top of page
Search
aislingreardon
May 9, 20196 min read

We aren't ready for Machine Super Intelligence

Updated: May 10, 2019

ARTIFICIAL INTELLIGENCE IS HERE

Whether it be the algorithms that decide what you see on a daily basis on your Facebook feed or smart fridges that order milk just before you run out we have integrated mathematical computation processes into our daily lives before we could take a moment to ever understand truly what the implications are.





But, it hasn't been the end of the world.

In fact, we have benefited hugely from the integration of artificial intelligence into our lives, and into our workplaces. When tasked with menial routine jobs AI systems can carry out the works faster with less mistakes than the human equivalent allowing humans to engage in more purposeful and engaging tasks.


But what is coming next is the biggest issue and if we weren't prepare for the artificial intelligence of today we sure aren't prepare for what's coming down the line;

Machine Super Intelligence.





Traditionally artificial intelligence would consist of hand crafted data inputted into an AI system to reproduce processed results.

However the new era of machine super intelligence is algorithm driven machine learning.

AI now longer required handcrafted data to be input into their system but rather they learn themselves from exposure to experiences and mistakes that are made and rectified.





This advances machines from the status of artificial intelligence, being able to apply their knowledge to various tasks in a particular field, to general intelligence across many different fields. In this capacity, machine intelligence will outpace human intelligence and the way in which our world and our culture works will be change forever.


Philosopher, Nick Bostrom likens AI to the Atom. Like the atom, Bostrom believes that the true power of AI will only lay dormant for a finite period of time. Only until human experimentation detonates its ferocity like it did with the atom in 1945. (Bostrom, N., 2015)


We need to be prepared for this explosion of Machine Super Intelligence so that we have a controlled environment for the eruption.

But how?


POLICIES AND LEGISLATION



The exponential speed of computer intelligence and machine learning has happened so quickly that we have had time to even contemplate the how to keep it under control.

"With the accelerating speed of technological development, regulators will be less and less able to keep up and produce constructive...guidelines. There just isn't time -- tech moves too fast." (Semykoz, M., 2018, V)


Up until now we have had autonomy over what computers process as we have had to manually input data that will produce logical objective answers. But the growth of these machines and human intellect has meant that these machines are being used to answer subjective and open ended questions without the input of the input of directly correlated data.


Machines are always learning and they give us an indication of what they have learned through their output but what are they learning in the meantime? How is this technology going to develop?


We don't know. But what we do know is whatever way these machines develop we need them to be aligned with human interest and human values. To be developed for the greater good and not as weapons of control or of destruction.


We need governments to seek advise and work side by side with technologists to ensure that we can introduce policies and legislation that control the purposes of and the uses of artificial intelligence.


When it comes to Machine Super Intelligence, misconceived or poorly defined goals can cost us greatly. Once the power of technological inferiority is taken from us by the AI (which it will be), we will need to make sure that all these years of development and all the data that we have been feeding these Super Machines have been aligned globally for the greater good.


The only way we can insure this alignment is to have Governmental Bodies introduce policies of development and regulations on what the conditions of AI development are.

It is now that we need to start creating a controlled environment for the intelligence detonation. Afterwards is too late.


CULTURE

Technology changes and develops quickly but culture does not.

Well, not as fast as this new way of machine learning.





" We know we can't compete with AI cognitively, and we can't perfectly comprehend the world AI is building. But we do have to live in it. " (Semykoz, M., 2018, IV) and so we need to make every effort to find human strengths within this new world.


We need to change how we see ourselves as humans. For years we have differentiated ourselves from all other beings on each because of our superior intelligence and now something else is coming along to threaten that. We need to shift our idea of human intelligence from computational, scientific and logical to emotional, ethical and philosophical.


" If AI beats us in the business of finding answers, humans are still better positioned to ask good questions....asking the right question is both a huge competitive advantage -- and one of the only things a human can do better than a machine." (Semykoz, M., 2018,III)


We have to change our education, our R&D, our search for meaning from a 'how' centric model to a 'why' centric model. We have to let go of the answers because AI will have them all far faster and far more concrete than we will ever have them.

We, in turn search for meaning.





But, " acknowledging the limits of our control is an enormous psychological challenge. We humans fundamentally do not like the idea. In fact, we'll go a long way to create the feeling that we control our future."


We need to try and change this fundamental human trait in order to best transition into the Machine Super Intelligence era and optimize our place in the world.

But, we're not ready.


ETHICAL BOUNDARIES

Concerns and fear regarding AI and Super Machine Intelligence range from Robots taking over the world to your iPhone knowing more about you than your partner but the most relevant and timely issue with Artificial Intelligence is that of ethical boundaries.





We are so ready to simplify our tasks as humans and workers and off-load responsibility onto computers, that our ethical responsibility has begun to slip also.


Zeynep Tufekci, techno-sociologist presses that "we cannot anchor computation for such subjective messy human affairs" (Tufekci, Z., 2016)


Humans can identify stereotypes or archetypes in life and in the way the world presents itself. They have draw on their experience to realise the reasoning behind these stereotypes and through human empathy and understanding make non-prejudicial and ethical decisions.


"AI programs are made up of algorithms, or a set of rules that help them identify patterns so they can make decisions with little intervention from humans. But algorithms need to be fed data in order to learn those rules — and, sometimes, human prejudices can seep into the platforms." (Harini, V., 2018)


Human empathy and understanding does not exists in computers.

Olly Buston, CEO of consulting think tank, Future Advocacy explained:

"For example, if an algorithm used to shortlist people for senior jobs is trained on data that reflects the fact that historically, more senior jobs have been held by men, then the algorithm's future behavior may reflect this, locking in the glass ceiling," (Harini, V., 2018)


This is only an example of something that is already occurring in the infancy stages of AI. When Machine Super Intelligence comes to fruition, we cannot comprehend the ethical difficulties we will have, handing control to beings created in our image with more logical and no empathy.





Machine Super Intelligence is near.

The technology is getting better and better every day and there is no way to stop or stunt it.

It is exciting, it presents so many new opportunities and possibilities but unlike any other revolution in history, it is happening to fast for us to comprehend - let alone adjust to it.

We aren't ready for Machine Super Intelligence.


--------------------------------------------------------------------------------------------------------------------------

BIBLIOGRAPHY





18 views0 comments

Comments

bottom of page