AI’S ECONOMICS ANALYSED
AI ventures into the diverse field of economics, raising concerns about this convergence of technology and economics.
Artificial Intelligence (AI) is reshaping business operations and, in the process, revolutionising the economy. With the ability to swiftly analyse immense data sets, detect patterns, and forecast future trends with precision, AI profoundly influences economic dynamics. This transformative power not only generates fresh avenues for employment but also enhances productivity levels across industries.
Several countries are already leveraging AI. China's technological prowess rivals only America's, thanks to its tech expertise and the financial muscle of its internet giants.
Changing Face of Businesses
In a rapidly transforming global order, AI has become the change maker, yielding multiple benefits across a broad spectrum covering productivity, process, and end-user benefits. As a starting point, AI has the capacity to automate repetitive tasks and streamline business operations, thereby boosting productivity and efficiency. For instance, it can optimise inventory management, curb waste, and enhance profitability.
In order to facilitate better decision-making, AI can sift through extensive data sets, furnishing valuable insights that empower businesses to swiftly make well-informed decisions.
AI can deliver improved customer experience by personalising the shopping experience for customers by analysing their preferences and suggesting products accordingly. This can increase customer satisfaction and loyalty, increasing business revenue.
AI can also reduce costs in various industries, such as healthcare, by automating tasks and reducing errors. For example, AI can assist in drug discovery, accelerating the process and reducing costs.
When assessing artificial intelligence through an economic lens, we inquire about its impact on cost reduction, much like any other technological advancement. AI can be viewed as driving down the cost of a fundamental input crucial to numerous business and daily life activities: prediction.
To comprehend the profound transformations that occur when technology slashes the cost of a valuable input, we can examine the example of another innovation: semiconductors. Semiconductors decreased the cost of arithmetic, triggering three significant outcomes.Firstly, there was a surge in the utilisation of arithmetic for tasks that already relied on it as an input. Initially, this primarily encompassed government and military applications in the 1960s. Subsequently, we expanded the use of arithmetic for functions such as demand forecasting because it became more accessible and economical.Secondly, the reduced cost of arithmetic led to its application in solving problems not conventionally framed as arithmetic issues. For instance, we traditionally addressed the creation of photographic images using chemistry (film-based photography). However, as arithmetic became more affordable, we began employing arithmetic-based solutions in camera design and image reproduction (digital cameras).Thirdly, the decline in the cost of arithmetic altered the value of other components—the value of arithmetic's complements increased while that of its substitutes diminished. In the realm of photography, the complements were the software and hardware utilised in digital cameras, whose value surged due to increased usage. Conversely, substitutes' value, such as film-based cameras' components, dwindled as their usage declined.
As the cost of prediction continues to drop, we will use more of it for traditional prediction problems such as inventory management because we can predict faster, cheaper, and better. At the same time, we'll start using prediction to solve problems that we have not considered prediction problems historically.
As AI becomes more entrenched in the economy, its impact on various sectors must be assessed. For example, in finance, AI adoption can bolster fraud detection, improve financial forecast accuracy, and promote economic stability. In healthcare, AI can enhance medical diagnosis precision, decrease healthcare expenses, and enhance patient outcomes, positively influencing the overall economy.
AI in Economics
Economists already leverage machine learning, a subset of AI, for data analysis and economic projections. However, genAI, which underpins tools like ChatGPT, is a distinct technology that is advancing rapidly. Anton Korinek of the University of Virginia anticipates its transformative impact on research, enabling better and more productive solutions to societal economic challenges. Beyond research, genAI aids in teaching economics and forecasting inflation, likely empowering economists rather than displacing their jobs—at least in the near term.Korinek's paper highlights how large language models, a type of genAI, can aid economists in various research tasks, including ideation, writing, data analysis, coding, and mathematical derivations.
GenAI tools commonly include ChatGPT, New Bing, Bard, Claude 2, and LlaMA 2. These tools assist in brainstorming research ideas, copy editing, summarising text, and even coding tasks. While genAI enhances productivity, its mathematical capabilities are still developing, occasionally producing inaccurate information.
Regarding forecasting, a recent working paper indicates that genAI, exemplified by Google's PaLM, outperforms traditional economists in predicting inflation. This suggests that large language models may offer a cost-effective and accurate alternative for inflation forecasting compared to conventional methods.
As for its impact on employment, while genAI initially enhances economists' productivity, it may eventually lead to job losses, particularly in fields like software development. Svenja Gudell, Indeed's chief economist, acknowledges that while genAI could create better jobs, transitioning to this state might be turbulent. Ultimately, economists will still require a human touch, especially in teaching and public presentations, despite the integration of genAI into their workflows.
AI Strategy and Business Models
The shift of AI from decision-making to prediction has the potential to reshape business strategies. For example, consider Amazon's shift from "shopping-then-shipping" to "shipping-then-shopping" if AI accurately predicts consumer purchases. This transformation, driven by AI's higher prediction accuracy, underscores an unprecedented convergence of technology and economics, shaping global economies profoundly. There are some key areas in which AI can reshape business models and economics.
Data-Driven Decision Making. Data-driven decision-making in economics leverages advanced analytics and AI algorithms to extract insights from vast datasets, enhancing precision and effectiveness in economic strategies.
Predictive Analytics. Predictive analytics utilises machine learning algorithms to forecast future outcomes based on historical data, finding applications across industries such as finance, healthcare, and marketing.
Financial Modelling and Algorithmic Trading. AI streamlines financial modelling and enables algorithmic trading, providing nuanced market insights and maximising returns through automated trade execution.
Automation of Routine Tasks. AI-driven automation enhances efficiency by handling routine tasks with unmatched precision, though it necessitates workforce upskilling to address job displacement concerns.
Enhanced Resource Allocation. Advanced algorithms optimise resource allocation by analysing data and making informed decisions, ensuring efficient economic distribution.
Personalised Economic Experiences. AI tailors economic offerings to individual preferences and behaviours, revolutionising marketing strategies and financial services to enhance consumer engagement and empowerment.
The Flip Side
Job Displacement: While AI offers benefits, it's crucial to acknowledge its potential to replace humans in specific tasks. As AI algorithms advance, they can automate tasks previously handled by humans, such as data entry, customer service, and decision-making processes.Integrating AI into the economy can result in job displacement, particularly in sectors where AI can automate repetitive tasks.AI adoption may lead to job displacement in certain sectors, impacting the labour market and economy. Short-term effects may include increased unemployment and reduced consumer spending. However, long-term benefits may include enhanced productivity, cost reduction, and efficiency, fostering economic growth and generating new job opportunities in emerging fields. The emergence of new roles may necessitate a shift in required skills, increasing demand for creativity, problem-solving, and emotional intelligence.
Bias: AI systems may develop biases if trained on biased data, potentially resulting in discrimination and unequal treatment of individuals.
Data Privacy and Security: AI systems rely on vast data sets to function, raising concerns regarding data privacy and security.
Dependence on Technology: Increasing reliance on AI in the economy may lead to a lack of diversity in decision-making and a higher risk of system failures.
Country-Specific AI
The advent of new technology brings both the promising prospect of enhanced prosperity and the daunting fear of being left behind. Many innovations, like online education courses, have stirred more excitement than tangible economic growth in emerging economies. There is apprehension that generative artificial intelligence (AI) might similarly disappoint the global South, with the primary beneficiaries thus far being Western early adopters and San Francisco startups, alongside America's leading tech firms, collectively boosting their market value by an astounding $4.6 trillion since ChatGPT's launch in November 2022.
Nevertheless, AI holds the potential to revolutionise lives in the emerging world as well. As its reach expands, this technology could elevate productivity and narrow human capital gaps at a pace previously unseen. Developing nations need not merely receive AI passively but can actively shape it to align with their requirements. Most significantly, it could help bridge income disparities between developing and affluent nations.
The potential of AI in developing countries is enticing. Like in the West, it will serve as a versatile tool for consumers and workers, facilitating easier access to and interpretation of information. While some jobs may be displaced, new opportunities will arise. Due to fewer white-collar workers in emerging economies, the disruption and gains for existing firms may be less pronounced than in the West. According to the IMF, a smaller proportion (a fifth to a quarter) of workers in these nations face the risk of replacement, compared to about a third in wealthy countries.
However, one of AI's potentially transformative benefits lies in improving and expanding public services. Developing economies have long grappled with a shortage of educated, healthy workers. For instance, primary school teachers in India contend with twice as many pupils as their American counterparts without adequate resources. Similarly, Africa needs more doctors, especially well-trained ones. Consequently, many children grow up inadequately educated and in poor health, unable to realise their potential in an increasingly competitive global job market.
Policymakers and entrepreneurs worldwide are exploring how AI can address these challenges. In India, for example, large language models are combined with speech recognition software to assist illiterate farmers in applying for government loans. Meanwhile, Kenyan students will soon be able to ask chatbots questions about their homework, with the chatbots refining and enhancing lessons based on feedback. In Brazil, researchers are testing medical AI systems to aid undertrained primary care workers inpatient treatment. By leveraging medical data globally, AI could enhance diagnostic accuracy. If AI can make people in poorer countries healthier and better educated, it could eventually facilitate catching up with the developed world.
Furthermore, AI can be customised to suit local requirements. Few indications exist that AI is subject to winner-takes-all effects, which benefited American social media and internet search firms. This suggests that a variety of approaches could thrive. In India, some developers are already adapting Western models with local data to provide efficient language translation services, circumventing the hefty capital costs of model creation.
Another emerging concept gaining traction in the West is the development of smaller, more affordable models tailored to specific needs. Rather than aiming to process all available information, these models focus on a narrower set of capabilities. For instance, a medical AI wouldn't need to generate whimsical limericks like ChatGPT, but it could require computing power and specialised data sets to adapt AI in diverse and practical ways.
India's outsourcing industry might face disruption as generative AI assumes some back-office tasks, yet it boasts a thriving startup ecosystem, millions of tech developers, and a government eager to enhance its digital infrastructure using AI. Countries in the Gulf, such as the United Arab Emirates and Saudi Arabia, are determined to cultivate an AI industry as they transition from oil dependency. With existing capital and talent imports, they're well-positioned for innovation and adaptation.
Each country will shape AI according to its unique needs and priorities. Chinese chatbots, for instance, steer clear of discussing sensitive topics like Xi Jinping, while Indian developers concentrate on breaking language barriers. Meanwhile, the Gulf region is constructing Arabic language models. While the global South may not displace America's leadership in AI, it stands to gain substantially from this wealth of expertise.
Assessment
The influence of AI on the economy is nuanced, encompassing various aspects. While there's potential for AI to displace human workers in certain sectors, it can also enhance productivity, lower expenses, and improve efficiency, thereby fostering economic expansion and generating fresh employment prospects.
As AI advances, its influence on the future of economics grows, presenting fresh frontiers and opportunities. Guiding this transformative journey with ethical considerations ensures AI's responsible application for societal welfare.
As AI develops, public policy has a clearly defined role. Such a big structural change
would necessitate a strong government role to enable policy so that
the benefits
are well dispersed across society.
Questions for Experts
How real is the threat of displacement of humans by AI in any field and specifically economics?
Do you see any particular areas within economics where AI can be easily
assimilated?
What are likely to be concern areas for this convergence of technology and economics?
