AI library

Research and usage

This curated guide and research summary is provided by researcher Abha Thakor and Non Stop News UK for collaboration projects, shared learning and professional development. It features just some of the items she has recently recommended in talks, training and social media. They have been selected to aid understanding and dispel misinformation, as well as to encourage the positive use and thinking for both Machine Learning and Artificial Intelligence.

Abha is an advocate for ongoing professional development for the professional institutes to which she belongs. This list is regularly updated to encourage greater understanding and as a CPD resource list. Most of the items featured on this page relate to or have potential for marketing, communications and PR.

Usage. This page and other forms of this guide are subject to copyright restrictions and intellectual property of researchers and journalists referenced. This document is also part of an academic research project being conducted by Non Stop News UK and its support work for AI for good campaigns. Permission to reproduce or store this list in part or whole is required prior to storage or republication. Please respect others’ work and research that is shared for CPD. We do use technology to monitor and address copyright and IP breaches, particularly by individuals reproducing sections of the website and resource sheets and republishing them as their own to sell products.

‪AI Read of the Week recommendations

Here is a small selection of Non Stop News UK’s recommended reads on #AI and #MachineLearning
The items featured on this page change regularly. They are aimed at helping beginners to AI better understand the subject and therefore influence and evaluate future development and their own usage of the tools which are already in our everyday use or soon will be.

You can keep up with regular recommendations on Non Stop News UK / Non Stop Business Support on social channels, shared reading lists and presentations.

A regular requested recommendation I receive is for practical introductions on AI. This book by Kirk Borne is one of my top 10 recommendations. It provides a good discussion and thinking points for managers and communications practitioners for their own professional development and to start assessing AI’s use and considerations in their workplaces or operational areas. Many points highlighted in the book can also be used to help dispel myths on the technology and its uses.‬ Getting practical about AI by Kirk Borne.

‘Machine Learning for Humans’ by Vishal Maini is a great, easy to follow introduction to ML. This webpage sets out each chapter

‘The Book of Why: the New Science of Cause and Effect’ questions predictive modelling approaches and systems which are currently common. For those new to AI, it provides an explanation of how it works. It is authored by Judea Pearl and Dana Mackenzie.

I like many of the items which also appear on this learning resource list produced by PureAI in November 2018.

AI awareness

Bot or Not app – this is a fun and easy to use application launched in May 2020. It won the Mozilla Creative Media Award in 2019. It uses a game approach to help raise awareness of what is a bot and what isn’t.

A quiz to help the public understand where AI is in use in their environment has been designed by We and AI (April 2020).

AI and emergency support

AI relating to supporting the coronavirus

Transcript for online meetings. These can help with the move to remote working during the emergency response. Tools include Chorus.ai which works with virtual conference platforms like Zoom.

Data mapping. There are a number of excellent AI data analysis tools. A number are developing in relation to the coronavirus pandemic. One has been created by ElementAI – a COVID19 calculator. Developers can join the open discussions on AI topics at OpenAI and A14E, Europe’s AI-on-demand platform.

More on the use of AI, machine learning and data in the COVID-19 section.

Social enterprise projects

Humans in the Loop, started in 2017 as a social enterprise. It describes itself as having a “vision to rebuild the lives of conflict-affected communities.” It garners AI to build what it calls the ‘workforces of the future’. When last reviewed for this resource in May 2020, it had a workforce of more than 250 conflict-affected people working on AI applications after receiving specialised ML annotation training.

OpenAI is a not for profit based in the USA. It states its mission is to focus on ‘positive human impact’. It was set up at the end of 2015. More in the AI resource section.

We and AI is a new not for profit organisation set up in February 2020 in the UK.

Women in AI groups

Just some of the women researchers, developers and writers working in artificial intelligence.

Video featuring Kathryn Hume explaining the basics of AI (Entrepreneur, August 2019). Kathryn is the VP of Product and Strategy at Integrate.ai

‘FAQ about AI by Hanan Salam’ (Medium, July 2019)

‘Become AI ready, jobs of tomorrow, new skills, career opportunities’ by Hanan Salam (Medium, July 2019)

‘Artificial intelligence – is it worth the risk?’ by Marisa Tschopp (Medium, July 2019)

Women in WI is a global community of women in Artificial Intelligence seeking to help close the gender gap in this area.

Women Leading in AI’s 10 principles of Responsible AI (PDF opens in a new window). The report sets out recommendations for Governments to regulate AI and drive its development. In the Executive Summary of its 2019 report, it states: “… we are still seeing many algorithms that discriminate against women and ethnic minorities.” It asks how AI can ‘help us build a better society’. The network was set up in 2018 to be a “global ‘think tank’ for women in AI with the aim to address the bias that can occur within algorithms due to a lack of diversity and inclusivity within the field of Artificial Intelligence.”

Deep Learning and #AI

Those who have worked with me or heard my talks on this will know this is my go-to recommendation for both those new to the area and wanting to start planning research and impact of using artificial intelligence. McKinsey’s feature on AI frontier applications and the value of deep learning (2018)

AI and game playing

If it has an algorithm then let’s build an AI version!

AI and the Rubik’s Cube

The traditional way to solve a Rubik’s Cube showing how algorithms play a part (Lifehacks, Pinterest)

There are now almost daily challenges appearing across the world on ho to solve this challenge using automation. Worth a read is Dan Maloney’s article on the Self-solving Rubik’s Cube (Sept 24, 2018).

University of California’s Irvine algorithm beats human record for solving Rubik’s Cube. 1.2seconds, 20 moves. (July 17, 2019). The University programmed DeepCubeA, a deep reinforcement learning algorithm. It said the algorithm “can find the solution in a fraction of a second, without any specific domain knowledge or in-game coaching from humans. This is no simple task considering that the cube has completion paths numbering in the billions but only one goal state – each of six sides displaying a solid color – which apparently can’t be found through random moves.”

Senior author Pierre Baldi, UC Irvine distinguished professor of computer science, said on the University website: “The solution to the Rubik’s Cube involves more symbolic, mathematical and abstract thinking, so a deep learning machine that can crack such a puzzle is getting closer to becoming a system that can think, reason, plan and make decisions.”

Google and its so described ‘super human’ game playing AI (Nova, Dec 6, 2018)

Internet of Things

An article often recommended to clients and workshop attendees by Non Stop Business Support to better understand the Internet of Things (#IOT). What is IOT (Wired)

AI in accessibility

Many of the items featured here have benefits for aiding access and reducing barriers for those with health accessibility needs. In this section, ideas and projects which have a potential benefit for use in communications will be highlighted.

Improvements to visual maps in Machine Learning to help reduce ambiguity for those with colour blindness (Google AI, August 2019). This article explores rainbow false maps used in algorithms in computer vision applications.

Research on how devices can track hand movements and recognise gestures in real time to improve VR’s receptiveness and accurately identify sign-language words. Google researchers write about how the “ability to perceive the shape and motion of hands” is vital in improving user experiences of technological domains and platform. They detail its usage to “enable overlay of digital content and information on top of the physical world in augmented reality”. (August 2019)

AI assistants

Nov 2018. How AI is being used to help digitise 5 million historical photographs for The New York Times.

Nov 2018. Alexa, Amazon’s AI personal assistant, is now available on Windows 10 computers. You can already use the Alexa app on Android, mobile ios and the range of Echo units produced by Amazon.

Earlier in 2018, the integration plans for Alex and Cortana were announced.

AI in business

An e-book has been published on managing AI and ML in business and ethical considerations. Produced by Znet and Tech Republic in April 2020. Sign-in to Tech Republic is needed for access and to view the additional research basis.

AI in communications

More in my specific guide on tools in communications, marketing and PR.

AI-generated clues to improve communication
Sample communications report, Cyrano AI
This company and its founder have had a positive reputation for their support of AIforGood.
There are some potential ethical issues in how this kind of technology may be misused and how it would pick up non standard neurological responses.

AI in the environment

This podcast by Eye On AI features Google Scientist John Platt (@johnplattml on Twitter) on how Machine Learning can be used to find economical zero-carbon energy sources and reduce the negative effects associated with climate change. (August 2019)

AI in health

How AI and robotics are changing healthcare (American Medical Association July 7, 2019)

‘Regulating healthcare AI and robots‘ by Nicolas Terry (SSRN, January 2019). It extols the “advances in healthcare artificial intelligence (AI) will seriously challenge the robustness and appropriateness of our current healthcare regulatory models.“

AI #algorithm that claims to beat the human brain (The Express)

Correcting the mistakes of robots using only brain and muscle signals. (BBC, September 2018)

‘Algorithms may outperform doctors, but they’re no healthcare panacea’ by Ivana Bartoletti (The Guardian, 26 July 2018)

AI in manufacturing

This post on Responsible AI focuses on the manufacturing sector, but many of its points are apt for other sectors too. It also gives an easy to follow explanation of biases. ‘Blindly following data does not create ethical AI systems’ by Kiran Krishnamurthy, The Manufacturer, May 2020

AI in journalism, media and research

Here are just a small selection of the products featured in chapters and magazine articles by Abha Thakor.

Changing access to transcription

AI transcription – examples include GoToMeetings. A useful review article on voice transcription by TechEmergence (September 2018)

GoToMeetings transcriptions feature

What topics are current?

Currents from Parse.ly – find out what people are currently topical and important trends with this free tool. It highlights how much focus is on a particular topic, based on a changing dataset of 600,000 articles (2019 figure). It analyses key words and traffic sources which are the main routes for those topics. It can be used to identify potential gaps in reporting subjects. This tool is also useful for marketing and PR, and campaign planning to include in analysis of topics in your area.

Saving your reading

Save articles for later with apps like Evernote and Pocket. If you want recommendations for similar reading, Pocket uses AI to offer suggestions to you.

AI in marketing

Also check apps listed in the journalism and PR sections.

Detailed and regular updated briefings on tools for productivity, marketing and PR are available to clients.
These are also available for personalisation for your organisation’s needs and online training is available.

AI is already in considerable use in marketing ranging from workflow and email management to generating online content and publishing systems from algorithms.

AI and social media

Inspecting algorithms in social media platforms – report by the Reset and the Ada Lovelace Institute, November 2020. It makes recommendations to policy makers for the regulator inspection (also known as algorithm audit) of social media platforms.

A good beginner’s article to marketing AI in social media from the Marketing AI Institute (21 April 2021). Option to provide your details for a more detailed document.

Voice output of text messages, emails, blurb and press releases

Abha Thakor writes

Do you consider how your emails sound? It is a must with AI and voice technology, as well as for your audiences who may be using accessibility and screen reader equipment.

As a journalist in both print and broadcast, I was taught early on to read out our news stories. As a trainer, I feature examples of how articles are frequently shared by our readers over a pint in the pub, are part of breakfast conversation or are read to relatives as part of catch-ups.

Remember press releases and the marketing text you hope will be picked up by newspaper and web articles can also end up being spoken by broadcasters and through accessibility tools. Non Stop News UK’s research into online training and digital accessibility projects have found even greater value for reading aloud your article or marketing.

Tools that allow you to ‘play’ written text your preferred AI-voice reader and the use of screen readers are becoming common use for train and bus commutes. They have been around for some time for accessibility. They are cheaper to access for both the end user and creator with developments in voice search and investment in virtual assistants from companies including Google, Microsoft, Apple and Amazon.

Considering how your marketing copy sounds should now be part of your main checklist before sending out releases or publishing materials. This also applies to email marketing blurb – a form of copy – may be received through an inbox, but may still be being read to a user.

Alexa updates mean marketeers must consider how emails sound not just how they look (MartechAdvisor, 2019)

AI in journalism, media and research

Changing access to transcription

AI transcription – examples include GoToMeetings. A useful review article on voice transcription by TechEmergence (September 2018)

GoToMeetings transcriptions feature

AI in PR

The CIPR launched the #AIinPR panel in 2018 to explore the impact of AI on public relations practice and foster dialogue in the public sphere about issues related to Artificial Intelligence. Abha Thakor has contributed to various projects including its collation of AI-related tools in PR and to the previous PRStacks book series.

An introduction to AI in PR – Skills guide for CIPR’s AIinPR panel member Jean Valin (PDF). This is the first in a series published in July 2019.

AI and productivity tools

Check Non Stop Business Support’s regular updated guides for clients featuring the latest productivity apps and processes fuelled by AI and ML. A few useful and quirky items are in this section to highlight the huge range available. Always check the origins of a product, how often it is updated and support, what is happening to your data and transferability, and security of the information you store to it. Secure passwords are essential.

Note taking – there are a range of AI solutions both for offline and online meetings. You can use apps like Otter.ai from live meetings and interviews. It is available on Google Play and iOS and has won a number of awards. You can create searchable files to improve their use and work collaboratively with other services it offers. In 2019, it partnered with Zoom for AI-live transcriptions.

SwiftKey keyboard from Microsoft (2019). Microsoft’s SwiftKey is designed to improve autocorrect and prediction accuracy as it learns your writing style. It will also suggest words, phrases and emojis that you’ve used and may be of use. You can also use it to type in five languages. It has customisable animal characters you can also send for that more creative and relaxed message.

AI and Music

One of my favourite projects in music is MuseNet. Developed by Open AI, it uses a deep neural network that can generate 4-minute musical compositions with 10 different instruments, and can combine a variety of music styles. It works on patterns of harmony and rhythm, and learning to predict the next token. Used this project in work to help adults with learning disabilities and Aspergers’ Syndrome. If you are a musician with an aptitude for technology, do support this project.

Music Transformer can help with the difficulty of writing longer musical score and maintaining the theme or coherence. It uses uses an attention-based neural network. In September 2019, it created a collaboration where you can generate piano music with this Magenta-based transformer which uses TensorFlow.

AI Resources

Open AI – this is a very resource based in the US. OpenAI builds free software for training, benchmarking, and experimenting with AI. It has education, platform and tools sections. It was launched in December 2015 and has continued to evolve. In its charter, it says it is focused on AI benefitting all of humanity.

AI talks and workshops

‪Hear the panel discussion from @TheAITalks on #ArtificialInteligence and Creativity (YouTube). Can #AI learn to be creative? An opportunity to look at #MachineLearning style professional development!

7 Dec 2018 Humans still needed in PR (PDF) according to an analysis of skills and tools in public relations published by the CIPR​. Jean Valin led this project for the #AIinPR panel to help understand the impact of technology, specifically artificial intelligence, on skills in the profession.

An #AI app that recommends pivotal tables and charts in Zoho’s new tools

AI in sport

An AI which can dribble a basketball!

IBM’s use of AI technology for professional tennis players (August 2018)

AI and video

CEO of Hootsuite Ryan Holmes critiques TikTak’s growth, its use of AI, and privacy considerations (August 2019)

TikTok’s advancement in social media predetermines video content delivered to user through algorithms ‘observing’ users’ preferences. It requires users to complete watching a video before it opens another and provides replay as an automatic. Interesting analysis on the Chinese video app, TikTok, by Shuhan Yu (UX Collective, March 2019)

Books featuring AI

‘Hello World: How to be Human in the Age of the Machine’ by Hannah Fry (Transworld Digital)

Explanation of Machine Learning and where it differs from AI. These terms are often incorrectly used interchangeably. This book also offers some interesting case studies. ‘Prediction Machines: The Simple Economics of Artificial Intelligence’ by Ajay Agrawal, Joshua Gans and Avi Goldfarb (Harvard Business Press, April 2018)

If you are wanting to understand the fundamentals well of how your organisation can start to use algorithms or influence your board’s thinking on AI, this is a good place to start. ‘Human + Machine: Reimagining Work in the Age of AI’ by Paul R Daugherty, H.James Wilson (Harvard Business Press, March 2018

An exploration through cognition in ‘AI and Future’ by Margaret Boden (Oxford University Press, 2016)

‘AI and Big Data on IBM Power Systems’

Voice search and shopping

Voice search allows users to speak search terms into a browser or app. This voice-enabled technology is now part and parcel of modern usage. Here are some of the ways users can access voice search and in some cases voice purchasing too. There are developments in both areas daily.

• Microsoft 10 operating system adds an AI-powered virtual assistant Cortana as part of its default installation
• Apple’s Siri has been around even longer and continues to evolve
• Google Assistant is making in-roads via its home hubs
• Google Chrome and more recently other browsers have a microphone symbol for users to speak their search term or find their product
• Alexa has an option for voice shopping / purchasing linked to users’ Amazon accounts
• YouTube’s voice search feature is now available on Android too
• Adobe’s advance into voice tech in 2018
• Google’s Voice Access app, Oct 2018

Trends and personalisation of voice searches are incorporated into how devices and apps ‘learn’ users’ interests and popular terms. ML and AI are used as part of this.

Voice shopping by High Street chain Argos using Google Home (September 2018)

Another useful article on how voice technology and its effect on business. (Computerworld)

Voice search and SEO

Why voice search will dominate SEO in 2019 (Forbes, December 2018)

Voice search on Google Chrome browser offered by ispeech.org

Voice search tips for YouTube on Android (Techboomers on YouTube, January 2019)

Mastering vertical search optimisation (Jim Wu, Search Engine Land, May 2018)

Write featured snippets for your website

‘Voice agents and interfaces will require many new marketing strategies’ by Barry Levine (April 2018). Good analysis of the NetElixr study and its suggestions for using featured snippets on websites which can be read by Google Home. Barry Schwartz writes about featured snippets and ‘Google’s new algorithm to better understand useful and timely information.’ (Search Engine Land, August 2019)

Building voice-controlled apps

Adobe’s move into this area as part of its stable for less technical users. Adobe purchased Saysspring Inc in 2018.

Copyright 2018-2020 Abha Thakor