<div dir="ltr"><div class="gmail_quote"><div dir="auto"><br><div class="gmail_quote" dir="auto"><div dir="ltr" class="gmail_attr">On Sat, Aug 13, 2022, 12:55 PM Stuart LaForge via extropy-chat <<a href="mailto:extropy-chat@lists.extropy.org" rel="noreferrer noreferrer" target="_blank">extropy-chat@lists.extropy.org</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><br>
Implementing neural networks as hardware on chips could do for <br>
training AI what ASICS did for bitcoin mining. Neuromorphic chips have <br>
a lot of potential IMO. Why simulate what you can instead reverse <br>
engineer?<br>
<br>
<a href="https://www.nature.com/articles/s43588-021-00184-y" rel="noreferrer noreferrer noreferrer noreferrer" target="_blank">https://www.nature.com/articles/s43588-021-00184-y</a><br>
<br>
----------------------------k<br>
Abstract<br>
<br>
Neuromorphic computing technologies will be important for the future <br>
of computing, but much of the work in neuromorphic computing has <br>
focused on hardware development. Here, we review recent results in <br>
neuromorphic computing algorithms and applications. We highlight <br>
characteristics of neuromorphic computing technologies that make them <br>
attractive for the future of computing and we discuss opportunities <br>
for future development of algorithms and applications on these systems.<br>
----------------------------<br>
<br>
In his recent Frontiers in Neuroscience article about phenomenal <br>
consciousness being mediated by the complex EM fields of the brain, <br>
Colin Hales wrote:<br>
<br>
"The creation of chip materials able to express EM fields structurally <br>
identical to those produced by neurons can be used to construct <br>
artificial neurons that replicate neuron signal processing through <br>
allowing the actual, natural EM fields to naturally interact in the <br>
manner they do in the brain, thereby replicating the same kind of <br>
signaling and signal processing (computation). This kind of in silico <br>
empirical approach is simply missing from the science." (Hales & <br>
Ericson, 2022)<br>
<br>
So Colin, it appears that the neuromorphic chips and computer <br>
architecture described in the Nature Computational Science article is <br>
exactly what you were suggesting right? So if these novel neuromorphic <br>
AI work as expected, would you believe one of these new machines to <br>
posses phenomenal consciousness or 1PP?<br>
<br>
Stuart LaForge<br><br></blockquote><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><br></blockquote><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><br></blockquote><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">Neuromorphic computers are not reverse engineering the brain. They are a highly parallel, fast, low energy implementation of a model of properties of brain signalling. None of the brain's signalling physics is involved.</blockquote></div><div dir="auto"><br></div><div dir="auto">A chip that reverse-engineers the brain's function is one that literally uses the physics the brain uses, in inorganic form. Nowhere else in science, EVER, has 'reverse engineering of a natural phenomenon been the creation of a computer-explored abstract model of the nature. I have called such a reverse engineered chip a 'neuromimetic' chip. There's a detailed discussion of it in TechRXiv. <a href="https://doi.org/10.36227/techrxiv.13298750.v4" target="_blank">https://doi.org/10.36227/techrxiv.13298750.v4</a> </div><div dir="auto"><br></div><div dir="auto">The likelihood is that you won't be able to fully appreciate the difference, but it is real and untried.</div><div dir="auto"><br></div><div dir="auto">I am building a 50,000+ scale version of one patch of membrane with one big ugly 'ion channel' in the middle of it. It produces a near-field that innately expresses the voltages modelled by the neuromohic chip. No model. No software. Just physics. If you place these fields inside each other they compute cognition. Literally. That is replication (reverse engineering) of brain signalling. It will be the first ever attempt to do so.</div><div dir="auto"><br></div><div dir="auto">Maybe then it'll be clear and the implications can finally be properly examined. </div><div dir="auto"><br></div><div dir="auto">I am calibrating the sensor positioner at the moment. Attached is a picture I hope will get through the byte limit of posting here:</div><div dir="auto"><br></div><div dir="auto">This endless argument will end before the end of the year. </div><div dir="auto">65 years of theoretical science being mistaken for empirical work has to stop.<br></div><div dir="auto"><br></div><div dir="auto">Cheers,</div><div dir="auto"><br></div><div>Colin</div></div>
</div></div>