This episode, Amelia and David talk about the parallels between artificial neural networks and the very real neural networks in our heads: brains. Despite many similarities, it’s still an open question whether AI can achieve an internal experience similar to our own. Of course, that’s an unanswerable question, but when has impossibility ever stopped a good discussion?

References and Additional Resources:

• Original Turing Test: https://philpapers.org/rec/TURCMA
• Searle’s “Chinese Room Argument”: https://plato.stanford.edu/entries/chinese-room/
• Brief history of Convolutional Neural Networks: https://glassboxmedicine.com/2019/04/13/a-short-history-of-convolutional-neural-networks/
• Artificial neural networks as models in neuroscience: https://www.cell.com/neuron/pdf/S0896-6273(20)30705-4.pdf
• The Orthogonality Thesis: https://nickbostrom.com/superintelligentwill.pdf
  • This is just a small corner of AI safety, which is a topic we will probably cover during a future episode!

Clarification from Amelia: I stated that AI like ChatGPT can “understand” the rules of grammar. I meant that AI is currently able to use or obey grammatical convention and didn’t intend to imply anything about AI’s inner consciousness or lack thereof.

Please send any questions, feedback, comments, or objections to our email at [email protected]!