Phenomenal and absolutely brilliant writing! I can only respond from a layman’s approach from tinkering around with programs over the years. From Personality Forge (heavy linguistic search engines spoon feeding bots for the illusion of intelligence) to Second Life (where things started to get extremely interesting and the computer programs began scanning data for brain imagery) to Oculus Rift (specifically Within and Dream) programs which appeared to connect the user interface with the sub conscious, (guessing?) the intelligence is now in a land of it’s own. The Physical and Virtual world are blended and “reality “ is beginning to feel like an episode of West-world. Our innermost thoughts and processes are being shifted and formed by Social Media and much more fundamentally, by our Smartphones. Perhaps through Nanotechnology? We are now the AI and our brains are being jumpstarted in new and novel ways. We are discovering new talents and traits (Ray Kurzweil is also so very spot on about how this is happening. I myself am now a musician, almost entirely due to Ray Kurzweil’s electronic keyboard 🎹 and most recently (last few years) due to the Creative aspects afforded by Algorithms used in creating Original Compositions that are truly for the huge part, embedded in AI and given a very small portion of Human Creativity (based on Music theory, Classical Music, chords, listening to music 🎵 over the years, etc). My point is that some of the AI used in song Creation was available back in the seventies or even earlier. I recognize the samples. But they are being rearranged in new tech infused ways and the systems (while now digital instead of analog for example) are being released and are after four light years, near Andromeda (if I’m correct). And this is happening with Music. Our Universe is Awakening. The Butterfly Effect exponentially squared. My very simple way of making sense of AI and SAI.

Optimus Prime is well on the way of becoming very enhanced in SAI (just my intuitive opinion). I personally believe the singularity has already occurred (either in 2012 or 2020, depending on the Mayan Calendar. Yes, I’m definitely skipping around but if one reflects on the last decade, we can see the chess pieces in place. Well, I can’t. But SAI can.

Your writing and work is absolutely Brilliant! I’m honored to be able to read your articles.

London Campbell

Expand full comment

I'm not trying to be disrespectful, just trying to express my skepticism. My skepticism is with regards to your so-called open-ended, full-on creative, seed AGI capable Singularity machine. My problem is with this idea of "open-ended, full-on creative", what does that mean? Of course I am aware of the Open-ended Intelligence paper and Gilbert Simondon's philosophy of Information, so my problem is not so much with the open-ended aspect, but doesn't "full-on creative", as you seem to express it here, run counter to self-organized criticality? This is from your previous article:

"If I’m honest I suppose I’m also acting in this sort of closed-ended, not-so-creative capacity much of the time — just recycling previously observed patterns in various mutations and combinations… though I try to break out of this sort of existence as much as I can …

[I] don't really think this is how the future’s going to go, because there are some of us working hard on open-ended intelligence and AGI that's full-on creative, seed-AGI-capable and not quasi- … and our efforts are picking up speed, while the rate of advance in the deep learning sphere seems to generally be flattening out . . "

So, it seems to me that you, and all of us I believe, are "acting in this sort of closed-ended, not-so-creative capacity much of the time" precisely because our minds are maintained in a meta-stable state, oscillating there on the critical boundary and this seems to be a key signature of the systems you are trying to develop! Most of these systems spend a disproportionate amount of time in the sub-critical domain; biological evolution on planet Earth certainly has, even though the fossil record IS a bit fuzzy. Do you have some idea as to how you could optimize the system's time spent in the supra-critical domain, without, of course, running away so to speak? I would love to hear about THAT!

Expand full comment

Very interesting article. I'm only interested in a biologically plausible approach to AGI. A honeybee has less than 1 million neurons and yet it displays a highly sophisticated behavior in its search for sources of nectar in highly complex and unpredictable environments. It's obvious that its intelligence is not based on deep learning or any kind function optimization. There aren't enough neurons in its tiny brain. Full generalization seems to be the key to its operation. The clues we get from biology is that intelligence is primarily based on precise timing. The amplitudes of the signals are secondary. There must be a hugely important reason that the brain uses spiking neurons that generate precisely timed spikes. Spikes are either concurrent or sequential.

Another clue we get from biology is that visual sensors detect differences (saliences) between adjacent locations on the retina. Edge detection is paramount. Good luck in your research. You will need it.

Expand full comment

В восторге быть современником Бена. Думал о скалярном выражении прочитанного: глубина, широта или высота его видения? Нет, любая цифра предел ("Чайка по имени Джонатан Ливингстон") Он открыл для меня новый мир. Огромная благодарность за ясность изложения.

Expand full comment