<html><head><meta http-equiv="Content-Type" content="text/html; charset=UTF-8"></head><body dir="auto">AIs are to us what jets planes are to birds. Nice comparison. LLMs are already better than us in many areas, just like planes are much faster than birds. They can summarize long texts in an instant, and generate comprehensive essays faster than we can. Different models excel in different areas, almost as different species <div>https://openrouter.ai/rankings<br><div dir="auto"><div dir="auto"><div dir="auto"><br></div><div dir="auto">I am not sure about the hard physical limits. If we take a normal DVD as a measure then a cognitive ability to process a few GB per hour should be enough "to understand" the world. This does not seem to be impossible. </div><div dir="auto"><br></div><div dir="auto">-J.</div><div dir="auto"><br></div></div></div></div><div><br></div><div align="left" dir="auto" style="font-size:100%;color:#000000"><div>-------- Original message --------</div><div>From: Pieter Steenekamp <pieters@randcontrols.co.za> </div><div>Date: 6/20/25 8:32 PM (GMT+01:00) </div><div>To: The Friday Morning Applied Complexity Coffee Group <friam@redfish.com> </div><div>Subject: Re: [FRIAM] AI </div><div><br></div></div><div dir="ltr">Just one thought to toss into the mix: humans didn\u2019t evolve to do astrophysics, drive Ferraris, or detect sarcasm on Twitter. We evolved to dodge predators, gather food, form social bonds, and pass on our genes \u2014 preferably in that order. The human brain is more like a rugged multitool than a precision instrument: built for \u201cgood enough, fast enough\u201d responses in a chaotic and often hostile world.<br><br>Now, if we set out to design a robot to function in today\u2019s environments \u2014 say, hospitals, homes, or corporate boardrooms \u2014 we\u2019re working with a very different set of goals. No need for snake-avoidance instincts or mushroom-edibility heuristics. No need for 30 trillion cells softly glowing in biophotonic harmony. No need for five trillion nerve impulses per second just to decide whether to scratch your nose.<br><br>So even though a robot might never replicate the full sensory richness or biochemical subtlety of the human body, it may not need to. It could get away with a leaner, more focused design \u2014 one that does specific tasks better than humans, precisely because it\u2019s not burdened with all our evolutionary baggage. Think of calculators: they\u2019re completely clueless about context, but they\u2019ll beat any of us in a mental arithmetic race, every time.<br><br>I wouldn\u2019t bet on a human-equivalent robot appearing next year \u2014 but ten years? Maybe. Especially if we stop trying to replicate every biological quirk and instead design for function. And when I say \u201cfunction,\u201d I mean not just doing what a human can do, but doing what the job needs \u2014 which is often a very different thing.<br><br>Take Demis Hassabis\u2019 current project: trying to simulate a single biological cell to improve drug discovery. Sounds simple \u2014 it\u2019s just one cell \u2014 but it\u2019s turning out to be a mammoth challenge. Meanwhile, a useful robot doesn\u2019t need even one biological cell. It just needs actuators, sensors, and some reasonably clever code. This illustrates a broader point: biological systems are complex because evolution took the long road. Engineering can often take a shortcut.<br><br>So yes, the human body is a marvel \u2014 a product of billions of years of trial and error. But that doesn\u2019t mean it\u2019s the most efficient solution for every task. It\u2019s just the one that happened to work well enough to keep our ancestors from being eaten.<br><br>After all, birds fly beautifully. But when we wanted to fly, we didn\u2019t grow feathers. We built jets.</div><br><div class="gmail_quote gmail_quote_container"><div class="gmail_attr" dir="ltr">On Fri, 20 Jun 2025 at 19:15, Prof David West <<a href="mailto:profwest@fastmail.fm">profwest@fastmail.fm</a>> wrote:<br></div><blockquote style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex" class="gmail_quote">Marcus made a comment recently about constructing an AI plus robotic body that provided the AI with sensory inputs comparable to a human being. It made me wonder about feasibility of such an idea.<br>
<br>
The average human body has about 100 billion nerve endings generating electrical impulses<br>
<br>
The average human (sex, weight, height sensitive) has about 30 trillion cells emitting ultra-weak biophotons; increasingly shown to play a role in inter-cellular communication<br>
<br>
It is extremely difficult to compare something like FLOPS for the brain, but best estimates suggest an average of 43 teraFLOPS, and up to 430 teraFLOPS for peak situations. Computers are capable of 1.1 exaFLOPS. But the brain uses 20 watts of power and the computer megawatts.<br>
<br>
Taking into account synaptic delay and refactory delay, each nerve ending could send a signal to the brain, or the brain could \u2018process\u2019 those signals at a rate between 10 Hz (cortex) to 1,000 Hz elsewhere. Also assume that the biophotons work mostly locally and maybe 1 percent actually end up triggering something akin to a nerve signal so, until we know more, it is unlikely that more than 30,000 to 300,000 additional signals reach the brain \u2013 less than noise, given what we know now. But that might change significantly in the future, especially as we learn more about quantum effects in the brain in general.<br>
<br>
The brain could receive 5 trillion discrete signals per second, but \u201cpre-processing\u201d reduces that to between 50 (average) and 500 million (peak) signals per second.<br>
<br>
.02-.03 percent of those signals are symbolic- originating in a phoneme, lexeme, word, number.<br>
<br>
Between .22 and 12.3 of the \u201cnon-symbolic\u201d signals process by the brain have a mediating effect on symbolic processing, in the human brain. Some of this can be simulated by an AI. Take sarcasm as an example: humans use a lot of non-symbolic signals to detect sarcasm with a success rate of about 95%. AI\u2019s must rely on context, on explicit labeling of training material, and, if available sound or images that can be analyzed. With a success rate of about 80%.<br>
<br>
Currently, an AI can simulate/emulate/equate to the roughly .02-.03 percent of the signal processing done by the human brain, i.e., that directly related to symbolic inputs. It can also deal with, roughly 80% (based on the sarcasm example) of the mediating non-symbolic signals (between .22 and 12.3 percent of signals processed by the brain.<br>
<br>
These numbers suggest, to me, that an AI is capable of simulating/emulating/equating-to about 1 to 15% of human brain signal processing. Of course, the human brain has all kinds of help elsewhere in the body, synthesizing, attenuating (reducing), and \u201cpre-processing\u201d signals. An AI has none of that help.<br>
<br>
So, it seems to me, that an AI must necessarily be a true idiot-savant for language manipulation and pattern recognition (image, sound).<br>
<br>
Only if we define human intelligence as nothing more than human abilities with language and visual/auditory pattern recognition can we say that artificial intelligence meets or exceeds (only in terms of speed) human intelligence.<br>
<br>
I used AI to generate all the numbers in the above.<br>
<br>
davew<br>
<br>
.- .-.. .-.. / ..-. --- --- - . .-. ... / .- .-. . / .-- .-. --- -. --. / ... --- -- . / .- .-. . / ..- ... . ..-. ..- .-..<br>
FRIAM Applied Complexity Group listserv<br>
Fridays 9a-12p Friday St. Johns Cafe / Thursdays 9a-12p Zoom <a rel="noreferrer" href="https://bit.ly/virtualfriam">https://bit.ly/virtualfriam</a><br>
to (un)subscribe <a rel="noreferrer" href="http://redfish.com/mailman/listinfo/friam_redfish.com">http://redfish.com/mailman/listinfo/friam_redfish.com</a><br>
FRIAM-COMIC <a rel="noreferrer" href="http://friam-comic.blogspot.com/">http://friam-comic.blogspot.com/</a><br>
archives: 5/2017 thru present <a rel="noreferrer" href="https://redfish.com/pipermail/friam_redfish.com/">https://redfish.com/pipermail/friam_redfish.com/</a><br>
1/2003 thru 6/2021 <a rel="noreferrer" href="http://friam.383.s1.nabble.com/">http://friam.383.s1.nabble.com/</a><br>
</blockquote></div>
</body></html>