I had a lunch with a analyst of another investment firm recently and she said she thinks Generative AI will act as a tailwind for hyperscalers revenue, because this will make companies accelerate their move to the cloud. Other people in the Street think the same

Obviously, companies requesting OpenAI tokens will be a tailwind for Microsoft. OpenAI training GPT-5 is a tailwind. Same goes for the other guys. I am particularly skeptical of this GPU renting business because you’re selling commodity hardware to the most technically skillful organizations in the world, this shouldn’t be great business by first principles. This is significantly different than the traditional cloud business where you’re selling cloud resources to companies that do other stuff than using the cloud. The dozens of different products that AWS has make it an IT in a box, something that by first principles should be good business (and it is).

The question here is whether there’s an effect where companies use more meaningfully more traditional cloud computing. Some possibilities that all I think won’t happen and aren’t happening.

• This will create so many companies that they’ll all need to have their CRM, HCM, ERP, hosting and all the regular stuff companies have. (Unclear that just because you have a deal with AI Lab X for GPUs they’ll use your serverless offering)

• This will accelerate migrations to the cloud

• This will generate so much new data it will need to be stored

I’ll rely on the specialized media, particularly outlets like SiliconAngle, SemiAnalysis, and Stratechery, comments by the hyperscalers themselves, and my overall feeling. Perhaps the discussion in the comments change my mind.

I won’t bet. If I’m not convinced this market resolves to NO on June 30th 2025.