2021 artificial intelligence trends – AI market has registered a record ££ billions dollars equity funding investment during 2020.
Organizations are seeking AI best practices for big data gathering and storage in different formats, AI models development, performance monitoring and process automation.
An interesting study published by CBInsights reports Enterprises data management with particular focus on AI automation, AI OPS and artificial intelligence “no-code” solutions.
For example AIOPS segment, based on machine learning for IT and DevOps automation, registered a considerable increase in equity funding, despite a slight decrease in the number of deals, moving from 433 millions dollars in 2019 to 620 millions dollars in 2020.
2021 artificial intelligence trends: take off of no-code ai
2021 artificial intelligence trends – No-code AI platforms such as crafter.ai, make AI applications deployment accessible to business managers, with no technical or engineering skills. No code solutions allow to automate business processes and integrate AI inside organizations’ workflows, using machine learning with no coding skill required, thus reducing development times and costs.
“Covid-19 has accelerated the urgent need for every business to create no-code, low-code apps and workflows in hours or days, not weeks or months.”
Satya Nadella, Microsoft CEO
Between 2017 and 2020 big tech companies such as Google, Apple, Microsoft and Amazon introduces no code AI solutions for computer vision applications, machine learning in-browser experiments, image classification, no-code model development, business intelligence applications, automatic choice algorithms.
The streaming analytics market is another successfulmarket segment, that will reach 52 billions dollars value in 2027.
The number of real-time data sources is growing with the proliferation of IoT, and traditional data analysis and storage methods may result in missed opportunities for enterprises, that currently want instantaneous analysis and decision-making capabilities. This has raised interest in stream processing technologies, where data is viewed as a “stream of events” that is constantly generated. Stream processing powers AI apps that are responsive in real time and benefit itself from the use of machine learning.
un-structured data analysis
The number of alternative data sources is growing and includes IoT sensors, images, social media posts, and surveillance videos.
Around l’80% of these unstructured data, without a predefined format is not accessible by organizations. AI makes possible unstructured data analysis and patent activity related to unstructured data types is growing.
“Transformer-based “ models
In Natural Language Processing “Transformer-based” are being developed. This is a new of natural language comprehension introduced by Google already in 2017, named “Transformer”, that is able to understand the context of words and relations between sentences, by analysing a huge amount of text from the Internet.
Transformer-based models are revolutionizing sentiment analysis, translation, reading comprehension, gaming, and more.
In 2020, Facebook has released an open-sourcing multilanguage translation model, named M2M-100, that can translate 100 different languages, without relying on English-centric data, and based on a dataset of 7,5 billions sentences.
Data Governance and explainable AI
Strong governance is the foundation of AI explainability, where Explainable AI (XAI) is artificial intelligence (AI) in which the results of the solution can be understood by humans. Infact, Enterprises may not know what an algorithm is “seeing” in the input data or why it arrives at a certain conclusion. Focus on explainability is fundamental to comply with regulations like GDPR and CCPA. The importance of this theme is showed by the increase in media attention, that reached a peak in 2020.
You may also like: