Artificial intelligence is slowly – but surely – becoming a part of everyday life. Some say human taxi drivers will soon become a thing of the past, as will fishermen, most factory workers and call center employees.
In fact, in the coming years, AI will become a major part of a host of activities that have traditionally been carried out by people.
And investors have taken notice.
According to a CB Insights report, the second quarter of 2019 saw $7.4 billion invested in AI startups. And in the six years starting in 2013, investments totaling $66 billion have been made in 3,600 AI startups spread across more than 70 countries.
Q2’19 sees record funding to AI startups at $7.4 billion
Have these investments been profitable? In many instances, investors have made significant returns. In 2018, there were 18 unicorns – private firms valued at $1 billion or more – among the AI-focused companies.
This list includes ByteDance, the Chinese internet technology firm that operates several machine learning-enabled websites, the most famous of which is TikTok, a video-sharing platform. ByteDance has been wildly successful, and its apps have a collective monthly active user base of 1.5 billion.
The Chinese company is valued at $75 billion. The Financial Times says that it is planning an IPO in Hong Kong in the first quarter of next year, a report that the company denies.
How powerful an impact will AI have? Andrew Ng, co-founder of online learning platform Coursera and an adjunct professor of computer science at Stanford University, is probably right when he says that AI is the new electricity. It’s something that is going to transform the way we live and work.
How AI is going to change the world
The list of sectors that AI will influence is practically endless. Here’s a quick overview of several:
Medicine: AI’s ability to process vast amounts of data can be put to good use in the field of healthcare.
In 2018, an AI algorithm, DLAD (Deep Learning based Automatic Detection), developed at the Seoul National University Hospital and College of Medicine, was used to detect abnormal cell growth. The project involved classifying large numbers of images to identify potential cancers. How did DLAD perform when compared to real doctors? In 17 of 18 instances, AI was better at identifying malignant cells.
Consider another use case for AI in medicine. Every year, about 25,000 people in the US develop brain tumors. The traditional approach involves the same type of treatment in each case. However, “precision medicine” that uses deep learning techniques can help to provide accurate diagnosis and individualized treatment that could result in far better outcomes.
Cybersecurity: According to Juniper Research, a consultancy, cybercriminals stole 12 billion records in 2018. In 2023, this figure is expected to rise to 33 billion.
Traditional methods to counter cyber threats are inadequate. That’s because they rely on past data. AI could help to provide a far higher degree of protection.
Experts in the field of securing electronic data know that passwords are particularly vulnerable to cybercriminals. Artificial intelligence-based systems could provide a solution. For example, Apple’s “Face ID” incorporates the user’s facial features as well as AI software to prevent unauthorized individuals from accessing the target device.
Another menace that practically everyone who uses email faces is phishing. Using this technique, online criminals fraudulently obtain sensitive data like bank account numbers and passwords by pretending to represent a trustworthy organization. It’s estimated that one trillion phishing emails are sent annually.
AI and machine learning can be used in thwarting phishing attacks. These techniques are especially useful in identifying phishing sources and differentiating between fake websites and real ones.
Precision farming: Cainthus, an Irish agtech company, has used facial recognition technology in an unexpected manner. Instead of identifying humans, the Dublin-based firm’s AI program recognizes the faces of cows. Its smart camera system tracks each cow’s behavior and sends notifications in real-time to the dairy farmer’s phone.
How is AI relevant in dairy farming? Cainthus CEO Aidan Connolly explains that the company’s facial recognition tech helps to tell the farmer how much each cow is eating and how much she drinks. It also serves as an early warning system for ailments that animals may develop. The time the AI program takes to identify an individual cow by its features? Only six seconds.
Banking: AI will have a tremendous effect on the banking sector. This technology will allow customer services to be automated to a far greater degree than is possible with today’s computer systems. Another benefit for financial institutions is that machine learning will permit banks to offer their clients more personalized services.
Banks will gain other advantages, too. For example, AI could help banks to identify customers who are planning to take their business to another bank even before the client realizes that she is unhappy with the bank’s services.
But there’s a flip side to this. San Francisco-headquartered Wells Fargo, the fourth-largest bank in the US by total assets, predicts that the banking industry will see 200,000 job losses in the next ten years. While the reduction in the number of employees will help the financial institutions cut costs, it will also lead to a large number of layoffs.
Mining: Even sectors like mining will benefit from AI. Freeport-McMoran, a large copper producer, recently announced that it would boost production at one of its mines by 90,000 tons by using machine learning. Interestingly, the increase in output would involve “very little capital investment.”
The company’s AI project has been implemented at its aging Bagdad mine in Arizona. Developed in association with the consultancy, McKinsey, the model tracks data from sensors located in the mine and recommends ways to boost output.
The benefits are huge. If the company were to increase production by 90,000 tons using traditional methods, it would have to invest an additional $1.5 billion to $2 billion.
AI stocks to invest in
Here are five companies that are at the forefront of AI technology. In the coming years, they could see sharp increases in revenues and profits because of their early mover advantage.
Baidu: Chinese multinational specializing in internet services and AI
|Price on January 1, 2019||$162.25|
|Price on November 7, 2019||$121.87|
|Gain/(Loss) this year||(25%)|
At first glance, an investment in Baidu doesn’t look so attractive. The stock’s price has lost 25% this year. It reported a financial loss in the first quarter – its worst performance since it went public in 2005.
However, Baidu has a lot going for it.
The company’s version of Alexa, called DuerOS, has been installed on 400 million devices. It receives 3.6 billion queries every month, and the number of daily active users on its platform is 188 million.
Baidu is also a leader in facial recognition technology. A recent competition in this field demonstrated that it is ahead of rivals Alibaba, Huawei, and China’s top universities. Another indication of Baidu’s leadership in AI is that the firm’s Apollo project, an open-source autonomous vehicle technology platform, has made rapid strides.
At the current price, the share could be a bargain.
Alphabet: Google’s parent is a leader in AI
|Price on January 1, 2019||$1,045.85|
|Price on November 7, 2019||$1,308.86|
|Gain this year||25%|
Google’s advertising revenues were $33.9 billion in the third quarter, a 17% increase from the figure a year ago. This revenue stream provides Alphabet with the money for its moonshot projects – ventures that address important needs, but which may not provide profits in the short-term.
Alphabet has used its surplus cash for a wide range of AI initiatives. These include:
- Waymo – autonomous driving
- Wing – autonomous drone delivery
- Verily – healthcare
- Calico – machine learning for anti-aging
If its AI ventures pay off, the company’s share price could see tremendous upside even from its current levels.
Microsoft: The world’s biggest software company
|Price on January 1, 2019||$99.99|
|Price on November 7, 2019||$144.26|
|Gain this year||44%|
The company that gave the world the Microsoft Windows operating system and the Microsoft Office suite has invested a billion dollars in OpenAI, a venture that promises to “create new AI technologies.”
OpenAI proposes to build artificial general intelligence – a system that has the capability of becoming a world-expert in multiple fields. The company says its AI system could have the combined skills of Curie (Marie Curie did pioneering work in the area of radioactivity), Turing (Alan Turing was a brilliant mathematician and computer scientist), and Bach (one of the greatest composers of all time).
If OpenAI is successful, it could allow Microsoft to become the world’s dominant AI company.
Amazon: E-commerce and cloud computing giant
|Price on January 1, 2019||$1,539.13|
|Price on November 7, 2019||$1,788.20|
|Gain this year||16%|
Amazon has used AI to predict what buyers will purchase next on its e-commerce site. The individualized recommendations have played a big part in helping the company to capture almost half of the US e-commerce market.
A recent survey illustrates Amazon’s stranglehold over the e-commerce market. Feeadvisor, a company that helps other firms increase sales on e-commerce marketplaces, found that 74% of buyers head to the Amazon website when they want to make a purchase.
The company’s Amazon Web Services (AWS) is even more successful. It is the firm’s cash cow, and owes part of its success to AI. To put things in perspective, consider the profits that AWS makes:
Amazon’s sales and profits for the 12 months ended December 31, 2018
|Amazon North America e-commerce||AWS|
|Sales||$141 billion||$26 billion|
|Operating Income||$7.3 billion||$7.3 billion|
How is AI used in AWS? The financial software company, Intuit, uses SageMaker, Amazon’s cloud machine learning platform, to help it to create fraud-detection models. Grammarly, the writing app, uses it to predict what a user is writing. New applications for AWS’s machine learning tools are being found as the company adds more features.
|Price on January 1, 2019||$86.97|
|Price on November 7, 2019||$91.06|
|Gain this year||5%|
The company offers users cloud-based APIs (application programming interfaces) that allow voice and video to be built into apps. A new product, Autopilot, can be used to develop and deploy intelligent bots and Alexa apps.
Twilio’s products have a wide range of applications. The industries they can be used in include insurance, hospitality, retail, and e-commerce. If Twilio’s APIs gain traction, its shares could provide an excellent return. The stock’s valuation has already risen six-fold from the June 2016 IPO price of $15.
Risks of investing in AI
Are AI stocks a sure thing? No – like any other investment, they carry a certain degree of risk. The company that an investor buys stock in could face competition from a firm with a superior product. It may get its marketing strategy wrong. A new firm could run out of cash and investor confidence before it can gain market traction.
All these are risks that every firm faces. However, AI companies are also exposed to additional perils.
AI can be misused: The technology used to make “deepfakes” – false digital content that looks real – is rapidly becoming cheaper and easier to access. And it’s also getting harder to detect fakes.
The repercussions can be severe. False videos and audio recordings can be used to tarnish the image of politicians and destroy political careers. They could also influence stock prices and be used by criminals to blackmail the victims they target.
If an AI company’s products are used for illicit purposes, it could face regulatory action, and its stock price may suffer as a result.
Untested technology can backfire: Remember that there’s no guarantee that an AI initiative will succeed. Even the most respected company could back a losing project.
Back in 2013, IBM’s Watson for Oncology project was touted as an initiative that would cure cancer. Several years later, it emerged that IBM’s supercomputer gave unsafe recommendations and dangerous treatment advice.
Beware of “herd mentality”: There are a lot of genuine high-potential AI projects that attract investor funds. However, many companies are using the words “artificial intelligence” to describe what they do with the intention of misleading potential investors.
A study by MMC, a venture capital firm based in London, found that of the 2,830 self-described AI startups in Europe, only 1,580 were genuinely using artificial intelligence technology.
The bottom line
Investing in the right AI stocks could help to boost investor returns. But there are risks, as well. Typically, new firms are more likely to fail than older, well-established companies. However, a recently established AI venture that succeeds is likely to provide higher returns.
So, what should investors do? Those who are willing to stomach more risk can invest in newer firms. Alternately, traditional tech giants like Amazon and Microsoft could be the safer bets.