The Pessimistic View of AI Development

I know we are just getting started on this, but is it even possible to catch up? So for example, if you take very successful AI models like charge GPT, they are being used to train even more powerful algorithms. At the same time, computing power takes a long time to build up - and it's very expensive. You basically need also a business model as the cloud hyperscalers show. How will all these inter-independence play out in the long run in terms of getting a head start in this international race? Well, so I have both a pessimistic view and an optimistic view. And which one is correct? I'm still at the beginnings of researching all this.