Technology part of the blog

Earlier this year, Gartner published its new hype cycle for Artificial Intelligence [1]. Business adoption of AI is experiencing significant growth – thus, according to Gartner’s 2019 CIO Agenda survey, organizations that have deployed artificial intelligence (AI) grew from 4% to 14% between 2018 and 2019 [2]. However, until now, only two AI technologies – GPU Accelerators and Speech Recognition – have reached the plateau of productivity. The greater majority is situated in the first half, with a lot of excitement and experimentation, but also disillusionment and steep learning curves to be expected in the years to come.

To dive into the details of AI adoption, we mined a large-scale dataset of Web data for industries and use cases which are affected by each trend. The following overview shows the amount of discussion for the top 15 trends in our dataset in 2019 (excluding umbrella terms such as Machine Learning and Deep Learning):


Figure 1: Amount of discussion per technology

As differentiated in this chart, our analysis integrates three types of Web resources with different degrees of specialization on AI:

  • General news are business and economic news without an explicit specialization, such as and
  • Technology blogs are focused on technological and digital topics, both from the technology and the business perspective. Examples are and
  • Blogs with AI focus are specialized AI and machine learning resources for practitioners, such as and

As can be seen from the chart, the bulk of discussion goes on in technology blogs and blogs with AI focus. Topics with significant ethical, infrastructural and regulatory stakes, such as Autonomous Driving, Quantum Computing and Artificial General Intelligence, also attract considerable attention in general news.

It is worth noting that the considered trends belong to different conceptual classes and levels of AI. Some of them, for example Computer Vision and Natural Language Processing, are whole subdomains of AI that can be relevant to a multitude of use cases and industries. Others, like AutoML, simplify and scale the integration of AI in the enterprise context. Autonomous Vehicles, Chatbots, Virtual Assistants and Conversational UIs are situated at the application level. Finally, concepts such as Augmented Intelligence relate to the human-machine interface and are mainly meant to spur the user acceptance of new AI technologies.

Artificial Intelligence across industries

The following chart shows how tightly the considered trends are associated with various B2B and B2C industries:

No alt text provided for this image

Figure 2: AI trends by industry

Let’s look at some highlights:

  • Computer Vision is highly relevant in Automotive since it is an important component of Autonomous and Assisted Driving technology.
  • Autonomous Vehicles are tightly associated with Automotive – not really a surprise. In our analysis, Autonomous Vehicles are not necessarily cars – they also include Aerospace devices such as drones and space robots. Finally, the Construction industry is developing highly specialized autonomous utility vehicles. Operating on a well-delimited, well-understood terrain, the adoption of these narrow-focus vehicles faces lower hurdles in terms of infrastructure, ethics and regulation.
  • Quantum Computing is gaining attention as traditional computers approach their limits in terms of processing power. Quantum Computing is widely researched in the domain of Aerospace, where it is used to address the complex challenges of flight physics. Its high relevance to the Energy industry is due less to the possible applications, but to the fact that it is itself a solution to a major anticipated energy problem: by 2040, energy needs for classical computational computing will exceed capacity that can be delivered by the worldwide grid [3]. Alternatives such as the quantum computer are set out to prevent this bottleneck.
  • Robotic Process Automation (RPA) is performed by virtual “software” robots as opposed to physical robots which are used in manufacturing etc. RPA is essential in the Finance industry, where it enhances and automates routine tasks such as verification, credit scoring and fraud detection.
  • Conversational UIs are especially relevant to B2C industries such as Entertainment, Fashion and Retail. Their distribution in our data roughly correlates with Chatbots, Virtual Assistants and Speech Recognition, a core technology behind voice interfaces. Most Big Data companies are pushing their own conversational solutions such as Amazon Alexa, Google Assistant etc. According to Gartner, the conversational interface is one of the technologies with the biggest transformational impact in a short-term perspective.

Artificial Intelligence across use cases

The following chart illustrates the relevance of the considered trends to industry-independent use cases and business functions:

No alt text provided for this image

Figure 3: AI trends by use case

Let’s consider some of the data points in detail:

  • As most of us have experienced, AI experiments are particularly popular in the domain of Customer Service. Especially in B2C, most customer requests can be roughly classified into a finite number of buckets, thus providing a fertile ground for training data creation and automation. Most customer interactions use language, and all trends related to language processing – NLP, Chatbots, Speech Recognition, Virtual Assistants and Conversational UIs – are highly salient in this domain. To a smaller degree, the relevance of these technologies is also visible for Customer Acquisitionand Marketing in general.
  • Quantum Computing is strongly associated with Manufacturing. According to IBM, Manufacturing is one of the most promising early beneficiaries of this technology. Quantum Computing can significantly scale up and optimize tasks such as chemical discovery, simulations for product development and supply chain optimization [4].
  • Augmented Intelligence is strongly present across most of the use cases. It should be kept in mind that Augmented Intelligence is a rather abstract concept, mainly used to communicate and also reassure that in the foreseeable future, AI will be “enhancing” (rather than substituting) the intelligence of human beings. The concept clearly demonstrates and helps to build awareness of the limits of AI on the application level. On the opposite end, Artificial General Intelligence – the intelligence of machines that can learn to perform any intellectual tasks formerly performed by humans – shows a loose association to most of the considered use cases.

Summing up

A great majority of the considered trends aim to increase the efficiency of existing tasks by reproducing fundamental functions of the human brain such as language and vision. They are necessary building blocks of the overarching vision behind AI, as reflected in the concept of Artificial General Intelligence. For the present moment, the popularity of Augmented Intelligence shows that AI has passed a “reality check” and deployment is smoothened by more realistic expectations about the cooperation between humans and machines. Finally, Quantum Computing is an active area of research which could allow to harness the combined potential of ever-growing data quantities and sophisticated algorithms, thus allowing for a “quantum leap” towards the general application of AI.


[1] Gartner (2019). Hype Cycle for Artificial Intelligence, 2019. Retrieved from

[2] Gartner (2018). Gartner Survey of More Than 3,000 CIOs Reveals That Enterprises Are Entering the Third Era of IT. Retrieved from

[3] SIA / SRA (2015). Rebooting the IT Revolution: A Call to Action. Retrieved from

[4] IBM (2019). Exploring quantum computing use cases for manufacturing. Retrieved from


Talk 1: Word embeddings and NNs in NLP: A (Linguistic) Overview (Janna Lipenkova)


Talk 2: Grammatical gender in word embeddings (Kate McCurdy)


Please visit the meetup page for further information and announcements!