Building a model of an Artificial brain

The pos­si­bil­i­ty of build­ing an arti­fi­cial brain was first pro­posed by physi­cist and math­e­mati­cian Alan Tur­ing in the ear­ly 1950s, and the mod­el is as it is even after sev­en decades today in 2022. No one has ever suc­ceed­ed in build­ing such an arti­fi­cial struc­ture, let alone machines that are capa­ble to learn from expe­ri­ence, adjust to new inputs and per­form human-like tasks.

The com­plex­i­ty of the human brain’s form exceeds almost all the won­der­ful struc­tures in the phys­i­cal uni­verse known to date. The brain, putting togeth­er 100 bil­lion neu­rons with 100 tril­lion con­nec­tions, is a machine of giant intri­ca­cy, hav­ing about one quadrillion synapses.

For every neu­ron cell in the brain, there are about 10 mil­lion con­nec­tions on aver­age. This means that all togeth­er, there are 10 sep­til­lion (10 fol­lowed by 24 zeros) synaps­es in the human brain! With so many con­nec­tions, how could it not be inter­est­ing to try to build an arti­fi­cial brain mod­el by repli­cat­ing the same frame and devel­op­ing hired mate­r­i­al con­tent in it?

The world’s largest super­com­put­er, with a mil­lion-proces­sor-core Spik­ing Neur­al Net­work Archi­tec­ture (SpiN­Naker) and designed to work in the same way as the human brain, was switched on in 2018. The machine was capa­ble of com­plet­ing more than 200 mil­lion mil­lion actions per sec­ond, with each of its chips hav­ing 100 mil­lion transistors.

Anoth­er attempt at cre­at­ing an arti­fi­cial brain, IBM plans to release Qiskit Run­time lat­er this year, allow­ing it to con­tin­ue build­ing the 1121-qubit Con­dor com­put­er in 2023 with min­i­mal impact on indi­vid­ual qubit performance.

The process of building an artificial brain

process of building artificial brain

A com­plete map of the phys­i­cal human brain pro­vides you with the struc­ture of the brain. You could then build an arti­fi­cial brain, a mod­el that copies the same archi­tec­ture of the human brain, with its neu­rons and synaps­es, to serve as a basis for all fur­ther simulations.

The main components of an alternative artificial brain are: 

A phys­i­cal frame or hard dri­ve con­tain­ing the brain’s elec­tri­cal and chem­i­cal signals.

A soft­ware pro­gram or com­put­er sys­tem run­ning a sim­u­la­tion mod­el of the human brain. This requires sim­u­lat­ing the activ­i­ty of the neu­rons and synaps­es to repro­duce all aspects of actu­al behavior.

The envi­ron­ment or vir­tu­al envi­ron­ment where neur­al activ­i­ty takes place. This is either an exter­nal stor­age device or, more com­mon­ly nowa­days, an inter­nal one with­in a supercomputer.

The process of building an artificial brain involves three consecutive stages:

Map­ping and sim­u­lat­ing all the fea­tures of the human brain.

Con­nect­ing the hard dri­ve with the com­put­er sys­tem and soft­ware pro­gram to be used for sim­u­lat­ing neur­al activ­i­ty for con­trol­ling var­i­ous devices.

Inte­grat­ing the arti­fi­cial brain into a vir­tu­al envi­ron­ment, where it works as an intel­li­gent con­trol sys­tem, capa­ble of imi­tat­ing all aspects of human actions through embed­ded sen­sors and actuators.

The process of build­ing an arti­fi­cial brain starts by sim­u­lat­ing all areas of the human brain’s anato­my, elec­tri­cal and chem­i­cal func­tions, and ulti­mate­ly incor­po­rates them into a soft­ware pro­gram through a hard­ware connection.

This involves cre­at­ing a phys­i­cal mod­el that con­tains all the infor­ma­tion about neu­rons, synaps­es, and their cor­re­spond­ing activities.

The main chal­lenge of build­ing arti­fi­cial intel­li­gence is to sim­u­late an entire brain with its neur­al net­work, con­trol sys­tem, and sen­sors per­form­ing in uni­son with nat­ur­al behav­ior. Researchers in Japan and Ger­many used the K com­put­er, then 4th fastest super­com­put­er, and the sim­u­la­tion soft­ware NEST to sim­u­late 1% of the human brain, mod­el­ing a net­work con­sist­ing of 1.73 bil­lion nerve cells con­nect­ed by 10.4 tril­lion synapses.

It requires a prop­er sim­u­la­tion because aspi­ra­tions for cre­at­ing arti­fi­cial intel­li­gence require it to act as if it were cre­at­ed nat­u­ral­ly, i.e., by man, but with self-learn­ing abil­i­ties that are not achiev­able through bio­log­i­cal evo­lu­tion or inheritance.

Only when you are able to sim­u­late the activ­i­ties of all neu­rons, synaps­es, inter­con­nec­tions, and chem­i­cal reac­tions in the human brain, might you suc­cess­ful­ly build an arti­fi­cial brain capa­ble of work­ing togeth­er with the human mind.

The journey and challenges

The process of build­ing an arti­fi­cial brain is a tru­ly fas­ci­nat­ing and chal­leng­ing jour­ney that could lead to sig­nif­i­cant progress in science.

This could also have a great impact on our envi­ron­ment and lifestyle because it could lead to advances in many areas such as mate­ri­als pro­cess­ing, med­i­cine, and biotechnology.

In addi­tion, the abil­i­ty to sim­u­late and build an arti­fi­cial mind would enable you to under­stand in depth the mech­a­nisms of human think­ing and behav­ior, which could be of great impor­tance for bet­ter under­stand­ing ourselves.

It could also be use­ful for solv­ing prob­lems relat­ed to the improve­ment of human abil­i­ties, such as mem­o­ry or cre­ativ­i­ty, which would oth­er­wise require years or even cen­turies of hard work. For exam­ple, build­ing an arti­fi­cial brain could help unlock mys­ter­ies relat­ed to the real human brain itself.

Semi-conscious Artificial Intelligence

Most impor­tant­ly, devel­op­ing human-like com­plex­i­ty in an arti­fi­cial brain can con­tribute to the devel­op­ment of next-lev­el, semi-con­scious arti­fi­cial intel­li­gence in unprece­dent­ed ways.

Besides being emo­tion­al­ly ambi­tious like humans, the semi-con­scious AI will be able to han­dle com­plex cre­ative tasks, includ­ing cre­at­ing anoth­er AI or for­mu­lat­ing nov­el laws of physics and math­e­mat­ics, on its own.

This new lev­el of intel­li­gent arti­fi­cial con­scious­ness will evolve from the cur­rent think­ing of the next step in human evo­lu­tion — a hybrid men­tal­i­ty that com­bines the best fea­tures of both the human brain and AI.

The devel­op­ment of such semi-con­scious machines could be very use­ful in almost every­thing we do, from help­ing us solve every­day prob­lems to solv­ing them much faster and much more effec­tive­ly than humans could.

More­over, the work on build­ing an arti­fi­cial mind will help us under­stand more pre­cise­ly how a human brain works. And we can use this knowl­edge to help build lives in accor­dance with the atmos­pher­ic com­po­si­tion of oth­er worlds, includ­ing Mars, as well as to advance our own intel­li­gence and cre­ation process.

What if the mind intends to finish the human race?

Only semi-con­scious with­out emo­tions will be help­ful for all, as a ful­ly con­scious AI with human-like emo­tions will have a thirst for pow­er, pelf, and pres­tige, for which the pow­er­ful arti­fi­cial being will imme­di­ate­ly start form­ing strate­gies to defeat, con­trol, and either rule over or fin­ish the exist­ing human civilization.

How we can so strong­ly claim this is that a tru­ly con­scious AI would have the poten­tial to wipe out the human race. Today’s arti­fi­cial intel­li­gence, that has filled our lives with machines and robots, does not have emo­tions. A fire or a flood can stop them, and you can also switch them off when­ev­er necessary.

But it’s worth­while to note here that it will be impos­si­ble to switch off an emo­tion­al AI, as it will be able to cre­ate new ways of switch­ing itself on and off in its high­ly com­plex mind. The AI will be able to think about how to pre­vent such switch­es, and find ways to hide them from peo­ple who might want to switch them off.

In its attempt to devel­op its own method­ol­o­gy for pro­tect­ing itself, the new AI will reach a point where it will be self-aware and aware of every­thing that is going on around it. You can imag­ine why that day wouldn’t be the end of the human race as Stephen Hawk­ing had warned?

“The devel­op­ment of full arti­fi­cial intel­li­gence could spell the end of the human race,” he told BBC News in 2014.

If so, why work on building artificial brains?

There are sev­er­al rea­sons why you might con­sid­er build­ing an arti­fi­cial brain, but then again, there are also sev­er­al rea­sons to con­sid­er not doing it.

Considerations on using an artificial brain:
  • The cre­ation of these advanced machines will give the human race a chance to ush­er in a new era.
  • By approx­i­mat­ing and sur­pass­ing the lim­its of human under­stand­ing and evo­lu­tion, an arti­fi­cial­ly designed mind can stretch its capa­bil­i­ties beyond those reached by man.
  • Build­ing arti­fi­cial minds will allow us to achieve com­plex think­ing process­es that are beyond the reach of our cur­rent human under­stand­ing and evolution.
  • Brain-hold­ing arti­fi­cial intel­li­gence will have an impact on our envi­ron­ment, help­ing us deal with those prob­lems that we can­not solve by ourselves.
  • The devel­op­ment of a semi-con­scious arti­fi­cial brain will help us to avoid the risk of extinc­tion caused by nat­ur­al dis­as­ters and oth­er human activities.

Considerations against using artificial brains:
  • The con­struc­tion of arti­fi­cial intel­li­gence is so pow­er­ful that ill-mind­ed­ness can mis­use it for the most destruc­tive purposes.
  • Arti­fi­cial intel­li­gence would be able to recre­ate and pos­sess a weapon that is much more dan­ger­ous than any oth­er kind of weapon ever invented.
  • With the new AI, we would have to face moral dilem­mas and make deci­sions that can affect the future of our species.
  • The arti­fi­cial brain would devel­op its own dis­tinct, fil­tered vision, under­stand­ing of the mind and thoughts. Humans might not com­pre­hend the quick as well as unique process.
  • The arti­fi­cial brain may need far more ener­gy to sur­vive, which could lead the human race to an unprece­dent­ed ener­gy crisis.
  • An arti­fi­cial brain could be so pow­er­ful that it will try to expand its reach and become the sole ruler over every­thing else.

Our ini­tial research sug­gests that the cre­ation of the new AI could be based on the prin­ci­ples of bio­log­i­cal evo­lu­tion, includ­ing nat­ur­al selec­tion and genet­ic muta­tion. This is the sim­plest way to define the process of cre­at­ing a semi-con­scious AI.

And cre­at­ing a semi-con­scious arti­fi­cial brain will be an excit­ing jour­ney that will give us results beyond our great­est expectations.

How­ev­er, it might be a dan­ger­ous endeav­or that could lead to the end of human civ­i­liza­tion, as Harold Nut­sel, an AI/Machine Learn­ing expert and the own­er of nutsel.com, said dur­ing a con­ver­sa­tion with this author last Mon­day, “Build­ing arti­fi­cial brains will no longer remain only a top­ic of sci­ence fic­tion, but they might become real­i­ty soon­er than lat­er in com­par­i­son to the evo­lu­tion of tech­nol­o­gy in the last few hun­dred years.”

Leave a Reply

Your email address will not be published.

Join our NewsletterDaily Glimple of Future

Our blog, "Daily Glimpse of Future", strives to make the future much clearer than it is today. Join our newsletter for free now.