Random Thoughts On Generative AI
OpenAI has more than 2 million developers that access GPT via APIs.
Groq has more than 250,000 developers using its Groq Cloud.
How many developers access Microsoft Azure in order to access GPT? That is unclear. I can understand why a Fortune 500 company would prefer to access GPT via Azure versus directly from OpenAI as Azure is an industrialized cloud with tried and tested SLAs, whereas OpenAI is new to the service side. I would expect for OpenAI to build out its cloud services business and invest in service delivery quality over time, although this is difficult to do when growth is rapid.
I expect for Apple to invest heavily in the quality aspect of the user experience for user flows that incorporate OpenAI.
Generative AI will succeed or fail based on the value-add created by various use cases, whether they be quality testing code, writing code, summarizing writings, transcription or powering AI agents to name a few (AI agents historically fell short because they lacked cross domain natural language capability, which is at the core of Large Language Models).
My view is that Generative AI will not create as much economic value as the Internet before it nor the forthcoming Blockchain / DLT revolution which will modernize various legacy systems such as payments (large and small transaction sizes), smart contracts, voting, health records, you name it.