World Reimagined

What Is Artificial Intelligence and Who Are the Major Players in the Space?

Credit: Photo by Possessed Photography on Unsplash

The global health crisis accelerated the creation of many digital technologies, but its greatest beneficiary may have been Artificial Intelligence (AI), given the recent acceleration of investment in the space. In this piece, we discuss what AI is, how it is being used, and who the major players are. 

While the science of Artificial Intelligence first emerged in the 1940s with the UK “Enigma” codebreaker team led by Alan Turing, it has only been during this past decade that AI has made its way into everyday life. The term “Artificial Intelligence” was first coined in 1956 by Dartmouth Assistant Professor John McCarthy to refer to hardware or software exhibiting intelligent behavior. But it wasn’t until the 1980s that AI technology evolved into “expert systems” that more closely resemble today’s technologies. 

Types of AI Technology

Today, there are various forms of Artificial Intelligence. 

AI is heavily used by the military as well. For example, in August 2020, an artificial intelligence algorithm developed by Heron Systems beat an experienced human pilot in a simulated dogfight 5-0 broadcast via a Zoom meeting. The pilot had more than 2,000 hours in an actual F-16, but the AI enjoyed a flawless victory.

To understand why, we’ll discuss just what AI is and why it performed so well in this particular application.

Types of AI and the Bias Problem

While AI is increasingly a part of our everyday lives, it is a broad term, and not all AI is the same. General AI is the type of AI imagined in science fiction, the sentient robot, and doesn’t yet exist. That being said, there are some fairly advanced AI systems that have been interacting with the public via chats and in some cases, helplines for a few years now. Nothing that could pass a real Turing Test but it is able to go at least one step beyond simple “Yes/No” questions.

Narrow AI is typically what is meant by the term AI, and at its core, is a mathematical method for prediction. More broadly, it refers to creating intelligent machines that work and react like humans within specific contexts. An example of a Narrow AI system would be Apple’s Siri or Microsoft’s Cortana systems that work within a specific environment for a group of specific tasks. Machine Learning (ML) is a subset of AI that learns via data sets rather than being explicitly instructed. ML uses algorithms that learn iteratively by going through enormous amounts of data and finding insights without explicit instructions. The effectiveness of ML algorithms is directly proportional to the size of the data – the larger the data sample, the more accurate the algorithm, which makes data transmission and storage core to the world of AI.

Since a programmer's explicit instructions do not limit Machine Learning, it can look at the data in various ways, some of which a programmer may have never considered, thus reducing the impact of programmer biases. However, since Machine Learning relies on data to develop and test its hypothesis, biases in the data could impact the outcome. In an example of a car identification, if the data provided did not include any cars manufactured by South Korean companies, the algorithm the AI develops could be different than had it had access to such information.

Deep Learning is a subset of Machine Learning and uses artificial neural networks with three or more layers of nodes to solve problems. Neural networks can be considered a bridge between neuroscience and Artificial Intelligence in that neural networks are sets of algorithms inspired by the human brain. The better we understand how our minds work, the better we can mimic them. Deep learning networks use layers of neural network nodes to analyze, classify and predict data. Deep learning requires a large amount of computing power and data, making it typically more complex than is necessary for some applications. The most advanced field in deep learning is Reinforcement Learning or Neuro-dynamic Programming

You may have heard GPT-3 mentioned in discussions around AI. It is the third-generation language prediction model in the GPT-n series created by OpenAI. It uses deep learning to produce human-like text and is part of the trend in natural language processing (NLP). A review of its capabilities in The New York Times this past April found that it could write original prose with human-level fluency. The data fed into GPT-3 is essentially the entirety of the online written word, which one could argue is a comprehensive source. But anyone who has spent time on social media knows it also contains the worst of human biases and prejudices, which means so can GPT-3. This brings up one of the existential questions concerning AI. 

Do we want AI to accurately reflect human thinking or perhaps some idealized version instead? If the answer is idealized, then who defines the ideal?

In the AI versus man F-16 dog fight, the AI program benefitted from the enormous amount of flight data fed to it and from the nature of an F-16 pilot. The pilot was highly trained with extensive experience following the rules he’d been taught, rules that the AI could quickly identify and counter. This brings up another current challenge for AI. It is not well suited for leaps of intuition into unchartered territory, given that by its very nature, it relies on predictability. This is one of the challenges facing AI in financial markets.

Artificial Intelligence in the Market

In 2014, $4.25 billion was invested into AI startups globally, rising to $36 billion in 2020 and more than doubling in 2021 to $77.5 billion. Like many other things digital, the global health crisis accelerated investment in this market. Overall, the global AI market in 2021 is estimated to have been around $327.5 billion and is expected to grow to well over $1.5 trillion by 2030, representing a CAGR of nearly 20%. 

According to McKinsey’s The State of AI in 2021, the adoption of AI continues to expand, with 56% of all respondents reporting its use in at least one function, up from 50% in 2020. The report found that the most significant geographical region for increased adoption was in emerging economies, including China, the Middle East and North Africa. Those areas reported 57% adoption in 2021 versus 45% in 2020. AI is increasingly a source of competitive advantage as well, with 27% reporting an improvement of 5% or more in EBIT (earnings before interest and taxes) thanks to AI, up from 22% in 2020.

Top AI Companies to Invest In

As of December 2021, China’s Tencent (TCEHY) was the largest owner of active machine learning and artificial intelligence patent families worldwide, with 9,614 owned. Baidu (BIDUcame in second with just over 9,500 patents, followed by IBM (IBMat 7,343 and Microsoft (MSFT) at 5,821.

The largest AI unicorn startup in 2021 was the privately-held Chinese tech company ByteDance, with a valuation of approximately $140 billion. The company's AI and machine learning algorithms customize its users’ feeds on TikTok and Douyin.

Leaders in the development of AI technology include companies such as (AMZN) (AI), a subsidiary of Alphabet (GOOG) called DeepMind, Meta (FB), and Salesforce (CRM). 

AI technology also depends on the speed and robustness of the microchips on which it runs. The standard CPU (central processing unit) is not ideal for AI. The development of the GPU (graphics processing unit) in the 1990s was a significant improvement for AI. Today, chips are being designed specifically for the execution of AI algorithms. These chips all fall under the catchall term of AI PU (AI processing unit) but have names like NPU (neural processing unit) and TPU (tensor processing unit). The OG of chips, Qualcomm (QCOM), is considered one of the leading AI chip makers, while MediaTek (MDTKF) is one of the newer players in the space, and of course, there’s Samsung (SSNLF), which is pushing the development envelope with new 3 nanometer wide chip fabrication and NVIDIA (NVDA).

The bottom line is that AI is a game-changing technology that is being rapidly adopted across a wide range of industries and has the potential to improve significantly human decision-making, business processes and resource usage. AI will likely continue to enjoy outsized growth as we face a potentially slower global economy in the coming months or even years.

The views and opinions expressed herein are the views and opinions of the author and do not necessarily reflect those of Nasdaq, Inc.

Lenore Elle Hawkins

Lenore Elle Hawkins has, for over a decade, served as a founding partner of Calit Advisors, a boutique advisory firm specializing in mergers and acquisitions, private capital raise, and corporate finance with offices in Italy, Ireland, and California. She has previously served as the Chief Macro Strategist for Tematica Research, which primarily develops indices for Exchange Traded Products, co-authored the book Cocktail Investing, and is a regular guest on a variety of national and international investing-oriented television programs. She holds a degree in Mathematics and Economics from Claremont McKenna College, an MBA in Finance from the Anderson School at UCLA and is a member of the Mont Pelerin Society.

Read Lenore's Bio

Chris Versace

Christopher (Chris) Versace is the Chief Investment Officer and thematic strategist at Tematica Research. The proprietary thematic investing framework that he’s developed over the last decade leverages changing economic, demographic, psychographic and technology landscapes to identify pronounced, multi-year structural changes. This framework sits at the heart of Tematica’s investment themes and indices and builds on his more than 25 years analyzing industries, companies and their business models as well as financial statements. Versace is the co-author of “Cocktail Investing: Distilling Everyday Noise into Clear Investing Signals” and hosts the Thematic Signals podcast. He is also an Assistant Professor at NJCU School of Business, where he developed the NJCU New Jersey 50 Index.

Read Chris' Bio

Mark Abssy

Mark Abssy is Head of Indexing at Tematica Research focused on index and Exchange Traded Product development. He has product development and management experience with Indexes, ETFs, ETNs, Mutual Funds and listed derivatives. In his 25 year career he has held product development and management positions at NYSE|ICE, ISE ETF Ventures, Morgan Stanley, Fidelity Investments and Loomis Sayles. He received a BSBA from Northeastern University with a focus in Finance and International Business.

Read Mark's Bio