2 ms·
You can't really debug an AI. The dimensions of it's understanding are quite literally beyond human interpretation, which makes it both smarter/more efficient t
by smoldesu 4y ago
You can't really debug an AI. The dimensions of it's understanding are quite literally beyond human interpretation, which makes it both smarter/more efficient than humans while also extremely dumb and context-unaware. Most of our attempts at adding a 'memory' to AI has been a hack thus far, which is why all of these prompts consist of people force-feeding word salad down the AI's throat for generally reproducible results.
- the8472 4y agoThey're debuggable, with effort. Here's the finance neuron in CLIP: https://microscope.openai.com/models/contrastive_4x/image_block_4_5_Add_6_0/1330 https://microscope.openai.com/models/contrastive_4x/image_bl...
- sdenton4 4y agoFor memory, check out: 1) Retro, which is essentially attention over large databases, and fast as hell. 2) S4 Layers, explicitly designed for handling long dependencies. These are orthogonal approaches to memory, and both very effective at what they do.