HackerNews中文版

我发现自己越来越多地将细节外包给人工智能。我开始故意忘记这些细节。我希望人工智能能够掌握它们。为什么？因为计算能力在这里。因此，知识也应该在这里存在。每次人工智能遗漏它不知道的内容时，我重新告诉它，这样做效率低下。我输入这些内容需要X时间，语音输入可能需要log(X)的时间。但接下来不可避免的往返沟通、轻微的误解、纠正等等，又会浪费更多时间。我意识到自己自然地倾向于让人工智能掌握所有的数据和知识。因为这本来就是它的职责。它需要进行计算，那么我为什么还要了解这些呢？人们认为将思考外包给人工智能是一条滑向深渊的道路。我并不认为这是滑坡，而是一条滑水道。这是不可避免的。就像重力一样，迅速占据主导地位。因为没有任何力量或激励能将水推回山上。人工智能应该掌握所有的数据。尽管这听起来很奇怪，但这就是应该如此。我现在找不到其他的解决办法。你有什么想法吗？

查看原文

I found myself increasingly outsourcing the details to the AI. I forgot the details, deliberately I think. I wanted the AI to know them. Why? Because that's where the compute is. So that's where the knowledge has to live.Me re-telling it to the AI every time it misses something it didn't know, is inefficient. It takes me X time to type it, and maybe log(X) to voice type it. But then there's the inevitable back and forth, the slight misunderstandings, the corrections, etc.I realized and found myself naturally sliding down towards, just letting the AI own all the data and knowledge. Becuase it should. That's the one that has to compute with it, so why should I know about it.People think AI is a slippery slope by outsourcing our thinking. I don't think it's a slippery slope - it's a waterslide. It's just inevitable. It's gravity, taking over rapidly. Because there's no force nor incentive pushing the water back up hill.AI should own all the data. As weird as that feels, that's how it should be. I don't know another way right now.Thoughts?

告诉HN：人工智能不是一个滑坡，而是一个水滑梯。