief investment officer, attributed the lower-than-expected second-quarter capital expenditure to delays in data center construction.

While Google might have invested over $200 billion in AI over the past decade, Scott Kessler, the global sector lead for technology media and telecommunications at Third Bridge, asserted that a significant portion of these investments might not be fully recognized by users and investors.

Analysts recognized one advantage Google holds – its in-house custom chip called the Tensor Processor Unit (TPU), which efficiently reduces costs associated with AI tasks.

James Cordwell, an analyst at Atlantic Equities, suggested that Microsoft might be aggressively acquiring Nvidia chips because it lacks its own silicon alternatives.

On the other hand, Google acknowledged that it will rely on chips from external vendors alongside its in-house solutions. Ruth Porat cautioned that this expenditure could impede profit margins and growth.

Gene Munster, the managing partner at Deepwater Asset Management, highlighted that both Microsoft and Google conveyed a consistent message about reaching an inflection point with their AI investments. However, the investors at Microsoft seemed more eager for additional progress.

In conclusion, artificial intelligence holds tremendous potential for tech giants, but they acknowledge the need for substantial investments before reaping tangible benefits. Microsoft’s approach involves acquiring chips and constructing data centers, while Google’s custom TPU chip helps streamline costs.

Both companies foresee a pivotal turning point, though they differ in their strategies, sparking interest among investors. As technology continues to evolve, the future of AI remains a focal point for these industry giants.