science & tech

‘AIs won’t really rule us, they will be very interested in us’
The pioneering computer scientist prophesies that machines smarter than humans will emerge in the next two decades

Juergen Schmidhuber, 54, is a computer scientist who works on Artificial Intelligence (AI).

one of the pioneers in improving neural networks, his techniques, the best known being Long Short-Term Memory, have been incorporated in speech translation software in smartphones. In this interview, he speaks of developments in AI, why the fear of job loss due to AI is unfounded, and his work. Excerpts:

What is the most exciting AI project under way in the world?

I would be quite biased because I’d say what’s happening in my lab is the most exciting.

My goal remains the same as it has been for a very long time: to build a general-purpose AI that can learn to do multiple things. It must learn the learning algorithm itself (that can help it master chess as well as drive a car, for instance) — true meta-learning, as it’s called. We’ve been at it for 30 years and it’s getting more feasible over time. On this journey, we are producing less sophisticated but more useful stuff, like smartphones.

How impressed are you by AlphaGo, a creation of Google DeepMind, that now beats human Go champions?

DeepMind is a company that was heavily shaped by some of my students. Shane (Legg), one of the co-founders, was among those who worked in my lab. It’s great that you can play Go better than any human. On the other hand, the basic techniques (in making AlphaGo) date back to the previous millennium. In the ’90s, there was a self-teaching neural network by IBM that learned to play backgammon by playing against itself. So board games are kind of simple in the sense that they can use a ‘feed-forward’ network. (These are layers of neural networks arranged to mimic neurons in the brain. The programe makes decisions based on how information moves up these layers.) There are no feedback layers and they cannot ‘learn’ sequences. These principles were developed when computers were 100,000 times more expensive than today. It’s great that Go (like chess), which is so popular in Asia, is among those that machines play better.

What is Long Short-Term Memory?

It’s a technique in speech recognition and translation that many major companies — Facebook, Amazon, Samsung — are using and is based on work that we did in the early 1990s. It’s a recurrent network, a little bit like in the brain. The brain has a hundred billion neurons and each is connected to 10,000 others. That’s a million-billion connections and each of them has a ‘strength’ that indicates how much one neuron influences another.

Then there are feedback connections that make it (the network) like a general-purpose computer and you can feed in videos through the input neurons, acoustics through microphones, tactile information through sensors, and some are output neurons that control finger muscles. Initially all connections are random and the network, perceiving all this, outputs rubbish. There’s a difference between the rubbish that comes out and the translated sentence that should have come out. We measure the difference and translate it into a change of all these connection strengths so that they become ‘better connections’ and learn through the Long Short-Term Memory algorithm to adjust internal connections to understand the structure of, say, Polish, and learn to translate between them.

Given that self-driving cars are a reality, do we need AI machines to be regulated or do you think it could kill innovation?

Self-driving cars are now so well understood that you can, in certain countries, take some of them out on the roads. It’s old hat. They have been there since the 1980s and were implemented in the Autobahns. They went at 180 km/h, or thrice the speed of the Google cars, and went from Munich to Denmark on the highways. Back then, computers were 100,000 times more expensive. Now, thanks to Deep Learning (a way of organising neural networks and the zeitgeist of the field), pattern recognition since 2011 has vastly improved.

But what about people in self-driven cars who could make mistakes?

The nature of AI is that it doesn’t know. Machine learning is all about failing and learning from failure. It’s not like the perfect robots of Isaac Asimov stories. Were a car to sense a situation that could potentially lead to an accident, 99% of the time it’s going to brake hard. There will be flashing lights that will warn people in cars behind you that this car is going to brake hard and you should do the predictable thing of braking hard too and not swerving blindly. In some situations, that too may not be the perfect thing… maybe there’s another way to save a life by doing something really complicated. However, if there can’t be a fix to self-driven cars to address this and it leads to, say, one life lost per 100 million per day as opposed to 10 as of today (where manual cars are the norm), then lawmakers would move to make it mandatory to have only self-driven cars on the road. There could still be mistakes, but the law of large numbers says that on average, there will be fewer deaths from self-driven cars. This is also provided, of course, that insurance companies and such carmakers aren’t driven to bankruptcy. If better traffic is key to the better running of society, then systems will shift accordingly. In that sense, it’s no different from what has happened in the past too.

What about AI’s potential to destroy jobs?

Interestingly, people have predicted similar things for decades — for example, in industrial robots. Volkswagen and other companies had hundreds of millions of workers who lost jobs to robots. But look at countries with a high per capita presence of industrial robots — Japan, South Korea, Germany. They all have low unemployment rates. This was because lots of new, unanticipated jobs came up. Who could have thought there’s a job where people make money being YouTube bloggers? Or selling Apps? Some make a lot of money, some don’t, but it’s still a lot of new jobs. It’s easy to see what jobs will be lost but harder to predict what new ones will emerge. Societies must think of alternative ways to adjust to these new realities. There was a referendum on universal basic income in Switzerland. It failed, but still got 30% of the vote. You wait another 20 years and it could be 55%.

Do you think it will be possible for AI systems to ‘learn’ ethical and moral codes?

Anything that can be taught via demonstration can be taught to an AI in principle. How are we teaching our kids to be valuable members of society? We let them play around, be curious and explore the world. We punish them, for instance, when they take the lens and burn ants. And they learn to adopt our ethical and moral values. The more situations they are exposed to, the closer they come to understanding values. We cannot prove or predict that they are always going to do the right thing, especially if they are smarter than the parents. Einstein’s parents couldn’t predict what he would do, and some of the things he discovered can be used for evil purposes. But this is a known problem. In an artificial neural network, it’s easier to see, in hindsight, what went wrong. For instance, in a car crash, we can find which neuron influenced the other. If it’s a huge network, it will take some time, but it’s possible. With humans you can’t do this. You can only ask them and very often, they will lie. Artificial systems, in that sense, are under control.

Do you believe in the concept of super intelligence (when AI evolves to a level that far exceeds human capability)? Is there a date by which, given current progress, machines could ‘rule us’?

I would be very surprised if, within a few decades, there are no AIs smarter than ourselves. They won’t really rule us. They will be very interested in us — ‘artificial curiosity’ as the term goes. That’s among the areas I work on. As long as they don’t understand life and civilisation, they will be super interested in us and their origins. In the long run, they will be much more interested in others of their kind and it will expand to wherever there are resources. There’s a billion times more sunlight in space than here. They will emigrate and be far away from humans. They will have very little, if anything, to do with humans. It won’t be like a (dystopian) Arnold Schwarzenegger movie.

But can we go extinct or be exterminated like Neanderthals?

No. So, people are much smarter than, say, frogs, but there are lots of frogs out there, right? Just because you are smarter than them doesn’t mean you have any desire to exterminate them. As humans, we are responsible for the accidental extermination of a lot of species that we also don’t know exist. That is true, but at least you won’t have the silly conflict of Schwarzenegger movies, or like The Matrix, where bad AIs live off the energy of human brains. That, incidentally, is the stupidest plot ever.

Thirty watts (what a brain produces) and the power plant used to keep the human alive is much more. When should you be afraid of anybody? When you share goals and have to fight for it. That’s why the worst enemy of a man is another man. However, the best friend of another man is also a man or a woman. You can collaborate or compete. An extreme example of collaboration may be love — that is shared goals towards having a family. The other extreme could be war. AI will be interested in other AI, like frogs are interested in other frogs.

What’s the limitation to AI now — code or hardware?

It’s a little bit about code although the basic ideas are from the previous millennium. We have some breakthroughs but the dominant theme is that hardware is getting cheaper every year. In 30 years, it’s going to be a factor of a million. Soon we will have a small device that computes as much as the human brain.

In our lab, we’ve profited a lot from hardware built earlier by companies such as NVidia. They didn’t care for Deep Learning and only about selling graphics processors to the video game industry. But then it turned out that these were exactly what was needed to make neural networks fast.

Security

Seeing through a glass darkly
To deal with the terror threat, there must be far greater sharing of intelligence among agencies worldwide

Yet another anniversary of the November 26, 2008 terror attacks on multiple targets in Mumbai has come and gone. Much has changed since then and terror has evolved into an even more dangerous phenomenon. Recent variants represent a paradigmatic change in the practice of violence.




A different genre

It is difficult to recognise the new generation of terrorists as a mere extension of the earlier lot of radical Islamist terrorists who were influenced by the teachings of the Egyptian thinker, Sayyid Qutb, and the Palestinian Islamist preacher, Abdullah Azzam, and adopted the practical theology of the Afghan warlord, Jalaluddin Haqqani. There is less theology today and the new age terrorist seems to belong to an altogether different genre of terrorism.



This is not to say that the 2008 terror attacks in Mumbai were not different in the methodology and the tactics used in the September 11, 2001 attack in New York City.

Nevertheless, the spate of recent attacks in Europe and parts of Asia, from 2015 to 2017 — beginning with the attack on the Charlie Hebdo offices in Paris in January 2015, the major incidents at Brussels and Istanbul Ataturk airports as well as the Bastille Day attack in Nice, France, all in 2016, to the string of attacks in London, Stockholm, Barcelona and New York, in 2017 — are very different in structure and the morphology from attacks of an earlier period.

Standing out from the crowd

A large number of terror attacks in the past three years have been attributed to the handiwork of the Islamic State (IS), and reveal its leaning towards the “nihilism” of Sayyid Qutb. It is this which distinguishes the IS from many of the other radical Islamist groups such as al-Qaeda and its affiliates. The IS’s recruitment techniques, especially its ability to proselytize over the Internet, including “direct to home jihad” as also its more sanguinary brand of violence, set it apart from earlier variants of radical Islamist terror.



Even while the IS has gained a great deal of prominence due to its brand of violence, other terror networks have continued to be no less active. For example, al-Qaeda and its affiliates. The Boko Haram in Africa has been responsible for more killings than most people would realize.

Closer home, the Afghan Taliban and the Haqqani network have carried out several spectacular attacks inside Afghanistan. The Tehrik-e-Taliban Pakistan (TTP) and the Lashkar-e-Jhangvi have carried out several attacks inside Pakistan.

  Pakistan provides the wherewithal and the support to terror outfits such as the Lashkar-e-Taiba and the Jaish-e-Mohammad to launch well-planned attacks on Indian targets.



Most of these outfits continue to adopt earlier methodologies. These have proved no less effective than those followed by the IS. The terror attack on a mosque in North Sinai, Egypt in November this year, which killed over 230 persons, is one such example. In December, the TTP was responsible for a terror attack on an agricultural training institute in Peshawar, Pakistan. Differences among terror outfits, do not, however, preclude a complicated pattern of relationships when it comes to operational aspects.

Incorrect perception

Understanding the constantly altering trajectory of terror is important before charges of intelligence failure are levelled. It has become axiomatic to attack agencies of intelligence failure whenever a major terror attack takes place. This need not be the case in every instance.

The usual charge leveled is of the failure of intelligence agencies “to connect the dots”.

Most often, this is not true. There are many other reasons for adequate intelligence not being available to prevent a terror attack.
The danger is that a wrong diagnosis could prevent further improvements in intelligence collection and analysis.



One common fallacy is that intelligence agencies have remained static, are rooted in the past, and that their personnel are inadequately trained to handle current day intelligence tasks. While there is room for improvement, it is a mistake to presume that intelligence agencies have not made rapid progress and kept up with the times. Intelligence agencies today are well-versed in the latest techniques of intelligence gathering and analysis. Agencies obtain vast amounts of information from both human and technical intelligence, not excluding signal intelligence and electronic intelligence, intelligence from satellites and photo reconnaissance, etc. This is apart from open source intelligence.



Agencies employ data mining techniques and are familiar with pattern recognition software. Today, noise and signals constitute valuable meta-data. Analysing meta-data has produced more precise information and intelligence than is possibly envisaged, and agencies well recognise the value and utility of this.



In addition, intelligence agencies have become highly adept in monitoring and exploiting open source material. Mapping and analysis of social networks is today a critical aspect of their work. This is especially useful when it comes to unearthing covert terror networks. Many intelligence agencies today have an extensive database of several thousands of terrorists and potential terrorists.



Admittedly, intelligence agencies, like many other organisations, are risk-prone. They do make mistakes. Intelligence analysts, like analysts in other fields, are particularly vulnerable. Problems also arise from inadequate sharing of intelligence across institutions and countries. All these, however, are a far cry from the charge of an inability or failure “to connect the dots”.



The real problem is that when dealing with terrorism and terror networks, no two situations in the actual world are identical. The nature of threats is such that they continue to evolve all the time. Both the 2001 terror attack in New York and the November 2008 attack in Mumbai were one of a kind with few parallels at the time. Anticipating an attack of this nature remains in the area of an “intelligence gap” rather than an “intelligence failure”. Most experts explain an intelligence gap as one denoting an absence of intelligence output while an intelligence failure is one where, based on available evidence, no warning was issued.



Newer challenges

One of the major challenges that all intelligence agencies face is a qualitative understanding of the newer, and many post-modern threats. These newer generation threats, including those by terror groups and outfits, often lie “below the radar” or beyond the horizon. Anticipating such threats and their nature requires intelligence agencies to be constantly ahead of the curve. Anticipating newer threats is only partly facilitated by today’s technical advances such as new computing and communication technologies. However, these alone are not often enough to meet today’s intelligence needs.

As problems become more complicated, and as terror networks become even more sophisticated, there has to be recognition that the situation demands better understanding of factors that are at work. Levelling mere charges or accusations against intelligence agencies of a failure to anticipate an attack by not “connecting the dots” could be misleading, if not downright dangerous. All professional analysts in whichever field they operate face the same problem as intelligence agencies, and vividly outlined by David Omand, a former U.K. Intelligence and Security Coordinator as “seeing through a glass darkly when the information available to them is incomplete or partially hidden”.



Alongside this, and to fill the gap, there is a case for far greater sharing of intelligence and information among intelligence agencies worldwide than it exists at present. This is important to prevent another terror attack on the lines of the Mumbai 2008 attack. It now transpires that certain foreign intelligence agencies had additional information about the possible attack which was not shared in time, and which led to an intelligence gap. This could have been avoided.

More important, such a situation should never arise in the future. Terror and terrorism is a universal phenomenon. Every nation is bound to share the intelligence available with it to prevent a possible major terror attack.

M.K. Narayanan is a former National Security Adviser and a former Governor of West Bengal

E- way bill |20th dec 2018 |the Hindu Editorial | Gs 3 current affairs

E-way bill

What is E-way bill & Rules to generate the E- way bill ? 

E way bill is an electronically generated document, required to be generated for the movement of goods of more than 50,000 INR from one place to other.

Required for inter-state and intra-state transportation. under GST regime this is necessary after 1st Feb 2018.

This is a replacement of VAT (value added tax) regime's way bill.
Way bill was a physically generated document necessary for inter-state movement of goods.


Who will have to generate the E-way bill ?

the E- way bill is required to be generated by every registered or unregistered person involved in the goods transfer regime .


The hindu newspaper article - 20th dec 2017 : 

The Centre needs to do more to ease the shift to e-way bills for transport of goods

Already grappling with the "GOODS AND SERVICES TAX TRANSITION" Businessman are now more anxious about how the E-way bill will roll out.

Every goods moving more than 10 kms will require prior registration and generation of an e-way bill through the GST Network, which will be valid for varying duration depending on the distance traveled.


The problem is that few states have implemented this system and few hasn't . This poses an interim headache for firms operating across States, as they will now face differing compliance requirements for inter-State trade and intra-State trade, depending on when individual States launch their own e-way bill systems.


The exempted ones - Over 150 items of common use, including LPG cylinders, vegetables, food grain and jewelry, will be exempt from such transport permits, which can be checked by designated tax officials by intercepting a transporting vehicle. Goods moved on non-motorised conveyance, such as carts, have been left out.


Why the gov is concerned ?

In October, the GST Council had decided to introduce e-way bills in a staggered manner from January 1, with a nationwide roll-out on April 1, 2018


After a monthly ₹90,000 crore-plus inflow in the GST’s first three months, revenue in October plummeted to just over ₹83,000 crore.


With States claiming a revenue shortfall of about ₹40,000 crore so far under the GST, the Centre, which has to fill that gap, is also feeling the pinch.

Way forward :

Budget is less than 2 months now and economic advisors feels that with the rollout of E-way bill, the nation's economy will again fell a bump in tax collection.

Plugging revenue leakages is essential, and encouragingly, Karnataka’s e-way bill experience in the first month saw very few glitches.

Given industry’s nervousness, the government must simplify the onerous rules proposed for e-way bills (a one-day validity for distances up to 100 km, for instance), ensure that the IT backbone is robust, and make inspections the exception, not the norm.