Cloud Vis a Vis Generative AI

Some random thoughts on comparing cloud evolution to generative AI.

Over the years the evolution of cloud computing from private to public to hybrid to multi-cloud was driven by a need for scalability, flexibility & cost-effectiveness. As companies sought to optimize their IT infrastructure, they adopted cloud computing solutions that best fit their needs.

Private cloud solutions involved building and managing an in-house IT infrastructure that provided cloud-like services then came public cloud solutions that involved using third-party providers like AWS, Google Cloud, or Azure. Hybrid cloud solutions combined private and public cloud services, allowing companies to leverage the benefits of both. Finally came Multi-cloud solutions involving using multiple cloud providers to achieve greater flexibility, resilience, and cost-effectiveness.

I believe Generative AI is still in its early stages of development, but its evolution is likely to be driven by similar factors of scalability, flexibility and cost-effectiveness. As companies seek to incorporate generative AI into their operations, they will need solutions that are adaptable to their specific needs and offer the right level of control and security. Take example of Morgan Stanley where they created virtual financial advisor to help their own financial advisors to drive greater efficiency using OpenAI.

Thus in the early stages of generative AI we are seeing how companies are relying on third-party providers like OpenAI or Hugging Face to provide pre-trained models and API services. As generative AI becomes more integrated into business operations, companies may start to develop their in-house LLMs to create and train their own models which is very costly process right now but maybe we have a breakthrough soon.

Now as generative AI becomes more widespread, there may be a need for hybrid solutions that combine both in-house and third-party models and mainly multi-cloud solutions that may also emerge, allowing companies to use multiple generative AI platforms to achieve greater flexibility and cost-effectiveness. I believe that will be the end goal here. Ultimately as this matures companies will seek out solutions that provide the right balance of flexibility, control, and cost-effectiveness, and that can be integrated seamlessly into their existing IT infrastructure.