Any innovation that eventually becomes mainstream must reach a tipping point where adoption moves from slow and niche to a “get out of the way” tsunami. We’ve seen it with online banking and investment, digital communications, medical advances, and innumerable other facets of life. Now, seemingly overnight, artificial intelligence might be at that tipping point.
50% of firms worldwide used AI in some way in 2022, up from 20% in 2017. (McKinsey)
The global AI market is set to grow by 38% in 2023. (Tractica)
The global AI market is expected to grow to $1.81 trillion by 2030. (GrandViewResearch
Artificial intelligence applications have been around for decades, but they have primarily been adopted by companies already known as forward thinkers.
For example, a recent Economist article neatly summarized some of the established AI uses within the tech firm universe— “Amazon's AI manages its supply chains, instructs warehouse robots and predicts which job applicants will be good workers; Apple's powers its Siri digital assistant; Meta's serves up attention-grabbing social-media posts; and Microsoft's does everything from stripping out background noise in Teams, its videoconferencing service, to letting users create first drafts of PowerPoint presentations.”
These are companies you’d expect to adopt innovative technology, such as artificial intelligence. But now salt-of-the-earth companies, such as Walmart and John Deere, are now finding practical uses for AI that go far beyond robots and tech apps.
For example, retail firms are turning to AI to streamline their supply chains. When Hurricane Ian shut down Walmart’s Florida distribution hub, the company used an AI-powered simulation model to reroute its deliveries from other hubs and predict how demand for goods would change after the storm. Walmart was able to analyze the data and set up new deliveries within hours rather than days.
John Deere recently shipped its first fleet of fully self-driving machines, which use artificial intelligence to recognize obstacles and maneuver out of the way. John Deere tractors can also use AI to spot weeds among the crops and spray insecticide on just the weeds, as well as automatically adjust the distribution of seed to reduce waste.
Companies are finding that AI isn’t limited to a single specialized task—it can be trained to cover a variety of tasks, as well as intuit what else needs to be done. Businesses are using it to analyze sales calls and predict which customers are most likely to buy, as well as generate ad copy and decision-tree sales manuals. Some financial firms are using it to generate a first draft of quarterly reports. It does more than place numbers in a spreadsheet cell. It produces a report of what those numbers mean.
This content-generating feature is one that companies are particularly interested in. One of the most talked about new tools is ChatGPT. This app can scour the internet for data and generate everything from professional-sounding news stories to research papers within minutes. (College professors are already trying to determine how they are going to separate the student-written paper from the AI-written paper. Hint: AI paper will probably be better.) A chatbot tool by DoNotPay can negotiate bills, cancel subscriptions, and obtain discounts by “talking” to a firm’s live chat feature. Last year, Nike bought a firm that uses AI algorithms to create new sneaker designs. Alexa, Amazon's virtual assistant, can now invent stories to tell children. Nestlé is using AI-generated images to help sell its yogurts.
The uses for a content-generating AI tool seem endless. But relying on an anonymous tool to generate content that arrives without the company or the reader knowing where the data comes from is also the one most fraught with danger. Foundation models are trained on the internet and thus reflect what they find there—data, facts, and figures, but also biases, violence, prejudices, and misinformation. Amazon and Alexa made the news last year when one of its AI-generated child activities challenged a 10-year-old girl to touch a coin to the prongs of a half-inserted plug.
Being able to generate first drafts, logos, predictive reports and presentations will be a huge time-saver for lots of companies. But they might then want to think about hiring fact checkers to check on the facts contained therein. Or maybe there is a second level of AI that can fact check the first level of AI. At that point, we might be getting into anthropomorphism territory. Isaac Asimov would be proud.
So, hype or tipping point? We’d say tipping point. When we are seeing an intersection of tractors and artificial intelligence, AI is here to stay.