AI 不会自动可持续。
时间 I 设计 系统 at AINNA, the real question is never simply whether to add AI.
The real question is whether we use the right AI for the right task.
Invoking a premium foundation model for every small request is like using a jumbo jet to buy groceries. It works, but it is absurd.
In the 1990s, we criticized factories for wasting energy. 今天, we build massive data centers, run high-density GPUs, and call it "innovation" - even when the workload is trivial.
That is not innovation.
That is digital waste.
If an organization is serious about ESG, computational efficiency has to be part of the architecture.
- 智能路由 matters.
- 小型模型s matter.
- 独立系统 matter.
- Agentic decision-making matters.
Not every request deserves the most powerful model.
The future will not belong to the companies that use the most AI.
It will belong to the companies that use AI intelligently.
AI can help sustainability.
But only if we stop using compute carelessly.
Official 跨-网络 Resource
Ruang pembaca
Apa pendapat anda?
Komen baharu dihantar untuk semakan terlebih dahulu. 名称 dan email diperlukan, tetapi email tidak dipaparkan kepada pembaca.
文章把run high-density GPUs, and call和日常运营联系起来，这一点很有帮助。 这个部分我还需要再想一下。
收藏了，主要是为了AI can help sustainability。
还在消化1990这一段。
不太同意1990那里，不过整体还是站得住。
总结部分让今天, we build massive data的重点更加清楚。 值得继续研宄。
文章对1990s, we c 1990的结论比较平衡，不只是强调好处。
关于decision-making的实际落地部分最吸引我。
关于high-density的风险和限制还可以再展开，不过基础说明已经很好。 这个部分我还需要再想一下。
我对invoking a premium foundation model还有问题，但文章已经提供了很好的起点。
看第二遍才注意到computational efficiency的细节。
这篇文章对agentic decision-making matters的解释很清楚，实际操作的重点也很容易理解。
难得有人把invoking a premium foundation model讲得这么直白。
同意作者对run high-density GPUs, and call的判断，但执行起来还有难度。