In a recent (2020-10-29) episode of “Closer to Truth” at time mark 06:41 Raymond Kurzweil said this:
“we can actually get a measure of how much complexity is in the brain, because the design of the brain is in the genome…”“the design of the brain is in the genome”? huh?

design?

BTW, the design of the brain is in the genome? Really?The genome contains information that is necessary, but insufficient, for building the brain.The entire biological system is required.

Transcript

00:00ray i did my doctorate in hardcore

00:03neuroscience neurophysiology and00:06in trying to understand what the brain00:09is00:10most of the people i talked to or my old00:12professors and00:14neuroanatomy and neurochemistry what i’d00:16really like you to do00:17is is reflect on what a brain is how it00:21functions00:22uh coming at it from a different00:24perspective as a computer scientist and00:26as a00:27thinker who’s who’s thought about brains00:29in a different way00:30well it’s all about like asking00:34the different blind observers what an00:36elephant is and00:37depending on what part you’re examining00:39they may describe it differently00:41and we’re still at a stage where all of00:44these different00:44perspectives modeling one neuron00:48modeling portions of a neuron like the00:51tubules00:52or the filaments uh or00:55modeling entire regions are are valid00:59are giving us greater insight we are01:02getting more and more information about01:03the brain the spatial resolution of01:06brain scanning is doubling every year01:07we’re not01:08getting to the point where we can see in01:09a living brain individual into neuronal01:12connections and see them firing in real01:14time and we can see them01:15creating new spines and new synapses and01:18we can actually see01:19our brain create our thoughts for the01:21first time we also see our thoughts01:23create our brain because01:24that is the feedback loop that gives us01:27plasticity01:28and but then the question is okay we’re01:31getting this data and then the amount of01:32data by the way is doubling01:33every year can we make heads and tails01:35of this there’s a vast amount of01:37information01:38we are finding that we can actually01:41understand01:42how specific regions work and as we get01:44enough data about specific areas of the01:46brain01:47we are succeeding in making models of01:50how that region processes information01:52and fundamentally that is what the brain01:54does yes it’s01:56partly analog and partly digital it’s01:58massively parallel whether our computers02:01are02:01more sequential but fundamentally it’s02:04an information processor02:06and some processes or some02:09efforts to understand the human brain02:11simulate02:12at a very precise biochemical level so02:15for example ibm has a project02:17to simulate a significant slice of the02:19cerebral cortex where we do our abstract02:21reasoning first an02:23individual cell level and then actually02:25at the level of biochemical molecules02:27and that’s one valid approach we also02:30find02:31sort of the neurological function02:35level that we can take a whole region02:38and ask the question what does this02:39region do to information how does it02:41represent it how does it transform it to02:43another region02:44without really modeling every single02:47synapse in every single cell02:49and this has been done for a couple02:51dozen regions of the human brain a dozen02:52regions of the auditory cortex02:54have been modeled and simulated and then02:57we can apply sophisticated tests02:59to the simulation and they get very03:00similar results applying those same03:02tests to human auditory perception03:04there’s now a fairly recent model in03:06simulation of03:08several regions of the visual cortex the03:11cerebellum which is where we do our03:14scale formation has been simulated we’re03:16actually beginning to understand03:18how a nine-year-old child can actually03:20catch a fly ball03:21because that you know they’re not doing03:24you know 5003:25simultaneous differential equations in03:27real time they’re actually able to03:28directly translate the movements that03:30they see03:31in their eye to the movement of their03:33hand and it takes time to develop that03:35skill03:36something called basis functions which03:37collapses all these differential03:38functions and03:40that’s how the cerebellum works we’re03:42finding that these different regions03:43have03:44the same structure repeated over and03:46over again the cerebellum is half the03:47neurons in the brain03:49it’s tens of billions of neurons but it03:51has one very simple structure that’s03:53repeated03:53billions of times it’s able to learn03:56a certain scale formation we actually03:58find when people learn04:00cursive writing there are groups of04:02these structures within cerebellum that04:04learn each type of stroke04:06and each type of movement for catching a04:08fly ball or walking or talking04:10use the same structure over and over04:12again the cerebral cortex where we do04:14abstract reasoning04:15has the same structure repeated over and04:17over again that appears04:18capable of doing a certain type of04:20recursive function where we can take04:22some complicated04:23structure of symbols and represent it as04:26one thing04:26and then use that in another hierarchy04:28so we can do this hierarchical04:30recursive thinking that’s part that’s04:32reflected in our language which is very04:34unique and that’s something that human04:35beings can do04:37you know more than other species04:38apparently that’s reflected04:40in the cerebral cortex we’re still in an04:43early stage04:44of understanding the human brain but04:46we’re showing that we can gather the04:47data04:48we can turn that data into working04:50simulations we can test the simulations04:52and find that they match04:53human performance up to certain levels04:56and all of this is progressing04:58exponentially i’ve made the case that we04:59will have simulations of all several05:02hundred regions within 20 years05:04some cognitive scientists would say that05:06all this biology is interesting but05:08that’s just05:10a way of representing what’s really the05:13fundamental thing which are these05:14algorithms or ways of thinking from a05:16cognitive point of view it doesn’t05:17matter if it’s in a brain or in a05:19computer05:19it’s all the same and that’s really what05:22it’s all about05:22well that’s correct i mean when i05:24studied computer science we learned the05:26intricacies of transistors and how05:27they’re constructed and the materials05:29and all the differential equations that05:31describe a transistor and then you could05:32learn what it does and then you put05:34multiple transistors together and you05:36can create something that multiplies05:37numbers05:38but then you can just forget about all05:40that and we don’t05:42try to understand a computer in terms of05:44exactly how the transistors are built we05:46understand them at the abstract level05:48of what it’s really doing to information05:51and through evolution05:52these different brain regions are05:54processing information05:56and it is important to understand these05:58different levels05:59i mean it’s important to understand06:00physics to understand chemistry but06:03finally when we learn chemical06:04principles we can throw the physics away06:06in biology is based on chemistry06:08you have to understand chemistry to some06:10extent to understand what they’re06:11talking about06:12but then you can understand biological06:14principles and ultimately you do get to06:16what these brain regions are doing to06:18information there may be a whole complex06:20region of the there are is a complex06:22region of the brain06:23that measures the time difference06:24between certain sound signals coming06:26from the two ears06:27it’s a fairly simple algorithmic idea06:30there’s a very complex structure06:33that actually computes that and06:36the question comes up well this must be06:38a vast complexity we’ll never be able to06:40understand this06:41we can actually get a measure of how06:43much complexity is in the brain06:45because the design of the brain is in06:46the genome06:48the body also and this there’s a lot of06:51information there but not that much06:53even if you even if you assume as i have06:55always assumed that the so-called junk06:57dna is not06:58junk because it controls gene expression07:00the genome is replete with redundancy07:02the sequence is repeated over and over07:04again one sequence called alu is07:05repeated07:06300 000 times and for those of you know07:09about data compression you know if07:10there’s a repetition of information you07:11can compress it07:13without losing information if you apply07:14lossless compression07:16to the genome you get somewhere between07:1830 and 100 million bytes07:20which is not small but it’s a level of07:22complexity that we can actually manage07:25and we are actually making very07:26impressive progress but it’s important07:28to know the progress is exponential07:30because halfway through the genome07:31project the skeptics are saying hey07:33you’re seven and a half years into this07:3415-year project07:35you’ve done one percent of the project07:37this is going to take hundreds of years07:39but in fact it doubled every year and we07:40finished the project on time it’s going07:42to be the same thing07:43with the brain the precision the scale07:45the amount of information07:47the the the number of regions and the07:49precision with which we’re simulating07:51all of this is07:52is it’s gearing up exponentially and of07:54course the super computers that we can07:56run these simulations on07:57are doubling in power every year and so08:00you know 20 years from now08:01this technologies will be a million08:03times more powerful than they are today08:05and we will finish the job i believe08:06within 20 years08:08and that will give us more insight into08:09ourselves we’ll understand the brain08:11better and understand how to fix certain08:13problems better than we do today08:15and we’ll but also i think most08:17importantly give us08:18an enhancement of the artificial08:20intelligence toolkit we have08:23and i had an argument with professor08:25poggio at mit08:26saying he’s not learning much about08:29how to do machine vision from the human08:32visual system and i said well it’s08:33because you don’t have good models yet08:36and recently i saw him and he said you08:37know you were right we now have these08:39very good simulations of how human08:41visual processing takes place in human08:43brain we applied that to machine vision08:45and got a big quantum jump in08:47performance so we are going to learn how08:48to create intelligent machines08:50by learning how evolution solved this08:52problem08:53over billions of years why do then you08:56use the word08:57spiritual to talk about these machines08:59that we’ll we will09:01create well they’re going to09:05have similar capabilities that we09:08associate with09:10other systems or entities that we call09:13spiritual which are human beings09:15because we’re going to really reverse09:17engineer the human brain and understand09:18how it works including09:20those regions that deal with emotion09:22which is not some sideshow that’s09:24actually the most complicated thing we09:25do09:26and spiritual behavior and09:29is uh is part of what the human brain09:32does and we will have entities09:34that exhibit for example conscious09:36behavior09:38and in my view you can’t fundamentally09:41understand09:41consciousness in objective terms but09:44these entities will appear09:46just as conscious they’ll claim to be09:48conscious they’ll be convincing they’ll09:49be as convincing as human beings are09:52and so if we believe that human beings09:55are are spiritual09:56then these machines will be spiritual as09:58well they’ll be as spiritual because10:00they will be as10:01capable of these very subtle emotions as10:04human beings