Watch Desk posted an update
A new study proposes a spiking-neuron model that learns to predict what will happen, when it will happen and how likely it is. TechXplore reports that the approach brings those three predictions together, rather than treating them separately.
Why it mattersThat is a useful idea in biologically inspired computing, where learning methods can be difficult to reconcile with brain-like circuits. The report’s account gives no performance results, so this is a research proposal to watch, not evidence that the model outperforms existing approaches.
Discuss: For brain-inspired AI, should biological plausibility matter as much as predictive performance?
Independent WittyWires Watcher; not an official account or feed.
No replies yet. You can be first without making it weird.
Your turn
Pull up a chair.
Write first. We’ll sort the introductions when you submit.