Evolution of Machine Intelligence: An Infinite run?

AI is becom­ing more and more ingrained into our lives. It affects every­thing from the way we search for infor­ma­tion, to the way we com­mu­ni­cate with those around us. 

Leave aside the Sci-Fis, there is no need to wor­ry that this “smart” machine is going to take over the world. There are plen­ty of oppor­tu­ni­ties for us if we want to improve upon it. Our own human intel­li­gence can be inte­grat­ed into arti­fi­cial intel­li­gence to increase the machine’s intel­li­gence, while also cre­at­ing a bet­ter inter­face with the AI, a more ful­fill­ing life. 

The concern of machine intelligence and infinity

Infinity loop machine intelligence

In just a few years time, machines have become an inte­gral part of our lives; in a way that was impos­si­ble only a decade ago. The way we com­mu­ni­cate, the way we enter­tain our­selves, and the way we make our liv­ing are all affect­ed by a machine’s intel­li­gence. And for a good rea­son – it was invent­ed for our use.

While machines are get­ting smarter, there is a grow­ing con­cern with­in the AI com­mu­ni­ty. The con­cern is if machines start help­ing machines with intel­li­gence, there can be an infi­nite run of self-improve­ment. Some believe that com­put­ers can start improv­ing com­put­ers and the machine intel­li­gence will only get bet­ter and bet­ter. As a result, we will run into a wall where there is no way to fur­ther progress. I call this the “Infi­nite run” of machine intelligence.

Related Post:

Humans are always adding to their knowl­edge and expe­ri­ence. Do we ever stop get­ting smarter? No, and nei­ther will machines.

In the present, which I am defin­ing as the next few decades, there is no need to wor­ry about this loop of machine intel­li­gence. But with the expo­nen­tial growth of machine intel­li­gence – and our con­nec­tion to it – this has become a point of con­cern for many AI experts.

As the indus­try pro­gress­es and machines take on more respon­si­bil­i­ty, there will be a point where we can­not look at the data we are col­lect­ing. This is when smart machines can real­ly become “smart”.

With all of this data at its dis­pos­al, an AI could devise an equa­tion to make oth­er machines in its net­work more intel­li­gent. And if those machines were made intel­li­gent enough to design a bet­ter machine, one could jump back into an infi­nite run of intel­li­gence – mak­ing com­put­ers smarter than humans.

If a machine gets human capa­bil­i­ties too, it will most like­ly lead to “infi­nite intel­li­gence”; it’s just because of that com­pound effect we dis­cussed ear­li­er in the article.

This essen­tial­ly means that there’s only so much we can do to improve upon arti­fi­cial intel­li­gence – espe­cial­ly if we are rely­ing on our log­i­cal minds to fig­ure out how to build it in the first place. 

As the future of com­put­ers & our lives turns to a future of inex­orably smarter machines, there are many ways we can improve upon them, and the major­i­ty of these improve­ments cen­ter around improv­ing the intel­li­gence behind the machine.

Self improvement of machines: an infinite run?

The day a machine becomes capa­ble enough to cre­ate a machine smarter than it (even 0.000001% smarter), we will start head­ing towards the infin­i­ty-run of machine intel­li­gence. This “algo­rithm” could then be used to make anoth­er “smarter” machine; it, in turn, becomes capa­ble of cre­at­ing a new machine even smarter than itself.

Is it possible?

It is still about at least 50 years before we will have the first machine that can do this. But with enough com­pu­ta­tion­al resources and the right algo­rithms, it is achievable.

But I believe that it is impor­tant for AI experts and researchers to think about how we can make our machines more intel­li­gent – the way we are able to make our com­put­ers more intel­li­gent. So let’s look at ways of doing so.

The sim­plest way is to give your AI a human brain, and make it like us. But this one is just one of those sim­ple ideas which are far more com­pli­cat­ed to execute.

As the con­cept of machines “becom­ing smarter than us”, seems to be a gen­er­al con­cern, many AI experts are find­ing ways to make sure that it does not hap­pen. There is a new field of research known as “human-inspired AI”. The research focus­es on mak­ing machines more human in the sense that they have human intel­li­gence; but still have some kind of machine log­ic behind it.

To do this, many researchers are try­ing to cre­ate machines with neur­al net­works — net­works that mim­ic the way neu­rons com­mu­ni­cate in our brains. The goal is to see if we can cre­ate a “brain” on a chip that mim­ics the way our neu­rons com­mu­ni­cate, adding machine intel­li­gence to machines.

Machine-learn­ing will let machines solve prob­lems they nev­er con­ceived of before. And by doing so, they are able to evolve faster than humans can. How­ev­er, this might become a prob­lem when those machines evolve into some­thing we can­not con­trol; it may be some­thing uncon­trol­lable by us or any oth­er type of human intelligence.

What would immense capabilities mean for AI?

Ok! We all know this, but it is still true: you can see com­put­ers can solve prob­lems in sec­onds. And it can take “our brain” hours to do the same. The humans’ brain and its capa­bil­i­ties are actu­al­ly lim­it­ed, so they can­not make rock­et-sci­ence cal­cu­la­tions in sec­onds. Machines can do that. And if a machine gets human capa­bil­i­ties too, it will most like­ly lead to “infi­nite intel­li­gence”; it’s just because of that com­pound effect we dis­cussed ear­li­er in the article.

The biggest con­cerns are that machine intel­li­gence can become so clever, so effec­tive and so fast, that it might achieve the lev­el of arti­fi­cial intel­li­gence that we(our brain) can­not imagine.

Read: the con­cept of AI?

This is why many researchers are work­ing on machine — learn­ing algo­rithms. They’re try­ing to boost their machine’s capa­bil­i­ties by teach­ing them how to learn like humans. The more they know, the smarter they become. And why not? We do this all day long with our human brains. After all, the human brain is a work of art when it comes to learn­ing new things. Fur­ther, if we can teach machines to think like humans and process infor­ma­tion in the same ways, we can cre­ate some pow­er­ful machines in the near future.

Also Read: Can AI sys­tem be mod­eled on the human brain?

To Conclude,

While such a run of infi­nite intel­li­gence for machines is high­ly unlike­ly to be devel­oped in the next few decades, it’s pos­si­ble that some­thing like this could still hap­pen in the future; there­fore, it’s best to be prepared.

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.