OK, so we have scratch pads for the brain. Now we need a scratch pad for AI.

现在我们有了大脑的草稿纸，我们需要拿到 AI 的草稿纸。

But even ask it to show you its scratch pad work.

而是能够让它向你展示它的工作流程。

Again, this scratch pad prediction task works.

这个草稿预测任务得到了同样的结果。

The scratch pad tells us a little bit about how the brain represents the language.

这些草稿本向我们透露了一些大脑表示语言的方法。

Still, these scratch pad prediction experiments have held up — multiple brain imaging experiments, multiple neural networks.

尽管如此，这些草稿预测实验仍然站得住脚——多个大脑成像结果，多个神经网络模型。

We're able to predict the neural network scratch pad from the brain and vice versa.

我们可以从大脑图像预测神经网络，反过来也可以。

Now we have two scratch pads, one from the brain and one from AI.

现在我们有了两张草稿纸，一张来自大脑，一张来自人工智能。

Though these scratch pad prediction tasks show above-chance accuracy, the underlying correlations are still pretty weak.

尽管这些草稿预测任务表现出高于随机的准确率，两者底层的相关性仍然非常弱。

Now we need a scratch pad for AI.

现在我们需要 AI 的便笺簿。

OK, so we have scratch pads for the brain.

好的，所以我们有大脑的便笺簿。