For quite a while, Oracle's playbook has been about being a slow follower until a market is proven out and then making acquisitions of key companies to capture marketshare and to establish a presence. For example, buying PeopleSoft or Siebel made Oracle immediately relevant in HR and CRM while lining up customers to potentially buy more of their software from Oracle.
That strategy doesn't work in Cloud because Oracle can't buy its way into the market -- all the players have significantly higher valuations than Oracle. It shows in its current strategy of trying to slow Cloud adoption (raising DB license costs for running on AWS/Azure and pushing hard for on-premise deployments of its own hardware), in limiting choice (refusing to allow licensing of the database for Google, IBM, etc. and pushing all Apps customers to Oracle Cloud) and in strongarming customers (financially engineering deals). The reliance on a slow follower approach has limited Oracle's opportunity to, at best, being able to host its database and Apps customers.
As a result, Oracle's ability to grow and to obtain more profit/cash flow is severely limited. Given that its stock price is closely tied to this, Oracle is in a world of hurt and every possible lever is being pulled to find ways to be "good enough" to capture database and Apps workloads while simultaneously doing everything to prop profit margins and the stock price up. Paying well for any "key" talent in Cloud design or ML/AI. Buying back stock. Cutting teams/divisions that aren't very profitable. Trimming back benefits. Keeping investments in Cloud data centers and networking to the minimum possible. Setting up hubs filled with lower income college grads to slowly replace high earning, experienced field personnel.
Oracle isn't making a big bet on Cloud as it isn't anywhere close to doing what the market leaders are doing. It is playing defense and trying to make bets that it can find ways to just survive. The noise that Oracle is going to beat AWS is all BS. The proclamations that Oracle's tech is great is too. It is all just marketing with the hope they can convince enough customers to move to Oracle "Cloud" and/or to not abandon their existing licenses and support payments.
Oracle doesn't have a Cloud at all. It has leased data centers with disparate products that require manual effort to connect to each other and that rely on the internet instead of dedicated fiber to communicate. It is really Oracle Hosting. That's not a big bet at all. The bet is that Wall Street and customers won't figure it out.
Excellent post by @UXQUKCZ-1fpw.