These days often we hear the phrase “With great power comes great responsibility”. It sounds cheesy and superheroistic, but I think the time has come for us to actually take these words seriously and act accordingly. Standing in this era of modern science and technology, I feel fortunate to see the world as it steps into a new realm of modernization, reaching towards superheroistic capabilities. Nevertheless, we should also give a thought to the “what if”s that might pop up in a pessimistic mind. What if something goes wrong? What if the good guy becomes the bad guy? To put these questions into perspective, in this article I will touch upon a few promising upcoming technologies outlining their darker sides for our future and the role of the scientific community to ensure safety.
- Artificial Intelligence :
Do you love robots? Or do you fear them? Popular sci-fi movies like The Terminator, showed us that robots can be powerful and also lethal at the same time. But when we talk about robots in real life, we might think that they will never be as smart as us. Isn’t there a ‘power off’ button at the back of every robot so that we can just turn them off when they go out of control? Naively speaking, the answer is yes. We can control the robots that we make to some extent, but what if they learn to disable the ‘power off’ button? Believe me, the concern is real, if not imminent. Companies like Google, Facebook and IBM have poured in enormous resources over the years to develop robust machine learning algorithms that can solve real life problems beyond the abilities of traditional algorithms. An example of such a problem is a basic classification problem where a computer needs to detect whether someone appears in a picture or not. In 1997 IBM’s chess engine Deep Blue defeated Gary Kasparov, the best human chess player at that time. In 2016, another engine AlphaGo defeated the best human player in the board game called Go, which requires much more intuition, strategy and creative thinking than chess. These events laid the foundation of showcasing the power of artificial intelligence in the context of human developments.
The most fascinating part of machine learning algorithms is that no one knows how it works exactly. It is like a bunch of connected neurons in the human brain. Just like every day you learn something new, more and more new connections form in your brain and other connections get deleted, the algorithm also creates and deletes nodes in its network based on its performance with respect to a certain task. In this way, it can be trained. But all these efforts were targeted at solving only a particular kind of problem like for example a chess engine only learns to play chess and cannot detect a human face. So the question arises, whether algorithms can be built that can perform any random task provided to them? Not yet, but efforts are being made to develop algorithms that learns to learn [1]! So what if it learns to not obey our orders? Or what if it learns to prank the user? We will have to see. Tech entrepreneur Elon Musk said in an interview, “I do think we need to be very careful about the advancement of artificial intelligence.”
Another interesting question that we can ask is, can machines develop consciousness? But understanding consciousness itself is a bit tricky. Scientists believe the nature of correlations between different parts of a brain gives rise to consciousness. Recent efforts in studies of correlations in a hallucinating brain are providing key insights in this area [2]. But the answer to the hard problem of consciousness, or the origin of the first-person perspective of our life experience is still unknown. However, with the evolution of the structures in an algorithm just like a natural evolution in the biological world, it might be possible to realize a machine with consciousness [3].
To summarize, we are not very far from that day when machines will be able to learn any new given task faster than human beings. Also, with growing research in neuroscience and the understanding of consciousness, we might be able to develop conscious machines in the near future. Thus it is very important to organize and implement research ethics to ensure that the acquired technologies could be used effectively and controllably.
- Quantum Technology :
Quantum computers are the future of classical computers and it is true that quantum computers will revolutionize our society and our understanding of nature. The basic difference between a classical (a mobile phone or a laptop) and a quantum computer stems from the fact that classical computers work with information encoded in bits which can take values 0 or 1, while on the other hand quantum computers can access all the intermediate values from 0 to 1. This is because quantum particles, like electrons remain in a so-called superposition state where they have certain probabilities of carrying the information 0 and certain probabilities of carrying the information 1. Without further delving into the details, let us see what kind of benefits we can reap out of quantum technologies.
Quantum computers can be used to solve certain kinds of problems like factorization of a large number, exponentially faster than classical computers [4]. This is important because most online security protocols, like encryption of password of your Facebook account, uses factorization of large numbers, and it would take thousands of years for classical computers to factorize such large numbers and break the encryption. With the advent of quantum computers, that might not be the scenario anymore. Quantum computers are way faster than classical computers with respect to solving these kinds of problems. Hence it can pose online security challenges. But there is a silver lining to this problem. In fact if communication channels can be protected using quantum security protocols like quantum key distribution systems [5], it would be almost impossible for an eavesdropper to breach the privacy of the channel. This technology arises from a phenomenon known as quantum entanglement. It says when two quantum particles are entangled quantum mechanically, then measuring the state of one of the particles fixes the state of the other particle. This is true even if the two particles are far apart in space. Quantum entanglement can also be used to develop high precision quantum radars that can track enemy aircrafts in an airspace using entangled photons, leading to ultimate precision in detection [6]. This is a bad news for countries that poured in huge sums to develop aircrafts that are almost invisible to traditional radar systems. It would be a new challenge to defeat quantum radars.
Whatever we discussed so far about quantum technologies do not pose a big threat to human civilization. But, there is another avenue of application of quantum computers. That is biology. Big pharma companies like Silicon Therapeutics and Pharmacelera are looking into these approaches to use quantum computers and artificial intelligence, to simulate large and complex molecules, that cannot be simulated on a classical computer. This might lead to the development of better and more potent drug molecules and enable us to tackle a wide range of health care problems [7]. If successful, it might lead to a steep inrease in the global average life expectancy. Thus in the future, population inundation might become an issue unless measures are taken to check it. Also, if these technologies fall in the wrong hands, the effects can be devastating. Using quantum simulations and artificial intelligence, it might be possible to design viruses which can have mutation abilities that can be hard to defeat. It can develop into a serious threat with apocalyptic potential.
With several countries pouring in funds for the development of quantum technologies, the key take-away of the above discussion is that these technologies are powerful and can change our world, but we should also be aware of the consequences of misuse and thus lay proper procedures and guidelines to prevent future mishaps.
- Nuclear Technology :
With the rise of populist forces around the globe, the need for developing a powerful defense infrastructure is soaring. As we move into the future, probably more and more nations will try to harness nuclear capabilities for combat. Although a nuclear powered state can enjoy an additional sense of security and pride, with more and more members joining the nuclear club, this extra bit will soon erode out. Moreover emergence of nuclear powered rogue states can significantly damage global peace through threats of nuclear attacks and in the worst case can lead to nuclear warfare.
Another big challenge for any kind of nuclear technology is nuclear waste management. Irresponsible handling of radioactive waste can lead to serious radioactive contamination. It is very hard to control such contamination if it reaches natural circulatory networks like a river system. This can snowball into severe disruption of biodiversity and ecosystems. An example of how such events might be triggered was the failed covert operation of 1965 conducted by US’s Central Intelligence Agency (CIA) and India’s Intelligence Bureau (IB). In this mission, the Indian and American agencies planned to plant a nuclear-powered sensor at the summit of mount Nanda Devi (7815 m), a Himalayan mountain peak overlooking China. The objective of this mission was to monitor China’s nuclear activity using the sensor. Among other things, the setup also contained seven Plutonium capsules required to power the instrument’s battery. But carrying these heavy equipment (weighing around 50 Kg) through such mountainous terrain was not easy. During the uphill journey, the crew was met with hazardous climatic conditions and were forced to leave the cargo midway at Camp IV (7300 m). The next day when they came back to look for it, they could not find it. It was never found since then despite several search missions undertaken by the Indian government. Now, if the nuclear capsules which are somewhere buried in the snow, start leaking out radiation, then it can cause nuclear contamination in the waters of Rishi Ganga, a tributary of river Ganga, that originates from this mountain. Luckily traces of such contamination has not yet been found in the waters of the surrounding region and some experts opine of nonexistence of fatal risks [8].
Hence the bottomline here is that, in order to prevent such circumstances from arising, strict regulations should be imposed based on the ethics and responsibilities of using nuclear technology and needless to say, the scientific community should play a strong and univocal role to empower this initiative.
- Concluding Remarks :
When a historian of a future time (say 2100) starts writing the history of today’s world, s/he will definitely have a lot of content and excitement in describing today’s world. The amazing things that are being developed today are paving the way for a new epoch of science and technology. But if we do not take care of the existing loopholes in our systems today, there is a possibility that that historian will never exist; as our civilization might cease to be.
Enough of pessimism! Let us hope for the best and prepare for the worst.
References :
[1] Duan, Yan, et al. “Rl2: Fast reinforcement learning via slow reinforcement learning.” arXiv preprint arXiv:1611.02779 (2016).
[2] Schartner, Michael M., et al. “Increased spontaneous MEG signal diversity for psychoactive doses of ketamine, LSD and psilocybin.” Scientific reports 7 (2017): 46421.
[3] Koch, Christof, and Giulio Tononi. “Can machines be conscious?.” Ieee Spectrum 45.6 (2008): 55-59.
[4] Shor, Peter W. “Algorithms for quantum computation: discrete logarithms and factoring.” Proceedings 35th annual symposium on foundations of computer science. Ieee, 1994.
[5] Shenoy-Hejamadi, Akshata, Anirban Pathak, and Srikanth Radhakrishna. “Quantum cryptography: key distribution and beyond.” Quanta 6.1 (2017): 1-47.
[6] Barzanjeh, S., et al. “Experimental microwave quantum illumination.” arXiv preprint arXiv:1908.03058 (2019).
[7] Peters, Martin B., Kaushik Raha, and K. M. Merz. “Quantum mechanics in structure-based drug design.” Current Opinion in Drug Discovery and Development 9.3 (2006): 370.
[8] Broad, William J. “Himalayan Spy Device Said to Pose No Radiation Risk.” Science 204.4398 (1979): 1180-1180.
Debanuj Chatterjee is a PhD scholar of physics at ENS Paris-Saclay. Views expressed are personal.