Superintelligence

A few years ago, at a team dinner in Munich, the topic of Artificial Intelligence (AI) popped up. One person pointed out that our societies have always been afraid to adopt something new. Electricity, an indispensable utility today, was demonized a little more than a century ago. Cars didn’t fare much better when they were introduced. So, isn’t all the brouhaha about Artificial Intelligence just another instance of societal inertia? What’s different this time around? Some agreed, some didn’t, and the conversation went on.

I remember this because the next day, a somewhat scary notion settled into me. Your electricity usage couldn’t determine and predict your sleep patterns. And if you used your car everyday, it couldn’t automatically determine where you would be the next day at 8 AM. No previous technological revolution was intelligent enough to predict human behavior.

The difference this time is that there are now mechanisms where, through collection of continuous data (mainly via smartphones), it has become possible for machines to not just understand your behaviour, habits, likes and dislikes, but also (mainly via social media) to shape them. 

In even simpler words, for the first time in our history, machines are capable of understanding us more than any other human ever will. And given our increasing dependence on technology, this capability is only going to increase. It is therefore important that our children are already aware of how intelligent machines are going to impact us, so at least they get a chance to be ready for any fallouts that may (I believe – WILL) happen. 

I am here to make a case for mechanisms we need to put in place today so our future generations get that chance.

Can you define intelligence? 

While there are several implications for intelligence, for example, response capacity of human brain, adaption and adjustment abilities or all-round mental efficiencies yet no one could completely understand its meaning and beauty.

In process of understanding and implementing intelligence, the concept of ARTIFICIAL INTELLIGENCE evolved as a powerful tool for improving machine design and efficiencies. And since then, humans are constantly in the quest of creating artificial intelligence by giving more power to the machines. Intelligence is abstract and unique. This super power identifies humans uniquely from other living organisms. If we are determined to impart this superpower into machines, we must be ready for analogous good and bad machines or applications, in case of technologies (Here, I am trying to draw an analogy between good and bad human beings based on the thought-process to good and bad machines which are empowered equally or even more than human beings.)

There are pros and cons associated with this issue. Just like every REAL-WORLD proposal, model or situation has both, positive and negative aspects related to it, ARTIFICIAL INTELLIGENCE is no exception despite being “ARTIFICAL”! 

Lets just look at them one by one.

BACKGROUND AND ORIGIN

Today, AI is a buzz word and our survival is becoming dependent on it. But during 1950s, people of unaware of this term. Only in 1955, John McCarthy introduced the term- “ARTIFICIAL INTELLIGENCE”. The idea was to implement of analogous human intelligence into machines. Humans used the available information and reasoning skills to make decisions and solve complex problems. If machines could master these skills, it would be a boon to the society. I wish this assumption was true!

Figure 1 Evolution of Artificial Intelligence

With the advent of computers, the concept of Natural language processing boomed. Computers did not only offer fast processing and a remarkable ability to store larger information but also provided reliable and affordable means to the people. Research related to AI expanded around in the 1980s and finally AI technology was successfully established but suffered from lack of funds and public attention. 

But today, AI is reaching new heights. It is a primary attraction for researchers and extensive research is being carried out to propose and generate best applications of this technology. In the last five years, AI research has grown by 12.9% annually worldwide, hence it can be realised how fast AI is growing. It is predicted that in the coming four to five years, China will become the biggest global source of Artificial Intelligence. Not to forget, India is another such country where AI research output and work in the filed of AI is vast and is still increasing.

PRESENT SCENARIO

We human empowered machines and robots to learn and generate outputs. We added intelligence in them and hence, they are able to solve complex decision-making problems beyond the human capabilities of manual processing. Also, we are the ones who are scared of the negative impact of AI on environment, society and mankind. Amazing, isn’t it?

I sometimes wonder how we humans possess strange qualities like over estimating some possibilities and exploiting resources recklessly and on the contrary, regret and formulate strategies to mitigate the harmful effects. 

Over the years, a significant resurgence is seen in AI technologies. AI has become commonplace in every aspect of life – like self-driving cars, more accurate weather predictions, or early-stage health diagnosis’, just to name a few and all we can do is to welcome the future with open arms. AI is being taught to perform tasks that require human thinking and reasoning. We are optimistic that that this human intelligence simulation being fed to machines will yield maximum profits and faster production. Workplaces and organizations are becoming “smarter” and more efficient as machines and humans are starting to work together. With smarter technologies in the workplace, machines are interacting, visualizing the entire production chain and is making the decisions autonomously. The industrial revolution has made a lot of advancements in business possible and clearly there is a long way to go. 

THE SIMPLE AI MANTRA

Acquisition of information with applied set of rules yields approximate or definite conclusions with self-correction.

Today, our food, shelter, clothing, e-commerce, healthcare, education, anything and everything that you can think of, is automated! This marks a huge shift in our lifestyles, work patterns and interpersonal skills. And with superintelligence and AI algorithms, machines are capable of outperforming human beings in very single aspect. 

Mind you, “super”intelligence

No wonder, if we could the future generations super intelligent, the world would be a better place to live!

THE BIG QUESTIONS: which we chose to ignore 

In the hustle of automation and machine intelligence, do we realise the major drawback of machine learning and AI applications, or let’s say the major challenge for us in using AI applications? 

I am talking about emotions or emotional intelligence.  A computer ordered to make humans happy might bypass the tricky stuff and just pump us full of drugs – mission accomplished. The difficulty will be instilling human moral values into a machine, and since we have spent roughly three millennia arguing over what human moral values are with no end in sight as yet, that is a bigger task than the relatively straightforward one of making a really clever computer. 

The second important question is to check if the rising demand and use of AI applications is worth? 

The answer to this question might vary with perceptions. While the possibility of a AI device to be a calm and harmless assistant, the AI device machine might also turn out to be a huge disaster that might destroy life, that is, it can possibly kill its own creator because apparently “it is more intelligent that its creator but unfortunately less emotional than the creator”.

PROS AND CONS OF AI

AI has a wide area of application. While AI is widely growing and its use is extensive in all domains of life, we cannot ignore some of the bad consequences of this advancement. Let us discuss these two aspects one by one.

PROS OF ARTIFICAL INTELLIGENCE AND SMART MACHINES

  • From mining to fuel exploration processes, these intelligent complex machines overcome human limitations. Plus, they possess additional properties of not getting tired! Machines do not fall sick! They are immune to health issues and hence are more reliable.
  • Machines use previous data to make decisions and hence algorithms make their outputs error free and feasible.
  • Digital assistance has made our lives easier and efficient. From GPS guiding us through routes to auto correct feature guiding us through typing. Other prominent areas of applications include banks, educational institutes and government offices. Digital banks and online education have been serving Indian community with the best reach and response. 

AMAZONGO: SMART SHOPPING APP has been a hot topic in AI applications. The technique in general, is to track the customer and the store through sensors on the shelves and several cameras surrounding the store.  Customer’s entry, exit, bill payment and notification regarding billing information is all handled through his/her amazon account in the Amazon Go app. And you can literally “WALK IN, GRAB WHAT YOU WANT & WALK OUT! NO LINES, NO CHECK-OUTS (no, seriously)”, as the brand line reads.

Figure 2 AMAZON GO: A new way of shopping

  • The bright side of one of the challenges posed by AI, emotional independence! Yes, you guessed it right, mood swings or emotional imbalances do not hamper the efficiency of a machine (which is prominent reason for inefficiency among humans).

This is why logical machines are expected to always be productive! Companies and industries are extensively moving towards using “robots” and “smart machines” to communicate with clients, finalise deals and for public relations. This is because machines are completely logical and exhibit excellent communication skills.

SOPHIA, THE ROBOT- SOPHIA has been designed to learn and adapt to human behaviour and work with humans.

Figure 3 Sophia- The intelligent

GOOGLE ASSISTANT– It makes an actual call for you in the background, to schedule an appointment for you. Google Duplex: A.I. Assistant Calls Local Businesses to make appointments

Figure 4  Google assistant making an actual call to book salon

   When used as sidekicks, Artificial intelligence can prove to be a blessing.

  • Medical applications and treatments through artificial intelligence is indeed a blessing. Scientists are working on robots that can provide medical care to senior citizens.

CHINA’S 5G REMOTE SURGERY– How can one forget the first every AI based 5G remote surgery tested in China where the doctor was 5okms away from the operation theatre and a robotic arm removed the liver from a laboratory animal. A delay as low as 100 milliseconds was achieved in this surgery. Doctors from big hospitals can help their colleagues and serve the needy at time of emergency apart by this technology.

Figure 5 China’s first 5G remote surgery

  • Growing digitization and AI also affect the economy of a country as technological and digital transformation is going to shake and break the traditions of business. If we have to elaborate this concept to a layman, we can say that this technological advancement is going to discard the role of a middle man. Transactions, selling and buying has all become an online phenomenon now. 

Figure 6 Life made easier with AI and digital transformation

This is just a glimpse of how fast and effectively AI is changing trends in all aspects.

Figure 7 PROS AND CONS OF AI

CONS OF ARTIFICIAL INTELLIGENCE AND SMART MACHINES 

  • AI is expensive to implement. It uses complex machines and software programs that require regular upgradation to cater the needs of changing environment. Not only the installation cost, but maintenance is equally expensive.
  • There is no way that man can challenge the ultimate creator. Though intelligence is added to machines via algorithms and training, but one cannot add feelings, emotions and creativity to these machines. Even artificial neural networks could not (till date) match the robustness and parallelism of the biological neural network, human brain. 

So, machines can perform and generate outputs as per their training and input data.

No matter how much a machine outgrows, it can’t inherent intuitive abilities of the human brain and sensitivity towards certain issues.

  • Increasing involvement of smart machines and robots is overpowering the requirement of employees and workforce. More and more industries are investing in capital intensive technologies. Human beings are on the verge of being replaced by machines if don’t possess the required skills that are in demand. 

A recent study shows robots will take over more than 20 million jobs by 2030, thus creating mass unemployment.

THE NEW COOK, FLIPPY– An AI Powered Robot that can Cook Burgers. His job is to flip hamburgers at the CaliBurger Restraunt in California. As per a worker at the Cali Group, People find hard to work as it is greasy, hot and needs constant use of wrist! This caused inefficiency in service. Robot made hamburgers taste as good as human-made burgers and also yielded great profits for the restaurant because of efficient service. 

Figure 8 Flippy: next generation cook

Figure 9 Caliburgers: Flippy’s home

  • EXCESSIVE EXPLOITATION OF RESOURCES AND ENVIRONMENTAL DEGARDATION is another important aspect. 

These rapid changes are welcoming mega disruptions in the near future. 

THE FUTURE: GOOD AND BAD IMPLICATIONS TO SOCIETY 

Though we are bound to expect the unexpected when it comes to Artificial technology, yet there are some aspects where we can predict the future of AIs and their implication to the society.

  • REINFORCEMENT LEARNING – This is also called the “REWARD AND POUNISHMENT TRAINING” wherein the machine uses previous data to evolve and learn. 
  • Generally, there are 5 levels of automatic transportation depending on the degree of human control and interference. The firth level of automatic transportation which is the ultimate level of autonomous transportation requires zero human interaction and interference to operate. 

One example of such cars can be a robotic taxi. While, the CEO of Tesla, Mr. Elon Musk also claims that they will be ready for this level in 2020.

  • Advanced medical care facilities to bring a revolution in providing personalised medical services and further reduce mortality rates.
  • Improved and sustainable assistance in defence sector by using robots or intelligent machines for hostile situations. This will protect he peace of the nation and boost security.

Not to undermine the power of AI to be equally dangerous and destructive, here are some points on how AI can cause bad consequences. Because, if you watched movies like Avengers, Star Wars, Interstellar, The Matrix or Terminator, etc, all that they have is Artificial intelligence causing destruction everywhere.

  • As AIs are programmed and are automated, they may be programmed to do something dangerous or unethical.  Basically, they might be used as automated weapons.
  • Problems including social manipulation and invasion of privacy by targeting users randomly can hamper the public image and private life of the users. Also, users are always spied on (checked on by target marketing or facial recognition algorithms) which renders users uncomfortable and unsafe.
  • Trafficking, selling and buying of dangerous drugs and human lives are extensively practiced through navigation technologies and unmanned vehicles. These crimes are increasing day by day and are a root cause of abuse and misbehaviour.
  • Chatbots are programmed to start a conversation with users with the intention to sway them to reveal their personal financial or account information. Today, chatbot is a trend because more than 50% of the people prefer to text than talk over the phone- according to Kaspersky.  As per the records, more than 1000 Facebook users were tricked into installing malware that led to their account’s information being hijacked in the year 2016. 
  • Hackers are now able to develop undetectable malware with the help of AI. These help them to control webcams, steal, upload, modify and even manipulate files. Criminals write codes of computer viruses, use password scrappers, and other tools to execute their malware.
  • Do you receive emails that don’t look quite right??  Emails or calls where the sender claiming to be your bank, phone service calls, etc. Someone who knows some coding and HTML can do this without any difficulties. Machine learning improves the process of phishing by crawling any platform, learning and using the language to produce a numerous of fake emails based on statistical data and observations.
  • Neural networks in AI are used by criminals to detect and identify passwords. Passwords which are predictable and basic can put you in great trouble. Criminals manage to crack a portion of the passwords with the help of neural networks. Today, it is easily possible for a machine to crunch on some numbers and predict unknown passwords.

All thanks to our efforts of imparting superintelligence to existing systems by enhancing the  learning paradigms, dynamics, connections and algorithms of neural networks.

PREVENTION IS BETTER THAN CURE: POSSIBLE STRATEGIES TO MITIGATE THE HARMFUL EFFCTS 

The major consequences of artificial intelligence and digitization revolve around cyber crimes and complicated evasion of private/ personal information. One can prevent such things from happening by following the given steps:

  • We require good and efficient managements systems for our wireless/ wired devices so that they do not connect with unwanted and dangerous devices. ‘
  • Installation of anti-virus and regular renew of anti-virus can protect you from a virus attack and also helps detect the crime easily. 
  • Having a strong password including alphanumeric and special characters is very important. It is necessary to use different passwords for various applications and changing them regularly.
  • Shutdown the unused systems as they are sensitive to attack by AI through tolls and services related to the device.  Also, shut down the “locations” option of your phone (mobile device) when not in use. AI systems usually track the users at all times and if the information reaches a criminal, it might put the user in great trouble).
  • AI based security cameras in industries, companies and offices can be used to not only capture a theft but complete details of the thief by face recognition and intelligence. 

The best solution for the problem is to realise the goal of sustainable development by giving equal importance and attention to- 

1) ECOSYSTEM- NATURAL CAPITAL

Goal: Maintain ecological resilience by generating AI applications to mitigate climate changes, pollution and imbalance by reducing greenhouse gas emission and other harmful emissions. Intelligent technologies to provide basis for environmental conservation. 

(2) ECONOMY- MANUFACTURED AND FINANCIAL CAPITAL

Goal: Improve resource efficiency and manufacturing practices by using AI as a sidekick.

(3) HUMAN WELL-BEING- SOCIAL AND HUMAN CAPITAL

Goal: Enhance social equity and fair burden sharing. 

Figure 10  Earth must be our focus