AI’s Long-Term Impact

Artificial intelligence technology will have a significant impact on the world’s economy in the months and years ahead, the McKinsey Global Institute forecasts in a new report. Certain companies and some countries will greatly benefit in the new era of AI, leaving others behind, the business and economics research arm of McKinsey & Co. predicts.

The key takeaway points of the report:

• “AI has large potential to contribute to global economic activity.”
• “The economic impact may emerge gradually and be visible only over time.”
• “A key challenge is that adoption of AI could widen gaps between countries, companies, and workers.”
• “How companies and countries choose to embrace AI will likely impact outcomes.”

The report says AI potentially could add economic output of about $13 trillion by 2030, increasing global GDP by approximately 1.2% each year. World leaders in AI (mostly developed economies) stand to realize an additional 20% to 25% in economic benefits compared with the present, while emerging economies might see half of their upside, according to the report.

“AI is not a single technology but a family of technologies,” the report notes. The McKinsey paper looks at five broad AI categories: computer vision, natural language, virtual assistants, robotic process automation, and advanced machine learning.

Source: McKinsey Global Institute analysts

“The impact of AI may not be linear, but may build up at an accelerating pace over time,” the paper continues. “AI’s contribution to growth may be three or more times higher by 2030 than it is over the next five years. An S-curve pattern of AI adoption is likely – a slow start due to substantial costs and investment associated with learning and deploying these technologies, but then an acceleration driven by the cumulative effect of competition and an improvement in complementary capabilities.”

It adds, “Many developed countries may have no choice but to push AI to capture higher productivity growth as their GDP growth momentum slows, in many cases partly reflecting the challenges related to aging populations.”

China is heavily investing in AI tech, the report notes, seeking to overtake the U.S. in AI leadership. Kaifu Lee, an AI expert and investor, said at a San Francisco conference this month that China is quickly catching up in AI, leveraging its vast amounts of data, while the U.S. continues to lead in pure research for AI. Alibaba, Baidu, and Tencent have joined iFlytek, a voice recognition specialist, in a Chinese “national team” to develop AI for autonomous vehicles, medical imaging, and smart cities.

The European Union wants to invest $24 billion in AI research by 2020, with France and the United Kingdom leading the way in European research and development, and in the area of AI ethics.

AI is making progress in three key areas, the report states. They are: Step-change improvements in computing power and capacity; an explosion in data; and progress in algorithms. It could bring significant changes to manufacturing, marketing and sales, and supply chain management.

Deep learning techniques – feed-forward neural networks, recurrent neural networks, and convolutional neural networks, in particular – together could enable the creation of $3.5 trillion to $5.8 trillion in value each year in nine business functions, contained in 19 nations. Amazon, Digiday, Neuron Soundware, and Quantum Black are among the companies already putting AI technology to work in a variety of functions.

AI is being touted for a wide variety of applications, including but not limited to aviation, computer science, education, finance, heavy industry, hospitals and medical care, human resources and recruiting, marketing, media, music, news and other information, online customer service, telecommunications, toys and games, and transportation.

Niranjan Manohar, research manager for connectivity and automotive Internet of Things at Frost & Sullivan, sees a lot of AI work going on, especially in automotive electronics. “With AI, the car needs to learn you,” he says. “That’s where the AI engine becomes crucial.”

At the same time, AI is “still a very nascent technology,” he adds.

Manohar divides automotive AI into artificial narrow intelligence, artificial general intelligence, and artificial super intelligence. Artificial narrow intelligence is the easiest for companies, performing “one particular function,” he comments. “General intelligence is a bit harder compared with the narrow intelligence, and the super intelligence everyone is expecting is something that is the hardest for any AI company.”

Within the context of AI, the narrow intelligence is “the weakest link,” Manohar says. Super intelligence could beat the world chess champion.

Automotive AI is now primarily about improving the customer experience in the car, according to the analyst. “It’s not a one-size-fits-all, and it depends on the application for which it is being developed now,” he says. “A good example when we look at the immediate work being done is development of a virtual assistant and a real-time interactive HMI in the car. That would be the first real-world applications of AI in cars. The next part, what we actually think, would be the collecting a bit of driver-related behavior and sensor data, and will be the next level to building accurate maps for Level 3 automation.”

While Tesla touts its Autopilot technology as Level 3 automation, it is actually closer to Level 2, Manohar asserts.

Gaining consumer acceptance for autonomous vehicles will be essential, he notes. There are instances in field testing where the vehicle gets stuck in deciding what its next action should be, such as turning right or left.

Related Stories
Security Holes In Machine Learning And AI
A primary goal of machine learning is to use machines to train other machines. But what happens if there’s malware or other flaws in the training data?
IBM Takes AI In Different Directions
What AI and deep learning are good for, what they’re not good for, and why accuracy sometimes works against these systems.
Energy-Efficient AI
How to improve the energy efficiency of AI operations.