The concept of Artificial Intelligence is not new.
By the mid-1950s, researchers had already developed programs enabling machines to solve problems and learn by themselves. The commercialization of AI started in the 1970s and continued into the next decade, with the deployment of expert systems, machine vision, and automated assembly lines.
However, one major issue was preventing the expansion of AI: the lack of large data sets. To be effective and accurate, artificial intelligence needs to be trained regularly with a vast amount of quality data; this prevents the decay or drift of the AI decision process (model). With the recent digitization of our society, the access to large digitized data sets is no longer a problem, and this context provides the required conditions for the deployment of artificial intelligence solution on a large scale.
It is now undeniable: the artificial intelligence revolution is well underway and it’s transforming several industry sectors. For instance, in the car industry, AI systems are used for collision detection, driving assistance, and maintenance planning. In retail, AI models are used to predict consumers’ online shopping patterns and improve their shopping experience by offering more personalized services such as sale assistance (chat bots).
In the banking sector, AI systems help identify fraud, detect anti-money laundering patterns, and better understand customers’ banking behaviors. Investment firms also use AI predictive models for trading on the stock exchange and investing in different markets. Finally, the manufacturing, warehouse, and supply chain sectors rely on AI systems to optimize manufacturing operations such as staffing, inventory control, energy consumption, and the supply of raw materials. AI is also used to create smarter robots for factories and warehouses.
Outperforming Human Beings
In general, there are three key areas in which artificial intelligence outperforms human beings:
- Analytical and strategic thinking
- Speech transcription, and translation
- Medical diagnosis
Regarding analytical and strategic thinking, artificial intelligence programs have been proven particularly better than humans in gaming. For instance, AI bots have been able to beat “human masters” in poker and chess. The complexity of these games requires strategy, intuition, and reasoning based on unknown information. As for speech transcription and translation, AI models have been proven faster and more accurate than human beings in transcribing speech to text and translating in different languages. Finally, AI programs have yielded more accurate medical diagnoses than human doctors, especially for cancer detection.
Artificial intelligence is also evolving into the realm of creative thinking. In September 2020, The Guardian published an op-ed completely written by an AI. The task was to write a 500-word essay on “Why humans have nothing to fear from AI.” The text was generated by GPT-3, an open source AI language model that uses machine learning to produce human-like text. Also, in 2018, the music industry welcomed its first album fully composed and produced by AI: “Hello, World!” AI is no stranger to the music industry since for several years singers and musicians have been using software that improve sound harmonization (e.g., autotune) and song composition.
Looking Into the Future
Looking at the future developments of artificial intelligence, three major trends should be highlighted. First, AI will transform the applications of other emerging technologies such as cloud computing, Internet of Things, and blockchain. This technology convergence will multiply the possibilities in terms of digital solutions and will increase drastically the penetration of AI in our society. Second, emerging technology will change AI and its capability. For instance, Quantum Computing will impact artificial intelligence by significantly improving its learning, reasoning, and understanding capabilities (Quantum Artificial Intelligence). Quantum Computing might be holding the key to unlocking a complex concept such as “Emotional AI.” Finally, in terms of application in our daily life, work places may change rapidly with the introduction of AI assistant that will support employees in their tasks and perhaps oversee their work in such areas as quality control and internal audits.
The rapid development of artificial intelligence in the past decade beg the question: should we be concerned by the future application of AI? There are certainly several ethical, legal, and socioeconomic challenges ahead. As a society we should consider critical questions, such as should artificial intelligence be eligible for human rights? For example, Sophia, an AI robot, obtained citizenship from Saudi Arabia in 2017. Also, how will the wealth created by AI be re-distributed in our society? This question is particularly relevant for industries that will rely heavily on robot and intelligent software to produce goods and services with minimal human intervention. How will our society manage the sizeable shift in the supply and demand of the workforce? More precisely, can we ensure that the artificial intelligence era will be inclusive and equitable?
Frederic Lemieux, Ph.D.
Frederic Lemieux, Ph.D., is the Faculty Director for the master's programs in Applied Intelligence, Cybersecurity Risk Management, and Information Technology Management and the professional certificate in Artificial Intelligence Management at Georgetown University. His research has focused on policing, homeland security, and cybersecurity.