AI: Still Early, But Not Too Early

研报英文原文证据摘录

AI: Still Early, But Not Too Early

ar

Case on Software is … a Hallucination” also dated today, along with our upgrades of

NOW, CRM, and CHKP. In addition, we introduce new AI subsectors and representative

companies in each – stay tuned for more here.

AI Macro Effects: We believe that (1) AI represents a technology paradigm shift that (2)

essentially all Software vendors will have to embrace, though history tells us that (3) the

leaders of a new paradigm are not often the leaders of the past one. But (4) previous

paradigm leaders often persist, even if (5) their climb in relevance is dampened. (6)

To bridge into the next paradigm is difficult and costly and typically requires scale and

alternative cash flow to fuel the shift. In addition, (7) previous paradigm job roles suffer,

but are often replaced by new roles, increasing the aggregate number of jobs.

AI Micro Effects: (1) Leading SaaS companies’ unkept promises of incremental

monetization of AI have probably done more harm to traditional Software stocks

than AI itself. (2)AI represents more risk to traditional Software that will benefit

from probabilistic solutions (e.g., CRM, some HCM, Development Software, ITXM, Work

Collaboration) than to traditional Software that requires deterministic outcomes (e.g.,

Payroll, ERP). AI will affect the financial framework of traditional Software subsectors

differently, with (3) IaaS exhibiting inherently higher COGS and inherently lower Opex,

which should result in greater absolute profit even if at lower margins. (4) It’s less

clear with most traditional Software which will see higher costs (that should come

down over time) and will not likely be able to charge incrementally for AI. (5) Pricing

models are still evolving, though token-based pricing will be a component. (6) There