iTechGuides is reader-supported. When you buy through links on our site, we may earn an affiliate commission. As an Amazon Associate I earn from qualifying purchases. Learn more
In 2018, artificial intelligence was drawing wider attention, spreading through business experimentation and deployment, and advancing in areas such as language technologies. But those signals did not mean AI was mature or broadly delivering value at scale: contemporary reporting also emphasized the difficulty of tracking a rapidly changing field and the barriers companies faced in scaling it.
What counted as an AI trend in 2018?
There was no single measure that captured the state of AI. Stanford’s 2018 AI Index Report organized evidence around the volume of activity, technical performance, relationships among trends, and selected areas approaching human performance. It also cautioned that the field was changing quickly enough to make it difficult to track, even for experts.
The report described AI’s rising prominence this way: “Artificial Intelligence has leapt to the forefront of global discourse, garnering increased attention from practitioners, industry leaders, policymakers, and the general public.” That attention was one part of the picture, not proof that a particular technology had become widely deployed.
Business adoption was spreading, but scaling remained difficult
McKinsey’s November 13, 2018 report, “AI adoption advances, but foundational barriers remain”, characterized business adoption as rapidly taking hold across the global business landscape. Its central qualification was that few companies had the foundational building blocks needed to create value at scale. Experimentation or deployment in a capability area should therefore not be read as evidence that most organizations had successfully integrated AI across their operations.
#1 Best Overall
The findings came from an online survey fielded February 6–16, 2018. McKinsey reported 2,135 participants, representing a range of regions, industries, company sizes, functions, and levels of tenure. These are respondent-based results from a defined survey period, not a census of every business.
Capabilities the survey covered
McKinsey’s survey asked about nine capability areas. Together they show how broad the term “AI adoption” could be in 2018:
Rank #2
- Natural-language text understanding
- Natural-language speech understanding
- Natural-language generation
- Virtual agents or conversational interfaces
- Computer vision
- Machine learning
- Physical robotics
- Autonomous vehicles
- Robotic process automation
These categories span different kinds of systems and use. A company reporting activity in one area does not establish that it was using all of them, or that the systems were producing value in routine production.
Recommended Free Tools
Language technology was advancing while remaining a challenge
Stanford’s December 2018 summary, “Artificial intelligence report finds advances in working with human language, global reach”, highlighted work with human language as both an area of progress and a continuing research challenge. This matters because language-related capabilities—such as understanding text or speech, generating language, and conversational interfaces—were also among the areas businesses were being asked about.
The two sources describe different kinds of evidence: one summarizes a research challenge, while the other reports business respondents’ activity across capability categories. Neither alone establishes how reliably language systems performed in every real-world setting.
AI education was reaching a wider international audience
Stanford’s December summary also reported a 16-fold increase in enrollment in introductory AI and machine-learning courses at Tsinghua University. This is a specific comparison for that university, as reported in 2018; it should not be generalized to all Chinese universities or to AI education worldwide. It does, however, offer a concrete example of growing educational participation beyond the usual technology-industry measures.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.How to evaluate a claim about an AI trend
Trend claims can sound comparable while measuring very different things. Before drawing conclusions, check:
- What is counted? A language system, a robot, machine-learning use, and robotic process automation are distinct capabilities, even when grouped under AI.
- What stage does the evidence describe? Research performance, a pilot, and routine production use are not interchangeable.
- Who and where were measured? Look for the geography and population, and, for a survey, its field dates and respondent base.
- Is the statement an observation or a forecast? A reported result describes evidence available at the time; it does not guarantee that a prediction came true.
That distinction reflects Stanford’s separation of activity measures from technical performance and related measures, and McKinsey’s use of a weighted online survey rather than a census. Read together, the 2018 sources show a field attracting attention, with business activity across a broad set of capabilities, alongside real limits in scaling and persistent technical challenges.
Quick Recap
Best Value
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
