Hardware Podcasts

Rippling Signals May Provide Working Memory in the Brain

EE Times Current · 4 Apr 2025 · 51m

What this episode covers

Dr. Terry Sejnowski of the Salk Institute discusses his 50-year career bridging neuroscience and AI, including his foundational work on the Boltzmann machine with Jeff Hinton. He proposes that traveling waves and Spike Timing Dependent Plasticity (STDP) provide the biological substrate for long-term working memory, a mechanism that could inspire more efficient neuromorphic hardware. The hosts, Dr. Sunny Bains and Dr. Julia D'Angelo, along with commentator Prof. Ralph Etienne-Cummings, debate the implications of these findings for AI, specifically regarding the 'shallow' nature of Large Language Models and the role of embodiment.

Most relevant to

established companies

Topics

Mentioned

Companies
Salk Institute, University College London, Johns Hopkins University, IBM
Products
Boltzmann machine, ChatGPT
Also
NURIPS, Telluride Neuromorphic Cognition Engineering Workshop

Related episodes from other shows

Other shows on the same ground, closest first.

  • The Next Byte Podcast · 21m

    201. Non-invasive Memory Booster

    Hosts Daniel and the other speaker discuss a non-invasive memory booster developed by researchers at EPFL, specifically the Hummelab and the Laboratory of Cognitive Science. The device uses transcranial temporal interference stimulation (TTIS) via scalp electrodes to stimulate the hippocampus, improving spatial memory recall by 18% in VR navigation tests. The episode contrasts this approach with invasive options like Neuralink and highlights the potential for treating dementia and cognitive decline.

    Medical & health devicesAI & machine learning
    5 chapters

The summary, chapters and topics are generated from the episode’s audio by this directory, not written by EE Times Current. The episode itself belongs to its publisher.