Is problem solving a problem for AI development?
❝ When thinking is submerged with thoughts of ‘not’… then knotting that ‘not’ is the only plot to progress ❞
The use of artificial intelligence (AI) is growing across a wide array of business segments. From medical sciences to space engineering, the influence of AI in business is widespread. According to a recent report, the global AI software market is showing signs of rapid growth and is expected to reach around the US $126 billion by 2025 (Liu, 2021).
These trends point to the fact that AI technology is being deployed to build solutions for all sorts of uses, both from a business and customer perspective. One of the key functions of AI, therefore, is to provide a decision-making answer as an output to a user-driven input or activity. As such, it won’t be an exaggeration to say that the AI system operates on the notion of giving an answer to a question in relation to the problem that the user may be interested in. Artificial intelligence technologies such as machine learning/deep learning and natural language processing (NLP) essentially infer and predict from the data fed into them and provide an output that is just like an answer to a particular question or query for decision-making and further information processing purposes.
For instance, an AI system trained on MRI scans helps with a cancer diagnosis and treatment protocols. The data fed into the system is processed by the system to generate MRI scans that help human decision-makers answer the probability of the presence or absence of cancer in the patient (NIH, 2018; Recht & Sodickson, 2020).
Similarly, an AI system using natural language processing (NLP) technology to convert voice to text is simply solving the problem of having text transcripts of voice data.
What it means is that AI’s functioning is predominantly problem-solving oriented. Luger (2005; p.25) concurs and highlights that AI programmes are designed to solve useful problems. As such, the term “intelligence” in artificial intelligence causes confusion for people, as AI does not possess any genuine intelligence per se.
When it comes to human intelligence, it is important to recognize the various challenges that we face to describe and define what intelligence is. First, despite a lot of progress and development of knowledge about ‘how the brain functions”, the understanding of what intelligence remains elusive. Is it a collection of various abilities or a single faculty? What is perception, intuition and creativity and how such concepts are developed? What are cognitive capabilities and how they are developed? (Luger, 2005). Second, human intelligence is not limited to problem solving and decision-making alone. It is much more than that. Human intelligence involves interaction with the environment dynamically and responding to emerging scenarios and situations. The inherently dynamic nature of human intelligence involving cognition capabilities is not question-answer sequence focused. Third, how do creativity and intuitive intelligence drive the actions and behaviours of humans? How do creativity and intuitive intelligence manifest in real-time? These are some other areas that need more work to understand the human cognitive processes.
For AI machines to be called intelligent in any extant imagination, they need to possess at least some limited intelligence capabilities similar to those a human mind possesses. It raises the question of the current AI systems’ abilities and functioning are limited by their focus on problem-solving? Is problem-solving focus a problem for the evolution and development of AI? If so, what should be done for the further development of AI?
What are the implications for project management?
Just like any other discipline, the limitations of AI, as discussed above, carry implications for project management (PM) too. In regard to PM, AI is expected to help planning and delivery of projects. Hence, the current data-driven focus of AI could lead to potential issues. If the data is biased and is of lesser quality, it will influence decision-making and planning for the projects. Moreover, if organizations do not collect and maintain data in an organized and transparent manner, it will make it quite challenging to use machine learning technologies for predictions in relation to projects. Since projects are done across all industries and economic sectors, it also adds to complications for gathering data in a transparent and structured manner to use for AI purposes.
With that in mind, if there are no significant technological advancements in relation to AI’s intelligence capabilities, then perhaps AI’s role in the delivery of projects seems to be limited.
However, all the developments in AI must not be at the cost of bringing adverse effects on humanity. While leveraging the potential of AI technology is important, any developments must be within the bounds of ethical and harmless use and should be aimed at aiding the human living experience.
David, E. (2020) How The Future Of Deep Learning Could Resemble The Human Brain, https://www.forbes.com/sites/forbestechcouncil/2020/11/11/how-the-future-of-deep-learning-could-resemble-the-human-brain/?sh=3566d2f9415c
Liu, S. (2021). Artificial intelligence software market revenue worldwide 2018-2025
Luger, G. F. (2005). Artificial intelligence: structures and strategies for complex problem-solving. Pearson education.
NIH, (2018). Artificial intelligence enhances MRI scans, https://www.nih.gov/news-events/nih-research-matters/artificial-intelligence-enhances-mri-scans
Recht, MP. & Sodickson, DK. (2020), New Research Finds FastMRI Scans Generated with Artificial Intelligence Are as Accurate as Traditional MRI, https://nyulangone.org/news/new-research-finds-fastmri-scans-generated-artificial-intelligence-are-accurate-traditional-mri
© 2021 Jiwat Ram, All Rights Reserved.