Content
summary Summary

Leading cloud providers and their customers are more cautious about generative AI. The focus is on cost, efficiency, and measurable value.

The buzz around generative AI seems to be cooling, at least a bit. According to The Information, several executives, product managers, and salespeople at major cloud providers - including Microsoft, Amazon Web Services, and Google - are cautious about their customers' willingness to invest in new AI services.

Reasons cited include high cost, lack of accuracy, and the difficulty of measuring the true value of generative AI in day-to-day operations.

Despite successful testing in small groups, some cloud customers have encountered problems when trying to roll out the technology to thousands of users, according to big cloud employees.

Ad
Ad

In a presentation to Amazon AWS salespeople, a Gartner analyst predicted that the industry is at the peak of the "hype cycle" surrounding large language models and other generative AI. He expects the euphoria to turn into Gartner's "trough of disillusionment." Eventually, however, the analyst expects widespread adoption.

How to make money with generative AI

Microsoft has seen an increase in cloud revenue from AI thanks to its exclusive partnership with OpenAI. However, Google and AWS have yet to see financial success from their recent investments.

Analysts estimate that Microsoft generated more than $1.6 billion from AI-related cloud services last year. This is largely due to sales of OpenAI software and OpenAI's use of Microsoft cloud servers.

OpenAI's investment in cloud computing capacity is estimated to be approximately $1 billion by mid-2023. OpenAI uses Microsoft's computing infrastructure, which means that a portion of Microsoft's investment of up to $13 billion will flow back to the company.

Google hopes to generate at least $1 billion this year in North America alone by selling AI services to cloud customers. This seems like a small amount compared to the $33 billion Google Cloud generated last year.

Recommendation

In addition to new AI services like Gemini, Google also offers older services, some of which have been around for nearly a decade, such as speech-to-text and image analysis. These services are included in the revenue expectations, reports The Information.

Google is also trying to sell generative AI directly to end users with its Gemini Advanced AI chatbot, which competes with ChatGPT. The company has integrated the most advanced version of the chatbot into a $20-per-month Google One subscription. So far, however, Gemini Advanced, as the new chatbot is called, has failed to make a mark, just like its predecessor, Google Bard.

ChatGPT dominates the market, but according to Similarweb, its web growth is also stagnating: the traffic record of 1.8 billion visits in May 2023 was not reached again. In February 2024, the ChatGPT website had about 1.6 billion visits.

OpenAI also offers ChatGPT and its AI models through a mobile app and API. These numbers are not included in the statistics and may show growth.

Ad
Join our community
Join the DECODER community on Discord, Reddit or Twitter - we can't wait to meet you.
Ad
Join our community
Join the DECODER community on Discord, Reddit or Twitter - we can't wait to meet you.
Support our independent, free-access reporting. Any contribution helps and secures our future. Support now:
Bank transfer
Summary
  • Executives and employees at major cloud providers such as Microsoft, Amazon Web Services, and Google are cautious about their customers' willingness to invest in generative AI services.
  • They cite high operational costs, lack of accuracy, and the difficulty of measuring real value.
  • One Gartner analyst predicted to AWS sales teams that the industry has reached the peak of the generative AI "hype cycle" and will now go through a "trough of disappointment" before achieving long-term adoption.
Sources
Online journalist Matthias is the co-founder and publisher of THE DECODER. He believes that artificial intelligence will fundamentally change the relationship between humans and computers.
Join our community
Join the DECODER community on Discord, Reddit or Twitter - we can't wait to meet you.