The Headline of the Future – Part II in an Exploration in AI
Artificial intelligence (AI) is generating exciting headlines lately. With stories covering seemingly infinite capabilities, from creating expensive art, to keeping the United States of America ahead in the international race to develop cutting-edge weaponry, to helping our homes take care of us, AI can be perceived as having suddenly embedded itself in virtually every aspect of our modern lives. While it is true that AI is a powerful technology with a broad range of business applications, AI is also a young technology still in the earliest stages of development. Due to its potential to achieve intelligence beyond that of humans, it warrants a thoughtful approach to remain beneficial and safe.
From Narrow AI to General AI
AI can seem like a nebulous concept to the average person, and separating fact from fiction can be a challenge. The Future of Life Institute (FLI), a multidisciplinary group of individuals whose mission is “to catalyze and support research and initiatives for safeguarding life […] including positive ways for humanity to steer its own course considering new technologies and challenges”, offers an easy-to-understand explanation of what AI is. They also emphasize that anticipating and avoiding problems with AI, rather than learning from costly mistakes, is paramount.
In its current form, AI is a software programming written by a human for the purpose of completing a single task. This is known as narrow AI, and it can drive a car, patrol social media platforms for inappropriate content, and diagnose diseases. Narrow AI is based on complex sets of rules that humans define and encode, and its successful implementation requires comprehensive datasets from which to draw input. Today, narrow AI enhances a human’s ability to do a job rather than replacing them entirely. Global businesses are embracing narrow AI as a means to not only gain competitive advantage but also to simply remain relevant in their respective markets.
The future of AI lies in Artificial General Intelligence (AGI), defined as software capable of learning and improving itself beyond the bounds of human programming and intervention. It is also referred to as strong AI or superintelligence in some contexts. AGI was once the stuff of science fiction, but developmental breakthroughs in recent years have accelerated conversations about its safety. High profile individuals like Elon Musk and the late Stephen Hawking have warned that AGI’s potential to become more intelligent than humans makes it a threat to humanity. At the same time, tech giants like Mark Zuckerberg and Larry Page choose a more optimistic outlook and are championing AGI’s swift development.
While excitement about AI drives conversations, the speed of its advancement toward AGI depends on investment by businesses and human ability to push the software programming to new realms.
AI Technology Systems’ Potential Impact on the Global Business Landscape
In June 2017, McKinsey Global Institute (MGI), an international business consulting firm, published a paper that discusses how narrow AI is being developed, and how it might impact the future global economy. It also offers a glimpse into how the foundation for AGI is being laid. The paper’s authors explore the possibilities associated with six AI technology systems:
- Smart robotics: Advanced sensor technology informs adaptable robots.
- Autonomous vehicles: Cars are capable of driving themselves.
- Computer vision: A computer has the ability to see and interpret images.
- Natural language: Human language can be processed and understood.
- Virtual agents: Automated chat bots provide users with quick, accurate information.
- Machine learning: Data analysis automates analytical model building.
These systems are examined in the context of how five global industry sectors – retail, electric utilities, manufacturing, healthcare, and education – are adopting them. Across these five sectors, narrow AI’s potential benefits are summarized and sorted into four categories of value creation:
- Project: Businesses can enhance demand forecasting, better inform research and development (R&D) efforts, and optimize the supply chain. For example, in the retail arena AI technology systems can help predict e-commerce purchases so sellers can keep the right amount of stock on hand, reducing the number of sales lost due to product unavailability by as much as 65%.
- Produce: AI systems integration can result in heightened operational productivity, reduced costs, and increased efficiency. In the world of manufacturing, a 3 to 5% production yield improvement is possible, along with as much as a 30% increase of material delivery time.
- Promote: Accurately target products and services to the right audience with an ideal message. In the healthcare industry, AI technology systems can help tailor treatment plans and boost patient engagement, and possibly bring a 5-9% expenditure reduction.
- Provide: A user’s experience with a business’s products and services can be enriched through the use of AI. For example, electric utilities can delight customers with $10-$30 in monthly bill savings, thanks to machine learning’s ability to automatically switch electricity supply deals.
MGI finds that investment in narrow AI is largely led by tech giants like Google, and urges companies across the global business landscape to prioritize AI adoption strategies or risk being left behind. They also point out that “machine learning and a subfield called deep learning are at the heart of many recent advances in artificial intelligence applications and have attracted a lot of attention and a significant share of the financing that has been pouring into the AI universe- almost 60 percent of all investment from outside the industry in 2016”. While the timeline remains a mystery, there is little question that the global business landscape is gearing up for an AGI sprint.
Persistent is the dilemma associated with advancing AI technology toward AGI: A company’s ability to keep competitive in a rapidly changing global economic landscape requires healthy investment and swift adoption, yet the potential power of the software introduces the risk of becoming a sorcerer’s apprentice. Therefore, businesses must actively anticipate problems and implement measured development strategies. Such a mandate is prescient and is also leading to some creative paths of exploration.
Maintaining Human Control
A commonly expressed worry revolves around AI’s ability to develop sentience and a malicious attitude toward its human creators. While fear certainly has a function in carefully planning for AI’s evolution, the Future of Life Institute identifies the concept of conscious and evil AI as a mythical worry. They say that alignment of human goals with those of the software is the key area of concern. Yes, AGI software may gain the ability to control humans by virtue of being smarter than us, but can we design it in a way that will prevent it from possessing such intent?
This is where the concept of digital ethics bears continuous discussion. Guido Wagner, an innovator at multinational enterprise software company SAP, asserts that “we do not know today what a self-organizing, super-intelligent AI will strive for in the end, but it is up to us to introduce from the beginning a basic set of needs suitable for a machine-based existence which neither impedes nor conflicts with the human hierarchy of needs”. Laying the ethical foundation within AI software introduces boundless complexity, but the work is necessary to keeping AI safe and beneficial to humanity.
Elon Musk, a prolific entrepreneur best known for helming headline-grabbing companies like Tesla and SpaceX, is not interested in taking chances with superintelligent software’s intentions. He aims to secure humans’ relevance and safety in an AGI powered future by directly linking our brains to AI technology systems. Championing his “if you can’t beat ‘em, join ‘em” approach, Mr. Musk has become a founding member of Neuralink, a venture whose mission is to develop brain-computer interfaces. Its long-term goals include:
- Creating neural lace implants that can be inserted into the human brain
- Using brain-computer interfaces (BCI) to integrate the mind with external software
- Enhancing humans’ ability to amass and process data via chip integration
- Providing new ways for the medical field to treat injuries
In addition to such exciting entrepreneurial efforts, partnerships between corporate and research community groups show promise. Earlier in 2018, Facebook AI Research (FAIR), powered by Mr. Zuckerberg’s enthusiasm for AI as a concept as well as his interest in its benefits for his business, announced the release of an open source AI framework called PyTorch 1.0. FAIR hopes to use this software to build an end-to-end deep learning system by unifying research and production capabilities. In support of this goal, the first developer conference for PyTorch 1.0 was held in October of this year.
Hope for the Best Outcome
AI is still a largely mysterious technology because its full potential will not be realized until it begins to surpass human intelligence. As with any concept or entity not widely understood, fear of superintelligent AI precipitates mythical worries and resistance by humans. Working to mitigate these concerns are some of the greatest minds in business, academia, and world government. Now the question is, how long will it be before the headlines celebrate the sudden awakening of true AGI?