What's the read on software, cloud, and internet moats, and where are experts most and least constructive?

The advent of AI represents a fundamental platform shift for software, comparable to the internet or cloud, that is reshaping the basis of competitive advantage . Experts agree that AI expands the total addressable market for software from the confines of IT spend to the much larger category of labor spend, as software can now directly perform work previously done by humans [1, 13, 23, 25]. However, there is a strong consensus that AI itself is not a durable moat; rather, it is a powerful but replicable tool for product differentiation [1, 4]. This ease of replication lowers the barrier to entry, commoditizing software creation and fostering a hyper-competitive landscape where achieving the scale necessary for a durable moat is more difficult for new entrants [1, 14]. For investors, this means shifting evaluation from IT budgets to the potential for capturing a portion of **global labor spend** [19, 25].

There is significant tension among experts regarding the durability of traditional software moats. One camp argues that classic moats like network effects