H
ello everyone. Reading this on your "smart" phone, or your tablet?
Everything seems to have an element of smartness now-a-days isn't it? But, does it really make the world a better place to live, for humans and other beings alike?
...we'll ponder upon this and some other questions in this article, and see what picture of an intelligent world are we able to paint...!
Ever heard a roaring robo in an utterly ugly voice? Heard the new Google Wavenet voice? Okay let's hear them and see for ourselves...because that will probably give us an idea of how much progress we have made in AI...
| The DECtalk system was one of the dominant text-to-speech systems in the 1980s and 1990s...(here is a demo) | |
| WaveNet is a deep neural network for generating raw audio. It was created by researchers at London-based artificial intelligence firm DeepMind. |
You can see how advancements in our understanding of Intelligence have led to a huge improvements in our capabilities of mimicking the reality. Also is evident that a new race has ensued between tech giants. A race to be ahead in AI, and a thirst for more user data.
1. The Idea of Intelligence
The AI that we have developed so far is far from intelligent. As a result the question "the difference between Ai and true intelligence" becomes very easy one for a layman to answer, not requiring much knowledge or insight. It's easy to say that a "being" capable enough to take its own decisions and perform tasks autonomously can be called as being "intelligent", but is that all? Or rather does it even partially answer the question? Maybe I should leave it for the reader to decide.
But, just thinking a bit on the topic and the traits of human intelligence, we can say that there are many aspects of Intelligence.
A child being taught to perform a particular task like speaking a word : does that require intelligence on the part of the child? Well, lets try to see for ourselves.
To speak a particular word, the child needs to produce a succession of sounds which themselves require him/her to be able to use his/her muscles in the mouth and the sound box properly. The thing that is hard is just the part where they learn to use a particular muscle for the first time, once done they can simply use that sound in other words. If a robot were given a set of hardware like speakers and other peripherals all properly connected to work and produce sounds when used in particular ways, then maybe we could call it (the robot) intelligent if he could learn on his own how to produce a particular sound by trial and error, when the exact methodology to produce the sound was not available to it.
But really is a brute-force trial-and-error learning a sign of intelligence at all?
Imagine you were given a 100 page telephone directory with 20 contacts on each page. Now you were told that you need to search for the name Martin Medro in it. Will sifting through each and every contact in the 2000 strong list and matching it to the given name be intelligent?
Not really!
Rather one would expect you to intelligently figure out a way by either looking at the letterings on the margins of the page which give a clue of the starting letter...and so on, so as to complete the task in the minimum amount of time...
But, is just the simple condition "minimum time consumption" sufficient enough to truly define intelligence?
Now-a-days computers are quite fast with advanced processors and GPUs and a load of RAM. Speed is no longer a concern, even a long brute-force search can be completed in a fraction of a second. So, is speed a proper substitute of "intelligence" or "smartness"?
Again, Not at All!
So, though intelligence requires learning through trial-and-error, a brute-force use of a methodology can never be considered intelligent, however fast it may be performed.
Hence-forth we can say that an intelligent being should be able to find for itself the most optimum method to perform a particular task given a predefined set of resources.
2. Can Intelligence be "taught"?
Not really! Intelligence has to do with thinking out of the box, yet not using brute-force use of random thinking process, but rather taking a directed weighted approach to discard the useless ans bring out the useful with as minimal effort in as minimal number of steps as possible.
The larger the weights and the more accurate the weights, the more intelligence the being will possess or exhibit. As stronger weights will help distinguish the useful from the useless faster.
3. Where does probability drop in?
Well, what do you do when you are given a one marker MCQ (qualitative) in an exam with only one correct answer, and you are not exactly sure of the answer?
We try to "guess" the best possible answer by assigning weights to the different options based upon our knowledge. These weights can be thought of in terms of probabilities. So if out of A, B, C and D. You guess the following:
| Option | Probability of it being the right answer |
|---|---|
| A | 25% |
| B | 70% |
| C | 2% |
| D | 1% |
Then of-course the answer would be B. Now, let's look at another similar quiz. This time the probabilities are intelligent being gets for the options is as follows:
| Option | Probability of it being the right answer |
|---|---|
| A | 23% |
| B | 26% |
| C | 24% |
| D | 27% |
In this case, the answer that the person gets is: D, but as you can see the probabilities are roughly the same, this indicates either the quiz was tricky or the person is not intelligent at all. Because the probability chart of a perfectly dumb person (dumb with respect to the level of hardness of the quiz!) would be as perfect as:
| Option | Probability of it being the right answer |
|---|---|
| A | 25% |
| B | 25% |
| C | 25% |
| D | 25% |
This is how neural networks actually work and form the basic elements of an AI system.
So, whatever we have seen thus far gives us a fleeting review of the very basics of AI. But, what we actually are concerned more in this article is the question about the impact of AI on human society. Are we ready for it? Do we need it? Who'll be responsible for it?
Let's first see why we should be concerned? Are we really in danger?
4. The difference between true / natural intelligence and AI
Can AI ever compete with natural human intelligence?
Natural Evolution is itself somewhat an intelligent process. There are infinite possibilities in the characteristic traits a monkey can have, and all are equally probable to take birth and come into existence. And then natural selection applies the weights to each of the different types of monkeys coming into existence and decides which genetic makeup of a monkey is the best suited for the environment.
That's how evolution takes place, and has been doing for the last 4 billion years of the existence of life on Earth.
Can we really compete with 4 billion years of intelligent evolution?...or are we unnecessarily messing the natural beauty? I don't know...maybe we can, maybe not...I leave at as open question for the reader to ponder upon.
5. Where did you meet AI last?
Google Search, Amazon Search, Microsoft Cortana?
Never met any of the above before? IMPOSSIBLE!
Almost everyone of us has interacted with one of these and now how futuristic it feels, but behind all that delusion possibly lies a stark reality. you know behind AI is data. Every AI system depends upon what we call Artificial Neural Networks (ANN), which are built to mimic the neuronal connections in our brain.
 | A representation of a simple neural network with only one hidden layer. |
| A computer simulation of the branching architecture of dendrites in pyramidal neurons. |  |
In the diagram of the simple neural network given above, all the circles or nodes are nothing but weights (or rather the probabilities) we've already talked above. So for example the aim of the neural network is to be able to identify the letters of the English alphabet (for this we'll need only on node in the input layer). We'll be providing the ANN with images of hand-written alphabets asking it to classify them (this is called the "Task"). But for it to be able to perform the "Task" properly, we'll first need to train it just like a new-born baby is trained to recognise the different alphabets. To train it, let's start with the letter A. We provide it with thousands of images of the letter A written by hand and ask it to register it as A. During this process it progressively changes the weights in its hidden layer to be able to give a final answer of A. Once it has been trained for A, we'll train it similarly for all the alphabets before finally testing it with a random alphabet to see how well it is at recognising them.
So, now we know that we need an enormous amount of data to train an ANN even to perform simple tasks like character recognition.
So,where does this data come from?...
6. Enter the "Big Data"
Wikipedia says:
Big Data:
Big data is a field that treats ways to analyse, systematically extract information from, or otherwise deal with data sets that are too large or complex to be dealt with by traditional data-processing application software. Data with many cases (rows) offer greater statistical power, while data with higher complexity (more attributes or columns) may lead to a higher false discovery rate.[2] Big data challenges include capturing data, data storage, data analysis, search, sharing, transfer, visualization, querying, updating, information privacy and data source.
We leak out a lot of data about ourselves all the time. The rise of Smartphones with tonnes of downloadable apps has only led to a steep rise in the amount of data we leave behind at platforms known and unknown carelessly not realising the potential risks.
All this data can definitely be analysed to recreate our everyday lives in extreme detail. Amazon knows it all Google only knows it better, and Twitter, Facebook and Instagram shouldn't be left out so we tell them our stories ourselves!!
All this data is painstakingly collected to the very tiny bit and analysed by algorithms developed my multimillion dollar companies to follow you every where you go, every ad you see, every choice you make is pre-biased by these friendly looking apps and services that earn money from selling your information directly or indirectly.
6a. The Delusion of Social Networks & Ai services
Ever wondered why the Social Networking Platforms are free?
 |  |
| (Mark Zuckerberg, Founder, Facebook) |
Really if Facebook was such an amazing platform, why didn't Mark Zuckerberg make you pay for it? ...or Amazon's Jeff Bezos charge you extra fee for using Amazon.com?...or Google say it will no longer give its search engine to public for free use?....
 |  |
| (Jeff Bezos, CEO Amazon ) |
Do you really think they are doing some charity work?!!....
6a.1. Are Social Networking Platforms social at all...or are they dividing our societies?
A very integral part of social interactions is giving each other space to live. That space need not be physical, but rather social. That fact that we need to embrace and listen to viewpoints other than those which align with our views, to listen to people who think differently, to accept each other and live harmoniously together.Do these social platforms allow for that? Or rather can they allow for that?
Facebook, Instagram and Amazon have one thing very much common to all of them. This is the fact that they tell you based upon your past search history, what you may want to buy next. But it is more like if I search very often for toothpastes they'll show me more toothpastes, and that's it. Or let's say if I search chocolates, they'll show me more chocolates.
They'll tell us what we should do / we may do based upon our past history. But, then that will mean that we can't change? Does that also mean that they want to us or rather force us to stagnate at what we are, and make our view-points stronger rather than more flexible?
So, if we follow a particular kind of people, Facebook and Instagram will show us only that kind of people saturating our minds in the long run to an extent that we cannot tolerate people with other view-points?... The same goes with Google Searches and YouTube.
OR...
Show us something that we may not like in the short run, but maybe helpful in building a better society in the long run, and thus put their own profits at risk?
6a.2. Reading the "useless" Privacy Policy.
Most of us tend not to look at the Privacy Policy, of the different products and services that we use daily. So let's get you the lessons in which you passed the exams just by clicking the button without reading the book, and see what you really are aware of, and what it really is:1) Instagram (Data Policy)
Here I am taking excerpts from the policy that can be read here: https://help.instagram.com/519522125107875
 | Did you know that Instagram reads all your image metadata? |
Image metadata is the data embedded in the image which includes the camera with which you clicked the photograph, the date, time and location (if available) at the time of clicking and other technical data like camera make, shutter speed, aperture, ISO etc.
So, does Instagram want to indirectly tell us that it thinks we can't find the right content for us on our own, so it will forcibly rob all our data to be able to show us the "right content"?
2) Whatsapp
Some extracts from Whatsapp's Privacy Policy: https://www.whatsapp.com/legal/#privacy-policy-information-we-collect
| Did you ever wonder, why Whatsapp needs access to all your contacts? |  |
...and why the app simply doesn't open without the contacts permission granted? Is Whatsapp (like Instagram) wanting to tell us that we wont be able to find our friends ourselves or probably we have a poor memory when it comes to phone numbers, and that only Whatsapp can do that task for us?
6a.3. Is Privacy more important or Convenience?
Posting your pictures on Instagram and Facebook and letting your friends know about your latest craze seems very "convinient" doesn't it?
Just download a "FREE" app pay a bit for the data required to run internet, and enjoy the feeling of staying connected.
Go to amazon.com and order some toothpaste without the need to get out of your bed, sounds very "convinient"?
But are we forgetting the fact that these are multi-billion dollar companies not a charitable organization working for you for free. Are we really so foolish that we think we are not paying for it?
We are paying them with our privacy, and our personal information.
So, what if we are paying them with our privacy?...let's have a look...
6b. Are we going into a technological dictatorship?
A perfect dictatorial rule can do whatever it likes, it can invade your privacy, can punish you for crimes you never committed, control your views through propaganda and force you to do what it likes and feels is good for you.Are we universally moving into such a scenario, this time it being not so tangible but hidden.
When a single company like Amazon or Google grows into a giant, it doesn't happen peacefully.
For all the customers that they are drawing to themselves they are killing other smaller rivals by taking away their share of customers. With no customers to serve to, smaller start-ups and companies simply die-off and the winners are giant companies like Amazon. Other larger rivals can be bought away with money!
Like Facebook bought Whatsapp!
That's how companies grow! Had these things not been happening, Google hadn't been as big as it is. And there would have been several other smaller Amazon's of sorts.
What happens when diversity is killed?
When companies like Amazon kill the diversity and are left unquestioned, and unrivalled, what all can happen?
You are forced to buy things starting from your napkin to chewing-gum to food and gadgets and everything else from Amazon!
You are basically caught! You have no where else to go. Wherever you go, you find Amazon!
 |  |  |
 |  |  |
 |  | ...and much more! |
There are over more than 71 services delivered by Amazon, one such list of services can be found here: https://www.minterest.com/list-of-all-amazon-products-and-services/
How much more dependent do we need to be on a single company?
Did you know Amazon even has it's own air transport services?
[ Amazon Air & Amazon Prime Air. ]
Amazon Air is a cargo airline operating exclusively to transport Amazon packages. Prime Air is a drone delivery service currently in development by Amazon. Operations were expected to begin in select cities starting late 2019; however, as of June 2020, the service has yet to materialize.
Amazon has killed the local markets and all the diversity around it.
...Is it really a good situation?
Are we forgetting that these companies lobby governments for their personal profits?
Don't tell me that Amazon has a passion for selling groceries and toothpastes which they have come to prefer over books. And that Facebook likes connecting people so that they can stay together, and that's where the story ends.
Oh some may say that they too need to pay their employees with salaries, so why not put some ads...?
Wait a second, what if I let go of my privacy for a bit of convenience, after all I am not doing anything wrong, so why be afraid?
6d. The down-side of letting go of your privacy
Sure, we may not be doing anything wrong, but does that mean that all our info can be outside in the hands of some third party we hardly trust? Does that mean that you can sell your info to robbers and thugs? Not really right?
Instagram and Amazon may not qualify as "robbers and thugs" in the usual sense, but the fact remains that like the former, the latter also use your private info to earn profits.
More over, they too know your weaknesses and make use of it to make a living. Not only this since these tech-giants are well known to have loads of data about our everyday lives in extreme detail, governments may fall back and forcibly retrieve this data maybe to "solve crimes". You would not want to get arrested because Facebook's data made you look like some Most Wanted criminal, and the police catches you based on your posts. That is becoming a reality in many countries round the globe....and people are unnecessarily being harassed. Though that may help police solve crimes better, but why at the cost of the privacy of others?...with power always comes a lot of responsibility. Will the authorities take this task responsibly?...
7. Data is the new war-zone, Wake up before its late!
 | UNDER Ai SURVEILLANCE!Stay Safe, Take Control of your Data and your privacy before it is too late. |
8. The Better Side of It
There is also a brighter side of AI where we have the self-driving cars being developed among other things...we'll talk about that in some later article.
I hope you learnt something new in this article, and that you'll be able to take better decisions in matters related to your privacy in the digital arena after having read this.
Thanks for sparing your time.
Best
M. Medro
Credits
- Surveillance Cover Image (Image by StockSnap from Pixabay) | Pixabay License
- Paul DecTalk, Source : Wikimedia Commons | License: CC-BY-SA 4.0 International
- Google Wavenet Voice Clipping [https://cloud.google.com/text-to-speech/] | License: Public Domain.
- Simple Neural Network | License : Public Domain
- Computer Simulation of dendrite pattern | License : CC BY 2.5
- Surveillance Sign Board, (Image by OpenClipart-Vectors from Pixabay) | Pixabay License
- Mark Zuckerberg image | License: Public Domain
- Jeff Bezos image | License : CC-By-2.0
- Amazon Prime Logo | License : Public Domain
- Prime Air | License: CC-BY-SA 4.0