Two major events have happened recently: the release of Deepseek V4 and the release of GPT 5.5. I'm also waiting for the study committee's notes.
But honestly, there's not much to say. By now, intelligence is no longer an issue for LLMs, and various benchmarks are just gimmicks. The real bottleneck lies in agentic capabilities. Based on current information, both new models have put significant effort into this area. GPT 5.5, undoubtedly paired with Codex, is aiming to push for a general-purpose agent. This wave of capability release is late but finally here.
Instead, Deepseek V4 deserves more attention because it challenges Silicon Valley, specifically the discourse power in the AI field from hardware to ecosystem.
Sounds grand, doesn't it? But even if you don't pay attention for a year or two, it doesn't matter. The tech industry never lacks new buzzwords and technologies. Good products won't impose extra learning costs on users; truly great products never make demands on users. So even if AI truly becomes mainstream, you won't feel too abrupt a shift.
Besides, the only area where AI has truly been implemented so far is information retrieval.
Wait for the study committee's notes, wait for real user experiences, and try it out yourself. If it helps you, that's great; if not, it means the product isn't ready yet or the technology hasn't matured. For things like LLM AI, data is of little reference value; your own experience is what matters.
Most technologies don't solve real-world problems; they solve productivity issues. The quality of life depends partly on mindset and partly on other topics that are inconvenient to discuss.
Take it easy. Everything will be fine.