Copyright thebftonline
By Stephen Kwame ODURO The general definition for Artificial Intelligence (AI) is “the ability of machines to perform tasks that typically require human intelligence, such as learning, reasoning, and decision-making.” This means that AI relies heavily on existing data to train models and improve performance over time. For instance, AI systems like Google’s DeepMind use vast datasets from medical records to diagnose diseases more accurately than some human doctors, demonstrating how data-driven learning enables machines to mimic human cognition. With each day passing, AI seems to be taking over human activities, from minor to major tasks. So far, it has evolved largely without comprehensive laws or legislation to manage its growth. In reality, AI is advancing faster than society is prepared for. As stated by Dario Amodei, CEO of Anthropic, “AI is starting to get better than humans at almost all intellectual tasks, and we are going to collectively, as a society, grapple with it.” This statement implies that, in a very short time, AI will replace many chores currently done by humans, such as most secretarial tasks (e.g., scheduling meetings via tools like Microsoft’s Copilot), paralegal duties (like document review using AI platforms such as LexisNexis), and even some executive functions. This is happening because AI is improving daily at tasks everyone performs, including strategic decision-making that some CEOs handle. The fear is that AI will soon become so advanced that it will replace entry-level white-collar jobs, spiking unemployment rates. This could occur within the next five years. On a recent visit to the US, I encountered food orders being delivered by robots through services like Uber Eats, which navigate busy streets autonomously to deliver meals. Similarly, companies like Amazon are deploying warehouse robots that sort and package items, reducing the need for human staff in logistics. These developments mean delivery personnel and warehouse staff could soon join the unemployment lines as AI-driven robots take over. This trend of job losses across sectors has been recognized by Academics and Economists, who caution that AI could replace jobs rapidly. For example, earlier this year, a World Economic Forum survey showed that 41% of employers plan to downsize their workforce due to AI automation by 2030, with industries like manufacturing (e.g., Tesla’s use of AI robots in car assembly) already seeing significant reductions in manual labor roles. It is widely believed that AI will improve employment overall, but records suggest otherwise. In the US, statistics indicate that large tech companies like Google and Meta have been hiring fewer entry-level employees, leaving new graduates in high-tech fields with limited opportunities—a phenomenon known as “job displacement.” For practical illustration, during the 2023 tech layoffs, companies such as IBM cited AI automation as a reason for cutting thousands of positions in HR and customer support, where chatbots like IBM Watson now handle inquiries. It is therefore crucial to create pathways to mitigate these challenges, such as reskilling programs tailored to emerging AI-related roles. Some argue that the effects of AI on job losses will be felt more harshly in less developed countries especially across Africa, than in developed economies. This is a flawed perspective. Manufacturers always seek low-cost production to boost profitability, and they will increasingly turn to AI automation, such as robots performing assembly line tasks. Despite lower labor costs in developing countries like Ghana, AI deployment will gradually replace cheap labor. For example, in the textile industry, companies like those in Bangladesh are already experimenting with AI sewing machines that outperform human workers in speed and precision, leading to factory job cuts. Currently, Ghana faces serious youth unemployment, and without intervention, AI automation could exacerbate this, skyrocketing unemployment and its consequences, potentially leading to social unrest. While some claim AI can reduce unemployment by creating new jobs, the question remains: Are we at a point where AI-generated opportunities sufficiently offset the jobs lost? One common argument is that “technology would automate lower-paying, lower-skilled jobs, and displaced workers can be trained for more lucrative positions.” However, AI advancements are outpacing new job creation by a wide margin. For instance, in the automotive sector, Ford’s integration of AI in quality control has eliminated thousands of inspection jobs worldwide, while the new roles in AI maintenance require advanced skills that many displaced workers lack. Again, Amodei noted that “if AI creates huge total wealth, a lot of that will, by default, go to the AI companies and less to the ordinary people,” as seen in the wealth concentration among tech giants like OpenAI, where profits from tools like ChatGPT benefit shareholders more than the broader workforce. How does society mitigate the advantages and disadvantages of AI? The challenge is that AI technology advances faster than human interventions. Recently, Meta announced its development of superintelligent AI models, which will accelerate automation and increase unemployment. Different countries have proposed solutions; for example, China has introduced a “Global Plan to Govern AI,” providing a framework that could especially benefit user nations rather than developers. A common international framework would establish processes for AI application. As we move toward widespread AI use, critical issues to consider include: Job Displacement (already discussed): Beyond delivery robots, consider how AI in healthcare, like IBM Watson diagnosing cancers, is reducing the need for junior radiologists. Privacy Concerns: AI processes massive data volumes, raising risks of breaches and misuse. A practical example is the 2018 Cambridge Analytica scandal, where AI algorithms exploited Facebook data to influence elections, compromising millions of users’ privacy without consent. Bias and Discrimination: AI can perpetuate unfair outcomes if trained on biased data. For instance, Amazon scrapped an AI hiring tool in 2018 after it discriminated against women by favoring resumes with male-dominated language, reflecting historical gender imbalances in tech. Ethical Considerations: AI can obscure accountability, as there’s often no clear source for decisions. In autonomous weapons systems, like those developed by companies such as Boston Dynamics, ethical dilemmas arise over who is responsible for lethal actions, with no global restrictions currently in place. Malicious Use: AI can destabilize societies; in Africa, where tribal conflicts occur, deepfake videos—created using tools like those from Adobe—could fabricate inflammatory speeches to incite violence, as seen in manipulated content during the 2020 US elections. Dependence on Technology: Over-reliance on AI for decisions could erode human skills. For example, pilots increasingly depend on autopilot systems in aviation, leading to concerns about skill degradation, as highlighted in investigations of crashes like the 2018 Boeing 737 MAX incidents where automation played a role. The Ghana Experience: Not long ago, Ghana’s Ministry of Communications held a retreat with a framework to: Demystify AI and its relevance. Explore trends and real-world applications. Address ethical challenges like bias. Align leadership with the revised National AI strategy. This is a positive start but already lags, given AI’s rapid pace. A better approach would be collaborating with advanced entities, such as partnering with companies like Google to train Ghanaian experts in AI. Emulating China, Ghana should enact legislation to regulate AI use, crucial for preventing malicious applications that could trigger instabilities. As experts advocate, “stronger regulation and ethical oversight as AI grows more powerful” must be prioritized. Additionally, universities with Computer Science departments should update curricula to include AI, perhaps incorporating practical modules on tools like TensorFlow for machine learning projects. In conclusion, AI is a necessary tool with profound benefits but also deep risks. While embracing AI, we must proceed cautiously to avoid exceeding our societal capabilities, ensuring equitable distribution of its advantages through proactive measures like global regulations and education. The writer is Computer Scientist with over 40 years of progressive experience in leading financial application systems development projects, designing, installing, and implementing for banking, insurance, brokerage and healthcare industries in the USA and Ghana. He is a former Managing Director of SIC Insurance Plc, Ghana’s largest indigenous general insurance company.
