Will Microsoft lead the AI world?

Microsoft’s ChatGPT integration to help the company outpace Google?

Microsoft recently announced a paid subscription model called ChatGPT Plus that offers priority access, faster response time and Microsoft’s own web search engine Bing will soon be powered by the next version of ChatGPT. 

As a premium subscription, the company's ChatGPT integration may be able to provide sophisticated answers to elicit new ideas and produce a more seamless internet search engine experience, which could pose a threat to market leader Google. 

Google’s answer to ChatGPT?

Google, on the other hand, is not too far away from entering the AI space. The business has been developing its own language models to compete with ChatGPT, such as BERT and Transformer-XL. 

Google has been attempting to incorporate these models, which are intended to give increased accuracy and efficiency, into a variety of products and services

What are Microsoft and ChatGPT building?

Microsoft and OpenAI are working together to bring GPT-3's cutting-edge AI capabilities to businesses and organizations, making it easier and faster for them to develop AI solutions.

The two companies are also working on a number of joint AI initiatives, such as developing new AI models and tools that can be used in various industries.

One of the key initiatives of this partnership is the integration of OpenAI's GPT-3 technology into Microsoft's Azure AI platform. This integration will make it easier for developers to build AI solutions on Azure, and will also allow businesses to access GPT-3's advanced AI capabilities without having to invest in their own infrastructure.

How has Microsoft shares performed?

OpenAI/ChatGPT – How does it work and how will it evolve? 

ChatGPT is currently accessible to everyone at no additional cost. Experts believe the capability of efficiently leveraging such services as augmentation to professional work will become the main explanatory variable of productivity in the coming years. Accuracy will continue to be a challenge for the next couple of years,  even though a new user database, new iteration and specialized APIs will help improve OpenAI models. 

How Does The Hardware Scale From Machine Learning? 

People tend to think that LLMs (large language models) are extremely expensive to train. But once a user base is scaled, such as the case with ChatGPT, inference cost will exceed the cost of training when the model is deployed. As generative AI applications gradually take market share from traditional search engines, the demand for hardware will increase given the significantly higher computing power needed for such models. 

Do You Expect Substantial Capex Spending For ChatGPT Like Models And Future Demand?

In the near future, cloud-hosted LLM will still be the mainstream solution, creating a new TAM for semiconductors and hardware. Over the long term, edge computing and a surrogate model that runs on existing local devices will be better alternative solutions as customers will own the hardware and will not need to pay every time for a new query.

Companies That Can Benefit from ChatGPT?

For many of our preferred plays only a small part of their business is currently related to AI, but as technology progresses, they could become the "picks and shovels" for building AI infrastructure: 

1)Logic and memory semiconductors for HPC/datacenter: TSMC, Samsung and Hynix

2) networking equipment and related hardware: TFC Optical, Inspur and ABF substrates

3) software and applications: Baidu and Sony. Competitive Dynamics? Traditional search engines will face fierce competition from LLM AI. Potential winners would be companies willing to aggressively integrate the technology into existing products and organizational frameworks.

This is not investment advice. Investments in the securities market are subject to market risk, read all the related documents carefully before investing. Past performance is not indicative of future returns.