arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1810.06985cs.AIcs.LOphysics.hist-ph
收藏

人类智能的不可计算性

Non-computability of human intelligence

Yasha Savelyev

摘要:

我们重新审视图灵提出的(最著名的)问题:人类智能可以完全由图灵机建模吗?我们表明,答案是\emph{no},假设一定的弱健全性假设。更具体地说,我们表明,至少有一些有意义的思维过程的大脑不能图灵计算。特别是一些物理过程不是图灵可计算的,这是不完全预期的。我们的论证与著名的卢卡斯-彭罗斯论证有一些相似之处,但我们纯粹在图灵机的水平上工作,而不使用哥德尔不完备性定理或任何直接的类比。相反,我们直接构造并使用哥德尔陈述的弱类似物用于涉及我们人类的特定系统,这允许我们在他们的论点中避开一些(可能的)元逻辑问题。

英文摘要:

We revisit the question (most famously) initiated by Turing: can human intelligence be completely modeled by a Turing machine? We show that the answer is \emph{no}, assuming a certain weak soundness hypothesis. More specifically we show that at least some meaningful thought processes of the brain cannot be Turing computable. In particular some physical processes are not Turing computable, which is not entirely expected. There are some similarities of our argument with the well known Lucas-Penrose argument, but we work purely on the level of Turing machines, and do not use Gödel's incompleteness theorem or any direct analogue. Instead we construct directly and use a weak analogue of a Gödel statement for a certain system which involves our human, this allows us to side-step some (possible) meta-logical issues with their argument.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑