Former high-ranking $MSFT employee shares his view on AI, $MSFT, and the cloud industry:
- The way he sees the AI space when it comes to Big Tech is that everyone is looking at it from a different angle. OpenAI was trained on whatever was available on the internet; $GOOGL Gemini's data set was its search and YouTube library. $AMZN used data sets from its buyers, consumers, and merchants. $META trained their AI from social media assets, trying to understand consumer behavior.
- He believes that GenAI presents a more compelling reason for companies to move the on-prem workloads to the cloud.
- He doesn't think genAI will be a big reason to switch between the hyperscalers—at least not yet. He thinks companies will layer AI on top of the cloud provider where the data is already located.
- In his view, even though $NVDA is dominant right now, we will start to see the growth in internal ASICs made by the hyperscalers. If they are in single digits in terms of the footprint right now, he thinks they will grow to double digits. Short-term $NVDA is still the king for the next two years; beyond that, the hyperscalers have money to do what they want.
- For inference, you do not need that much computation as it is a much easier task than training.
Disclaimer: Investing carries risk. This is not financial advice. The above content should not be regarded as an offer, recommendation, or solicitation on acquiring or disposing of any financial products, any associated discussions, comments, or posts by author or other users should not be considered as such either. It is solely for general information purpose only, which does not consider your own investment objectives, financial situations or needs. TTM assumes no responsibility or warranty for the accuracy and completeness of the information, investors should do their own research and may seek professional advice before investing.
