I recently had the chance to meet great founders & CEO's building exciting deep tech companies including Hon Weng C. & Alex Wiltschko. HUGE thanks to the "big JB" for bringing us together. I could not be more excited about this field. Osmo is creating smell through software or AI and the other is a biocompute company Cortical Labs. Koniku sits somewhere in the middle, start with "smell processor" or "smell cyborg" then build outwards. We believe a myriad of application layers will be built on our "smell processors".
Cortical labs showed you can teach cultured cells to play pong - read the paper. I saw the hardware and it reminded of the long years and many nights in the lab in Zürich, Umeå, St. Gallen & London..
When I look at our hardware and others you can make out how we got here. You need the eye for it and understand who did what & why certain design choices were made. Here is some background: Jerome "Jerry" Pines (RIP) - a Caltech, particle physicist built the first MEA's (70's & 80's) - multichannel systems (MCS) in Reutlingen ran with it. MCS was acquired by Harvard Apparatus for a $11M 😳!!! I met Jerry once in Reutlingen - amazing man! Peter Fromherz from Max Planck, Martinsried - brought a shit ton of rigor to the whole enterprise - added active elements to ordinary glass electrodes. Prof. Fromherz graciously "reviewed" my M.Sc. thesis. He probably went meh!
I don't believe anyone has done more work to popularize MEA's & expand the reach of MEA's than Steve M. Potter, PhD. Daniel Wagenaar PhD thesis is still the best work I have seen to date in terms of characterizing neural populations on chip - if you can find better, i'll be waiting for it. But you should go read it nevertheless.
In our old lab at the ETH Zürich, Henry Baltes & Andreas Hierlemann led the INPRO project. An EU project - if you can believe it. Within the INPRO project, we built one of the most advanced chips at the time (~2004). Sadik Hafizovic's PhD used Liquid State Machines (and the INPRO Neurochip) to recognize a character sequence using cultured neurons. The INPRO chip forked (by Urs Frey) to create MaxWell Biosystems.
Your take home message is this: "AGI will be done through biological neurons on chip. I am yet to see empirical evidence that says otherwise." The commercialization of all that wealth of knowledge is still at the ground floor. Hey, look at Nvidia...since 1993. So I am not bothered about it. Soon it will be everything everywhere all at once. If you know, you know...if you don't, you chase trends & hype.
Have an amazing weekend & Eid Mubarak once again :-).
P.S. Yes, there is a redacted box. Rules are rules! Ask Hon Weng C. for it, he might like you enough to show you. Plus he is super kind & extremely good looking guy :-). There is also Thomas Oxley, he built the cool stentrode for BMI.
