Amazon recently announced that the company will launch its own generative AI service called Amazon Titan. This announcement is just the latest development in a noticeable trend of large tech companies investing in generative AI.
Generative AI, particularly text generation services, has grown in popularity in the past year or so, thanks to services like ChatGPT. These services have made it possible for anyone to take advantage of the power of pre-trained language models from their own browser.
The popularity of generative AI products cannot be overstated. The official website of OpenAI, the company behind ChatGPT, has already reached the top 20 most visited websites globally as of April 2023.
Given the demand, it seems natural for tech giants to invest so much in this space. Nonetheless, each player in Big Tech has their own distinct approach to AI, with some having already gained a significant advantage.
In this article. we’ll delve into some of the top products and investments that top tech companies have announced over the past few years. We’ll then try to see how these companies intend to leverage these investments for long-term profitability.
Generative AI as a Business Model for Big Tech
What is behind the trend of massive technology companies like Google and Microsoft creating their own language model services? The answer likely varies depending on the company.
The Evolution of Search Engines
The flexibility of language models in performing all sorts of tasks already makes it an excellent technology to invest in on its own. However, for Microsoft and Google, a lot of the interest seems to come from integration with their respective existing services.
Microsoft’s partnership with OpenAI has allowed their search engine Bing a chance to compete in the search engine market. The new Bing interface is trying to create more than just a list of relevant links.
Bing’s chat feature allows you to look up information more organically, with the search engine offering personalized replies for every user.
Microsoft is also planning on adding ChatGPT-like features for their Microsoft 365 Suite of business tools, including Word, PowerPoint, and Excel.
Similarly, Google’s Bard project seems to be the company’s response to the popularity and potential of ChatGPT. Bard also uses a chat interface that can answer user queries and generate text based on the user’s prompts.
Google has billions of dollars of revenue on the line if another search engine like Bing offers a better AI chatbot service. The threat of losing its market share and position as an industry leader in the AI space is certainly enough incentive to invest in language models of their own.
Currently, Microsoft’s first-mover advantage with its partnership with OpenAI makes it a better service than what Google has to offer for consumers.
However, considering both companies have massive amounts of capital and computing resources, we can safely say that both companies will continue to innovate their respective products.
AI as a Service
While the search-chatbot hybrid products by Google and Microsoft mostly serve the average consumer, Amazon seems to be taking a different approach.
A significant portion of Amazon’s revenue comes from its earnings with Amazon Web Services (AWS). The company’s AWS division focuses on cloud computing and cloud storage services primarily targeted to the business sector.
AWS is already ahead of Microsoft, Google, and China-based Alibaba in terms of market share as the largest cloud computing provider in the globe.
AWS includes dozens of highly specialized products that cater to various sectors. Amazon spends a large amount of its budget on research and development costs on these products to justify higher margins.
Offering cloud services that scale well and are leagues ahead of competitors allow Amazon to earn a significant user base in the business sector.
With that in mind, Amazon’s generative AI service Amazon Titan appears to be yet another product that will attract more customers to the AWS ecosystem. Amazon Titan includes two services: Titan Text, a text generation model which works similarly to ChatGPT, and Titan Embeddings, which allows clients to create more personalized search results and recommendations.
Amazon Titan falls under a broader suite of AI tools known as Amazon Bedrock. The Amazon Bedrock platform gives corporate clients API access to powerful AI models from partner startups and AI labs.
Conclusion
Tech companies, like all other commercial businesses, make decisions based on profitability. Since there is certainly interest in both the consumer and business sector for access to generative models, it is a wise move to invest in creating products that fulfill that demand.
However, these companies may soon bear a heavy responsibility. Since products powered by large language models are becoming more accessible than ever, the risks are becoming more apparent.
Language models are susceptible to hallucinations and may sometimes display biases and even hateful speech. Both Google and Microsoft have taken the initiative to address concerns about transparency, fairness, and interpretability in their respective generative models.
However, with the lack of proper legislation that address AI, there is still some skepticism on whether Big Tech will enforce ethical use of their AI.
Leave a Reply