We encounter news about artificial intelligence (AI) constantly. The coverage oscillates between enthusiasm for AI’s potential benefits and anxieties about job displacement and even threats to humanity.

This article won’t delve into artificial general intelligence or a dystopian AI seeking human destruction. Instead, we’ll concentrate on AI as it exists today—primarily algorithms designed for predictions. We’ll explore the economics of AI and its potential impact on business, drawing heavily from two recommended books: Prediction Machines: The Simple Economics of Artificial Intelligence and Human + Machine: Reimagining Work in the Age of AI.

This article unfolds in three sections:

• The Evolution of Technology provides a historical overview, highlighting parallels between past technological shifts and the current AI age.
• AI Economics: Strategy and Business Models examines how enhanced prediction accuracy through AI can reshape strategies and business models.
• Human and AI Interaction investigates how humans and AI can complement each other’s strengths.

The Evolution of Technology

Let’s examine historical events that mirror our current perceptions of AI, illustrating how technology adoption has transformed our thinking. We’ll trace the journey from rudimentary arithmetic to specialized artificial intelligence development companies.

The Age of Electronic Computers

Computers excel at arithmetic. Before their modern incarnation, “computer” referred to humans performing calculations—a role now known as a “human computer.”

Technological advancements made computing increasingly affordable and efficient, leading us to view problems through an arithmetic lens. Photography exemplifies this shift. Historically, photo manipulation involved chemical processes. Today, algorithms within widely accessible software suites enable mathematical image editing.

This exemplifies our response to cost reduction: we seek solutions through readily available technology. AI follows the same trajectory. As AI becomes more economical, its advantages are being harnessed by a growing number of sectors and individuals.

The Internet Age

The internet’s rise brought about dramatic cost reductions across industries. The declining cost of distribution fueled the e-commerce boom, forcing businesses to adapt or perish. As online accessibility increased, so did our reliance on it. Encyclopedias were eclipsed by Google and other search engines.

The Age of AI

AI is becoming more affordable, both in terms of computing power and tools. New tools and libraries expedite machine learning development, while the growth of GPU computing further reduces costs.

This affordability is reshaping our thinking, prompting us to frame diverse problems as prediction challenges. While sales forecasting is a clear example, autonomous vehicle development was not perceived this way 15 years ago. Back then, autonomous vehicles were confined to controlled environments like factories, relying on if-else programming.

The paradigm shift occurred when engineers began viewing it as a prediction problem. By training AI on human responses to various conditions, they developed onboard software enabling autonomous vehicles to navigate thousands of miles. Essentially, AI learned from human actions to predict its own, showcasing the power of framing problems through a predictive lens.

AI Economics: Strategy and Business Models

A crucial question arises: will AI revolutionize business strategies and models? If AI is seen solely as a decision-making prediction tool, its impact on strategy might seem limited. However, when viewed as a highly accurate forecasting tool, the implications become more profound. Prediction Machines: The Simple Economics of Artificial Intelligence offers a compelling illustration.

Amazon ships purchased items to our designated addresses. While browsing, their recommendation engine suggests products. We may not purchase every recommendation, but Amazon identifies potential interests.

Imagine Amazon achieving high accuracy in predicting our purchases. If we consistently buy 80% of recommended items, they might opt for a “shipping-then-shopping” approach, delivering items before purchase.

This change would necessitate adjustments to Amazon’s strategy. With returns anticipated, pricing models, packaging, and return handling would require adaptation. This strategic overhaul stems from AI’s enhanced prediction capabilities.

By considering the implications of highly accurate and cost-effective AI predictions, we can unlock further thought experiments like this.

Human and AI Interaction

What does the future hold for human-AI interaction? Will it be characterized by competition or collaboration?

We’ll focus on Human + Machine: Reimagining Work in the Age of AI, which posits scenarios where humans augment AI and vice versa.

Humans can complement AI in three key ways: training, explaining, and sustaining.

Training AI Models

AI relies on data for learning—the training phase. In the future, we may see dedicated “training agents” responsible for tailoring AI training to specific business needs. In a factory, this might involve robot training; in e-commerce, it could entail historical data aggregation.

Explaining AI-generated Results

Understanding the reasoning behind AI’s outputs is crucial.

A trade-off often exists between AI’s explainability and accuracy. While tools are being developed to shed light on black box AI decisions, a dedicated role might be necessary to interpret and communicate AI’s reasoning.

Sustaining and Overseeing Deployed AI Systems

Ensuring AI systems function as intended is paramount.

The 2015 incident where a Volkswagen factory robot fatally crushed a worker highlights the need for roles focused on AI system oversight and safety.

AI Complementing Humans

AI’s capabilities can empower humans by providing faster and more precise predictions. These “superpowers” enhance our actions, highlighting a key advantage of AI.

Amplifying Human Capabilities

AI tools enhance human abilities. In Human + Machine: Reimagining Work in the Age of AI, the authors cite Autodesk’s Dreamcatcher software, which leverages genetic algorithms to generate design iterations.

A designer aiming to create a lightweight, affordable, and durable chair can use Dreamcatcher. AI generates designs based on these criteria, presenting them to the designer. The designer then selects a design and adds their creative touch.

This exemplifies how computers have historically augmented human capabilities, but AI takes it to a new level by expanding the scope of assistance.

Interacting With AI Products and Services

AI can act as an interactive assistant. Amazon’s Alexa, Google Home, and Apple’s Siri exemplify this.

As these agents improve, they become increasingly integrated into our lives, much like smartphones have become extensions of ourselves. These interactive AI assistants augment our abilities.

Augmenting Humans on the Factory Floor and Beyond

Factories provide examples of AI-driven physical augmentation. While many factories employ robots, these are typically rule-based systems operating within safety cages.

AI-powered robots can collaborate with humans, designed for safe interaction and movement within a factory setting.

The Benefits of AI Outweigh the Drawbacks

Despite concerns about AI-driven job displacement, Markus Schӓfer, Chief Technology Officer of Development & Procurement at Mercedes-Benz Group, emphasizes a shift towards human-centered industrial processes, stating, “We’re moving away from trying to maximize automation, with people taking a bigger part in industrial processes again.”

New technologies inevitably transform our ways of working. However, just as the plow didn’t eliminate the need for farmworkers or computers didn’t replace mathematicians, AI’s role is not to supplant humans, but to propel us towards a new era of progress.