Do you think the AI hype cycle will carry over into 2025? For a little while longer, yes. There will still be executives with FOMO hoping to show they can opt

Do you think the AI hype cycle will carry over into 2025? For a little while longer, yes. There will still be executives with FOMO hoping to show they can optimize head count using LLM's. In my opinion the companies able to reduce headcount using LLM's did not really need LLM's but needed to work on big-data management but lacked the talent or priorities or budget to do so. "AI" gives them a way to outsource this to the cloud so it's more of a OpEx vs CapEx issue at the moment. Once those companies are sorted out I believe it will start to plateau. So perhaps more like 2026/2027 to plateau and probably shortly thereafter to start losing the hype assuming people are not tripping over SLA issues due to third party dependencies, rate limits, privacy incidents and network outages at which point it becomes a bigger cost issue for standing up and migrating to self hosting LLM's or the big-data internal development and integrations they originally needed to do.