Artificial intelligence has become widely deployed across several industries. 2020 is set to become a banner year for the innovative technology.
Of course, bad news always gets preference and catches people’s minds. Some of the popular bad news in AI have been related to fake news generation, autonomous vehicles killing pedestrians, AI systems attacking a production facility and data biases creating problems in AI applications. Nonetheless, there are still some positive signs, we have seen innovative devices being deployed in hospitals, AI tools helping disabled people, robots being used in increasing set of domains as well as AI assistants and smart devices guiding people in day to day queries and chores. The speed of evolution, adoption, and research in artificial intelligence is accelerating. It will be important and essential for society to know what lies ahead on the road so that we are prepared for the worst and hopeful for the best. Here are 5 predictions of artificial intelligence in 2020.
Although one of the main drivers for the success story of AI in the last decade has been the availability of exponentially increasing data; now the data itself is becoming one of the key barriers in developing futuristic applications using AI. Advancements in the study of human intelligence show that we are very effective in adapting to unseen situations which contrasts with the current capabilities of AI.
In the past year, there has been a significant activity in AI research to tackle this issue. Specific progress has been made in reinforcement learning techniques that take care of the limitations of supervised learning methods requiring a huge amount of data. Deep Mind’s recent achievement is on top of the success stories in this domain. The Star Craft-II system developed by them – taking the throne of the Grandmaster – is a game-changer and an indicator of the tremendous progress and potential of this technology.
Generating data through simulation will grow at a much faster pace in 2020. For many complex applications, it is almost impossible to have data of all variety related to the different phenomenon of that problem. For example, Autonomous vehicles, Healthcare, space research, prediction of natural disasters, video generation are some of the areas where high quality simulation data will be much more effective. In most of these cases, real historical data will be too limited to predict new situations that can occur in the future. For example, space research is coming out with new inventions every day and nullifying old assumptions; in such a scenario, any AI application using historical data is bound to fail. However, simulations of new possibilities with high precision software applications can alter the direction of AI applications in these domains.
Even in the cases where the applications are starving for additional training with local data; public and private organizations are coming forward to share and collaborate for data requirements. Leaders are becoming more conversant with the requirements of AI and the mindset is changing. All these factors combined will have a dramatic effect in 2020 and will bring the dependency of AI on data to a lower level. AI will come out of the data cage.
We have seen the prototypes and demonstrations of content-generating Robots in the form of user reviews, news stories, celebrity images, funny videos, music compositions, short stories, and artistic paintings. This is going to become sophisticated with the advancement in self-supervised learning led by NVIDIA, Google and Microsoft; which are pushing the boundaries to new frontiers.
Most popular online retail stores, food portals, hotel & travel aggregators, etc. are based on customer reviews. Till now, these were written by real customers and real humans. Most of us were putting our faith in the crowd and take their reviews at face value. This has become a key component in driving new sales in different business segments. So, we were relying on crowd intelligence. But with these new content generation robots, all such businesses will be flooded with AI-generated reviews and it will be very easy to fool the customer.
Another Critical area is opinion formation regarding various news, events, and issues concerning the society. Social media, online campaigns, messaging through different mobile apps have become a key resource to build public sentiment on important issues. This is another area that is facing immediate danger of Artificial machines taking over human beings to form an opinion.
Next year this trend will consolidate and there will be a visible effect on democratic governments. AI may become a key driver and a primary campaigner for elections. Those organizations, Individuals or parties having AI supremacy will be able to win the elections and drive the world.
With the tremendous success and improvements brought by BERT and GPT-2, the Language translation is coming of age. People talking to anyone outside their community will be talking through the language of AI middleware. In 2019, we have already seen devices which can help you converse with people speaking other languages. The offerings are going to become more qualitative and inclusive. More and more languages are being added with an amazing pace in such conversational devices. The impact of such technologies coming in mass usage will result in a plethora of applications being developed resulting in a great impact on business and society. Movement of people, skills, and knowledge across borders with different language speakers will become more common. This will also bring transformation to the cinema, performing arts and travel industries. This phenomenon will also affect the higher education sector and affect different countries in diverse ways. It can prove to be an economic bonanza or a disaster based on the way the countries plan and embrace the changes. Proactive leadership that understands the future impacts of these technologies will be crucial to bringing a considered transition of society and the happy future of these countries.
AI is working in the same way for the powerful as industrialization and digitalization. People with resources are deploying and utilizing new-age technologies to their advantage. They have resources to invest in new applications and become the first adopters of technology. The power of artificial intelligence is being combined to optimize manufacturing and energy production. It is also being used to increase the efficiency of distribution networks, delivery chain, connectivity. Every big business is progressing to further increase AI adoption including airlines, shipping corporations, mining companies and infrastructure conglomerates. Eventually, AI is further intensifying the divide between the haves and have-nots. Common people are becoming pawns in the hands of AI applications. Their privacy is under attack. As the cost of labor is devalued due to automation and new technologies, the wealth will be owned by a tiny percentage of the people in the world.
Genuine voices, groups, and organizations should strive for the development of technology with a human face. Already, the UN and other groups are working for sustainable development goals. Now, it is time that a proper framework is put in place involving all stakeholders so that the pace and direction of technology remain under the control of humanity. It will involve developing comprehensive moral, ethical, legal and societal ecosystems governing the use, development, and deployment of AI tools, technologies and applications.
Few countries in the world are already in the advanced stages of developing lethal autonomous warfare systems. Sea Hunter, an autonomous unmanned surface Vehicle for anti-submarine warfare is already operational. China is in the final stages of deploying an army of micro swarm drone systems which can launch suicidal incognito attacks on adversary infrastructure. Other permanent members of the Security Council are working on Holistic Warfare Systems which are fully integrated with other functions of the government. With a complex set of adversaries in place, Israel is working to use AI as a force multiplier and to take fast decisions in the prevailing nebulosity of hybridity. AI also helps greatly in asymmetric warfare.
AI has unlimited potential to launch cybersecurity attacks of complex nature which will require adversaries to have superior AI capabilities to counter. As major financial systems of the world are online including banks and stock markets, they may become easy targets of future AI systems for blackmailing and threatening governments.
In recent years we have seen a significant increase in AI-related defense budgets to help AI-enabled weaponization. This is going to further accelerate in the coming year(s). Precision attacks on individuals, distribution and infrastructure networks of the countries will be enhanced by AI. We have already seen a precision attack powered by the US and Israeli Cooperation in Iraq, which resulted in the killing of Iran’s Top Commander.
With all these trends in the pipeline, it will be vital for the organizations, countries and the world to set their artificial intelligence strategy in place. Having competent individuals who are experts in AI will be indispensable and essential for survival in this new decade. We will need more data scientists and AI engineers who understand human and machine operated ecosystems to make emotionally sound judgments that are beneficial for future progress.
Deepak is a Professor and Department Head of Computer Science and Engineering at Bennett University, Greater Noida. He is also a chief consultant for algorithmguru.com, a resource for algorithms. Deepak is considered as one of the best Algorithm Gurus in India. His active research interests are designing efficient deep learning algorithms. He is Senior Member of IEEE, Senior Member of ACM and Life Member of CSI, IETE. He served as chair of IEEE Computer Society, India Council from 2012-2016 and on the Board of governors of IEEE Education Society, USA (2013-15). He previously worked as a Software Engineer in IBM Corporation Southbury, CT, USA. Deepak has 100+ publications in various International Journals and conferences with 500+ citations and Google h-index of 13. In his 20 years of experience, he has delivered 200+ invited talks across India. He is an ABET PEV and serves on ABET criteria review committee of Computing Accreditation Commission. He is also an active blogger in the Times of India with the tag name of “Breaking Shackles” and is passionate about transforming the landscape of Higher Education. For more details about his accomplishments: http://www.gdeepak.com. Deepak holds a PhD in Computer Science and Engineering from Thapar University.