What is automatic text summarization?

Automatic text summarization is the data science problem of creating a short, accurate, and fluent summary from a longer document. Summarization methods are greatly needed to consume the ever-growing amount of text data available online. In essence, summarization is meant to help us consume relevant information faster.

While summarization has been a field of study for decades, it has certainly grown in popularity in recent years. In 2017, Salesforce announced certain breakthroughs in the field of abstractive automatic summarization, and the use cases have proliferated across the enterprise.

Earlier in 2014, data scientist Juan Manuel Torres Moreno published a full book on the subject titled “Automatic Text summarization”, where he provided 6 reasons why we need automatic text summarization tools:

  • Summaries reduce reading time.
  • When researching documents, summaries make the selection process easier.
  • Automatic summarization improves the effectiveness of indexing.
  • Automatic summarization algorithms are less biased than human summarizers.
  • Personalized summaries are useful in question-answering systems as they provide personalized information.
  • Using automatic or semi-automatic summarization systems enables commercial abstract services to increase the number of texts they are able to process.

As described by Agolo, a Microsoft-backed summarization startup,  a document summarizer must generally overcome a set of challenges:

  • Determining which sentences are the most salient.
  • Making the summary cohesive and readable.
  • Minimizing the number of references to ideas and entities not mentioned in the summary. (i.e. coreference resolution).

Types of automatic summarization

Automatic summarization can be used in a variety of applications. Depending on the use case and type of documents, summarization systems can fall into different categories.

Abstractive vs. Extractive

When a human is given a corpus of text to summarize, they might rewrite the main points in their own words. This is called abstractive summarization and it requires high-level human skills like the ability to combine multiple perspectives into coherent natural language. As of 2018, the state of the art for abstractive summarization is not yet up to par, so many automatic summarization systems opt for a technique called extractive summarization.

Extractive summaries are excerpts taken directly from the input documents and presented in a readable way. The summary does not contain any rephrasing of the ideas presented in the original text. Extractive summarization methods employ AI-powered techniques to identify the most important sentences directly from the source.

Illustration of Salesforce’s model generating a multi-sentence summary from a news article. For each generated word, the model pays attention to specific words of the input and the previously generated output.

Single-document vs. Multi-document summarization

When summarizing a single document, the summarization system can rely on a cohesive piece of text with very little repetition of facts. However, the chance of redundancy increases with multi-document summarization systems. An ideal multi-document summarizer maximizes the important information included in the summary while minimizing repetition.

Indicative vs. informative

The taxonomy of summaries largely depend on the user’s end goal. For example, journalists or analysts looking to skim information as fast as possible would be interested in the high-level points of an article. So this use case requires an ‘indicative’ type of summary.

On the other hand, when the reader is looking to get more granular, summaries may require more detail. For example, a summary might need to allow topic filtering to let the reader further drill down the summary. This type of summary is considered to be ‘informative’.

Document length and type

The length of the input text heavily impacts the sort of approaches a summarization system can take. The largest summarization datasets, like Newsroom by Cornell University, have focused on news articles, which usually range 300 to 1,000 words. Extractive summarizers can be very effective when dealing with relatively short documents like news or blog articles. On the other hand, a 20-page report or a chapter of a book can only be summarized with the help of more advanced approaches like hierarchical clustering or discourse analysis.

In addition to length, documents may also fall into different genres. It is very different to summarize a news article to a financial earnings report or a technical white-paper. These are very different types of documents that may require entirely distinct summarization approaches.

Recommended articles

As a recap, here is a list of articles that cover the basics of automatic summarization. These articles were actually summarized by Frase’s summarization engine, which uses AI-powered extractive summarization.

New AI Breakthrough from Salesforce Research Boosts Productivity with Text Summarization (salesforce.com)

  • Salesforce Research is tackling this exact challenge and today we’re excited to announce two new breakthroughs in natural language processing towards the goal of automatically summarizing a long text and serving up coherent, digestible highlights that help you stay informed in a fraction of the time.
  • Text summarization is a very tough challenge, especially for longer texts such as news articles, and the work we are doing at Salesforce Research is pushing the state of the art.
  • I’m honored to work with Caiming Xiong and Richard Socher to introduce a more contextual word generation model and a new way of training summarization models with reinforcement learning (RL) .

Introduction to Automatic Text Summarization (blog.algorithmia.com) – Jan 05 2017

  • Without an abstract or summary, it can take minutes just to figure out what the heck someone is talking about in a paper or report.
  • Automatic text summarization is part of the field of natural language processing , which is how computers can analyze, understand, and derive meaning from human language.
  • By keeping things simple and general purpose, the automatic text summarization algorithm is able to function in a variety of situations that other implementations might struggle with, such as documents containing foreign languages or unique word associations that aren’t found in standard english language corpuses.

A Gentle Introduction to Text Summarization (machinelearningmastery.com) – Nov 28 2017

  • Automatic text summarization methods are greatly needed to address the ever-growing amount of text data available online to both better help discover relevant information and to consume relevant information faster.
  • After reading this post, you will know: Why text summarization is important, especially given the wealth of text available on the internet.
  • Automatic text summarization, or just text summarization, is the process of creating a short and coherent version of a longer document.
  • These deep learning approaches to automatic text summarization may be considered abstractive methods and generate a wholly new description by learning a language generation model specific to the source documents.

Taming Recurrent Neural Networks for Better Summarization (abigailsee.com)

  • Abstractive approaches use natural language generation techniques to write novel sentences.
  • In the past few years, the Recurrent neural network (RNN) – a type of neural network that can perform calculations on sequential data (e.g. sequences of words) – has become the standard approach for many Natural Language Processing tasks.
  • The decoder’s ability to freely generate words in any order – including words such as beat that do not appear in the source text – makes the sequence-to-sequence model a potentially powerful solution to abstractive summarization.
  • Explanation for Problem 2 : Repetition may be caused by the decoder’s over-reliance on the decoder input (i.e. previous summary word) , rather than storing longer-term information in the decoder state.

An Overview of Summarization – agolo (blog.agolo.com) – Nov 03 2016

  • A summarization system with what’s called a generic trigger will find the most important topics in a given input text and summarize it without further guidance.
  • A generic trigger for summarization is useful in cases where the user does not yet know the contents of the text to be summarized.
  • Agolo’s summarizer takes these factors into account at various points in the summarization process.

A Quick Introduction to Text Summarization in Machine Learning (towardsdatascience.com) – Sep 18 2018

  • Text summarization refers to the technique of shortening long pieces of text.
  • Machine learning models are usually trained to understand documents and distill the useful information before outputting the required summarized texts.
  • With such a big amount of data circulating in the digital space, there is need to develop machine learning algorithms that can automatically shorten longer texts and deliver accurate summaries that can fluently pass the intended messages.
  • However, the text summarization algorithms required to do abstraction are more difficult to develop; that’s why the use of extraction is still popular.
  • As research in this area continues, we can expect to see breakthroughs that will assist in fluently and accurately shortening long text documents.

How to Make a Text Summarizer – Intro to Deep Learning  (YouTube)

What is Natural Language Processing (NLP)?

Natural language processing (NLP) is an area of artificial intelligence that helps computers understand human natural language.  Often referred to as the engineering side of computational linguistics, NLP focuses on extracting meaning from unstructured data. NLP includes many different techniques for interpreting human language, ranging from statistical and machine learning methods to rules-based and algorithmic approaches.

NLP has removed many of the barriers between humans and computers, not only enabling them to understand and interact with each other, but also creating new opportunities to augment human intelligence and accomplish tasks that were impossible before. NLP enables real-world applications, including:

  • Automatic summarization: the process of creating a short and coherent version of a longer document. (machinelearningmastery.com)
  • Sentiment analysis: the process of determining the emotional tone behind a series of words, used to gain an understanding of the the attitudes, opinions and emotions expressed within an online mention. (brandwatch.com)
  • Named entity recognition (NER) locates and classifies the named entities present in the text. NER classifies entities into pre-defined categories such as the names of persons, organizations, locations, quantities, monetary values, specialized terms, product terminology and expressions of times. (blog.paralleldots.com)
  • Parts of speech tagging: the process of marking up a word in a text as corresponding to a particular part of speech (noun, verb, adjective, etc.), based on both its definition and its context—i.e., its relationship with adjacent and related words in a phrase , sentence , or paragraph. (en.wikipedia.org)

Subcategories of NLP include natural language generation (NLG) — a computer’s ability to create communication of its own — and natural language understanding (NLU) — the ability to understand slang, mispronunciations, misspellings, and other variants in language. (cio.com)

Business applications of NLP

NLP has been widely applied across many industries. Some examples include:

  • Enterprise question answering tools leverage NLP to enhance customer experience and improve administrative activities by allowing users to ask questions in everyday language about products, services or applications and receive immediate and accurate answers.  Many companies are successfully using customer support chatbots to streamline some of the work that would traditionally fall to representatives. Models built with NLP algorithms are the brains of these chatbots. They’re trained using text data from past conversations between your customer support agents and customers.
  • Optimizing Customer Satisfaction: manually sifting through product reviews and surveys can be prohibitively time consuming, but NLP can be used to build data models that generate insights to help optimize customer satisfaction. Sentiment analysis is used to classify text into positive, neutral, or negative categories.
  • Classifying medical records: Researchers at MIT in 2012 were able to attain a 75 percent accuracy rate for deciphering the semantic meaning of specific clinical terms contained in free-text clinical notes, using a statistical probability model to assess surrounding terms and put ambiguous terms into context. (healthitanalytics.com)
  • Ad placement: NLP can help in intelligent advertisement targeting and placement. Media buying is usually the largest channel in an organization’s advertising budget. So, it is important to ensure that the advertisement reaches the right eyeballs. Browsing behaviors, social media and emails contains a lot of information imbedded that can give a lot of insights about consumer preferences. NLP can be used here to match keywords of interest in the texts to target the right consumers. It can also be used for disambiguation or identification of the sense in which a word is used in a sentence.
  • Reputation monitoring: with increased competition in diverse market, monitoring reputation is essential to avoid getting drifted away in the tide. With a plethora of information sources abut companies like social media, blog posts, and reports, it becomes imperative to utilize these sources to get more insights about the reputation and reviews of the company. NLP is the best way to understand and extract insights from these sources.
  • Helping hiring manager: NLP can help hiring managers to filter resumes. Automated candidate sourcing tools can scan CVs of applicants to extract required information and pinpoint the candidates who are right fit for the job. This will save a lot of time and give a more efficient solution.
  • Market intelligence: NLP can help to monitor and track market intelligence. Since markets are influenced by information exchange, using event extraction, NLP can recognize what happens to an entity.

These are only a few examples of the many ways NLP can be used to unlock valuable information from text data. To learn how to get started implementing these techniques in your business, below are some resources that might help you dive deeper into the world of NLP:

Relevant Wikipedia articles

  • Natural language generation (NLG) is the natural language processing task of generating natural language from a machine representation system such as a knowledge base or a logical form. Psycholinguists prefer the term language production when such formal representations are interpreted as models for mental representations.
  • Computational linguistics is an interdisciplinary field concerned with the statistical or rule-based modeling of natural language from a computational perspective, as well as the study of appropriate computational approaches to linguistic questions.
  • Question answering (QA) is a computer science discipline within the fields of information retrieval and natural language processing (NLP), which is concerned with building systems that automatically answer questions posed by humans in a natural language.
  • Textual entailment (TE) in natural language processing is a directional relation between text fragments. The relation holds whenever the truth of one text fragment follows from another text.

  • The bag-of-words model is a simplifying representation used in natural language processing and information retrieval (IR). Also known as the vector space model. In this model, a text (such as a sentence or a document) is represented as the bagof its  words, disregarding grammar and even word order but keeping multiplicity.

Open Source NLP Libraries

These libraries provide the algorithmic building blocks of NLP in real-world applications.

  • Apache OpenNLP: a machine learning toolkit that provides tokenizers, sentence segmentation, part-of-speech tagging, named entity extraction, chunking, parsing, and more.
  • Natural Language Toolkit (NLTK): a Python library that provides modules for processing text, classifying, tokenizing, stemming, tagging, parsing, and more.
  • Standford NLP: a suite of NLP tools that provide part-of-speech tagging, the named entity recognizer, coreference resolution system, sentiment analysis, and more.
  • MALLET: a Java package that provides Latent Dirichlet Allocation, document classification, clustering, topic modeling, information extraction, and more.

NLP Courses

In case you are looking to get your feet wet with NLP, these are 2 popular online courses for beginners:

  • Stanford Natural Language Processing on Coursera: “This course covers a broad range of topics in natural language processing, including word and sentence tokenization, text classification and sentiment analysis, spelling correction, information extraction, parsing, meaning extraction, and question answering, We will also introduce the underlying theory from probability, statistics, and machine learning that are crucial for the field, and cover fundamental algorithms like n-gram language modeling, naive bayes and maxent classifiers, sequence models like Hidden Markov Models, probabilistic dependency and constituent parsing, and vector-space models of meaning.”
  • Udemy’s Introduction to Natural Language Processing: ” This course introduces Natural Language Processing through the use of python and the Natural Language Tool Kit. Through a practical approach, you’ll get hands on experience working with and analyzing textAs a student of this course, you’ll get updates for free, which include lecture revisions, new code examples, and new data projects.”

NLP YouTube videos

  • Natural Language Processing with Deep Learning – Stanford University
  • Introduction to Natural Language Processing – Cambridge Data Science Bootcamp
  • Natural Language Generation at Google Research

What Are The Core Elements Of An Effective Landing Page?

A strong landing page is a key element of any inbound marketer’s toolkit. But what exactly is a landing page, and when would you want to create one?

In answering this question, we used Frase’s AI-driven research tool to summarize the most relevant pieces of content on the internet about the role of landing pages in a B2B or B2C marketing strategy.

Our sources agree on a number of items:

  • A landing page is intended to be the first point of entry for a visitor to your business. It is a standalone web page, generally created with a marketing or advertising campaign in mind.
  • The primary purpose of a landing page is to get your visitor to do something, whether sign up for a newsletter or product demo, or click through to another targeted part of your website.
  • The hallmark of a strong landing page is its ability to convert, in other words to entice the visitor to undertake the action you are seeking.
  • An effective landing page should be simple in design, easy for the visitor to understand, and not overloaded with choices.
  • A good landing page is light on links and strong on an over-riding Call To Action (CTA).
  • It’s preferable to be directing visitors to dedicated landing pages rather than your home page, particularly for paid search campaigns.
  • With any marketing campaign, it makes sense to be testing different versions of your landing page to assess its effectiveness.
  • A major reason businesses don’t set up custom landing pages is because marketing departments are not sure how to create them easily.

What’s clear is that landing pages play a major role in effective inbound lead generation. What’s also apparent is that landing page best practices are often overlooked. Here is a collection of the main sources we reviewed in drawing our conclusions.

Why Landing Pages Are an Indispensable Part of Marketing (blog.hubspot.com) – Jul 28 2017

  • When she reaches the bottom of the article, she notices a call-to-action (CTA), which is essentially an ad for one of your offers — a free painting consultation to help her decide which color scheme would work best with the size and type of nursery she’s working with.
  • The landing page provides some additional information and details about what she will get out of the free consultation, convincing her it’s worth providing her contact information on the landing page’s conversion form in order to take advantage of the offer.
  • Landing pages not only enable you to generate new leads; they also allow you to track reconversions of existing leads, which you can then use to identify which prospects are more engaged with your business.
  • By tracking and analyzing the metrics associated with your landing pages , you can collect a lot of insight into your marketing performance, such as how your various marketing offers compare, how visitors and leads are converting on your landing pages over time, and more.

What is a Landing Page? (unbounce.com)

  • To fully understand the difference between a landing page, and the other pages on your website, such as your homepage, it’s important to consider the differences between organic search traffic and paid search traffic.
  • You could choose to send them to your website’s homepage, or to the preferred option – a standalone landing page created specifically for that ad campaign.
  • Lead Generation landing pages (sometimes referred to as lead gen or lead capture pages) use a web form as the Call To Action, for the purpose of collecting lead data such as names and email addresses.

What is a PPC Landing Page? (instapage.com) – Apr 20 2015

  • Instead of relying on organic traffic to your website, you buy traffic for your page by paying a publisher, like Google, to show your ad when your visitor does a search for your relevant keyword(s).
  • The entire purpose of your ad is to get a user to a PPC landing page so you can get a visitor to take an action, click on the CTA button, and become a customer.
  • However, the page has a number of navigation links, and even though they are at the bottom of the page, they do still make the conversion ratio 5:1, which they should test to see what the effect is on their conversion rate.

What Is a Landing Page and Why Should You Care (blog.hubspot.com) – Jul 28 2017

  • A good landing page will be targeted to a particular stream of traffic – say from an email campaign advertising a particular whitepaper – and, because it is targeted, and because it has an interesting offer behind a lead capture form, you will convert a higher percentage of your website visitors into leads with which you can follow up.
  • When you know a stream of targeted traffic will be coming to your website, you can increase the likelihood of converting that traffic into leads by using a targeted landing page .
  • Make sure you have a landing page creation tool that allows you to create and test many different landing pages to see what works best for your business.

Overcoming Your Fear of Local Landing Pages (moz.com)

  • Obviously, some things will be the same on various pages (like a phone number, a license number, mentions of products, slogans, etc,), but if you are going to the trouble to uniquely market 3 different landing pages for 3 different towns, definitely do make the effort to write something unique for each.
  • You can (and should) do so within the text of other pages of the site, yes, but I advocate also having a high level menu labeled something like ‘Cities We Serve’ that lists out all these cities, provided there aren’t too many of them to make the list crazy long.

Exactly What Is The Difference Between A Copywriter And Content Writer?

Ever scratched your head over the distinction between a copywriter and a content writer? How is one truly different from the other?

At Frase we’re not ones to sit on our hands. We scoured the web using our cutting-edge research technology to identify the online sources most closely answering our question.

Having found the 12 most relevant sources offered up by Frase’s Research Assistant, we reviewed the custom summaries generated by the system and whittled them down to the five most relevant articles from such publishers as Skyword and Copyblogger.

The insights we unearthed are illuminating.

One critical distinction between a copywriter and a content writer our sources are agreed upon involves the intention of the writing. Copywriting is seen largely as marketing copy, intended to persuade. Slogans, taglines, “About us” copy, even headlines designed to entice clicks, are examples of the copywriter’s craft.

By contrast, content writing is viewed as informing and educating. Blog postswhite paperslonger-form product descriptions all fall into this category.

But it isn’t always cut and dried. Our expert sources acknowledge that good copywritingoften includes informational content, and that informational content may incorporate elements of copy. A white paper needs a catchy headline after all or nobody is going to read it. Content without copywriting is a waste of good content, as Copyblogger says.

But don’t take our word for it. Here are Frase’s AI-generated summaries of our five sources, with links out to each of the original articles.

What’s the Difference Between Content Marketing and Copywriting? – Copyblogger (copyblogger.com)

  • If you’re writing great articles that people would love to read, but you’re not getting the traffic you want, the problem may be ineffective copywriting: Your headlines might be too dull.
  • Just like a product has to have a benefit to the buyer, your content has to be inherently rewarding to readers or they won’t come back to your website .
  • It’s tricky to show readers your blog is a cool place to hang out when you don’t have lots of readers yet, but we have a few tips for you .

Copy vs. Content Writing: What’s the Difference? (skyword.com)

Topics: landing pagecontent marketingonline mediasocial mediadirect mailblog postend usercopywriting

  • Because there’s general confusion on the differences between copy and content, you can decide how to define each category on your own terms, and expect that they may be used interchangeably depending on who you work with.
  • I consider most of the work I do to be content because it’s longer form, such as this blog post, and I’m almost always giving advice.
  • When you start working on a project, it’s more important to ask yourself why the project requires the written word, rather than whether it’s copy or content.
  • The recruiter I met with may always use different words to describe the services I offer, and that’s OK. As a content creator, it’s most important to figure out what you’re good at, and show samples to anyone who’s curious.

Content Writer vs Copywriter: What’s the Difference? (koozai.com)

Topics: search engineblog postsdisplay advertisingmarketing materialwritten wordMetaskill setscomment section

  • They both act as a cornerstone for the other; think of them as two different sides of the same coin, as although there are similarities in the two skill sets, there are also some clear differences, too.
  • When a Content Writer creates a piece of content they are most likely considering the use of keywords, meta, and how shares and links to the piece will amplify the content.

Content Writer vs. Copywriter: What’s the Difference? (marketingprofs.com)

  • I may not be in the market for a car or a pizza or new type of toothbrush, but effective copy can persuade me to consider one.
  • If the decision is to provide a core dump that attempts to be all things to all people or write copy that tries to sell something to readers or viewers or listeners who don’t want to buy, it’s time for client management to undergo remedial training.

Copywriting vs. Content Writing: What’s the Difference Between the Two? – Quietly Blog (blog.quiet.ly)

  • The ultimate objective of copywriting is to sell an idea whereas the ultimate objective of content writing aims to create valuable content to help the audience understand your brand and generate interest.
  • Bottom line: A copywriter is a professional who writes marketing copy; a content writer can be anyone producing content.
  • What you really need to know is that brands of all kinds need copywriting and content writing to stay fresh, so there’s plenty of opportunities for writers out there to try their hands at both.

Can AI be creative?

Generally speaking, Artificial Intelligence (AI) is disrupting industries by either replacing humans with full automation or augmenting their capabilities. In other posts, we’ve discussed how AI will dramatically impact content creation in both ways. While AI might not be consciously creative yet, AI tools can certainly help humans find inspiration and make more creative decisions.

In theory, AI is meant to help content creators automate repetitive, time consuming tasks that don’t require creativity. However, there have been plenty of experiments that explore the intersection of AI and creativity. Thinking of AI developing creative skills makes you think of the idea of Artificial General Intelligence:

Artificial general intelligence (AGI) is the intelligence of a machine that could successfully perform any intellectual task that a human being can. It is a primary goal of some artificial intelligence research and a common topic in science fiction and future studies . Artificial general intelligence is also referred to as ” strong AI” , ” full AI” or as the ability of a machine to perform ” general intelligent action” . Academic sources reserve ” strong AI” to refer to machines capable of experiencing consciousness . (wikipedia.org)

To get you thinking about the creative AI , we’ve summarized a collection of popular stories that discuss the subject:

Robot Art Raises Questions about Human Creativity (technologyreview.com)

Topics: art world, Piet Mondrian, color field, David Hockney, Max Ernst, Rembrandt, earlier work, Leonardo da Vinci

  • AARON has been able to make pictures autonomously for decades; even in the late 1980s Cohen was able to joke that he was the only artist who would ever be able to have a posthumous exhibition of new works created entirely after his own death.
  • The unresolved questions about machine art are, first, what its potential is and, second, whether—irrespective of the quality of the work produced—it can truly be described as “creative” or “imaginative.” These are problems, profound and fascinating, that take us deep into the mysteries of human art-making.
  • The neural net is provided with an image made up of a blizzard of blotches and spots, and is asked to tweak the image to bring out any faint resemblance it detects in the noise to objects that the software has been trained to recognize.
  • That may or may not happen, and “if it doesn’t, it means that machines will never be creative in the same sense that humans are creative.” The processes of such an artist involve an interplay between social, emotional, historical, psychological, and physiological factors that are dauntingly difficult to analyze, let alone replicate.
  • One day, Cohen suggests, a machine might develop an equivalent sensibility, but even if that never comes to pass, “it doesn’t mean that machines have no part to play with respect to creativity.” As his own experience shows, artificial intelligence offers the artist something beyond an assistant or pupil: a new creative collaborator.

Can robots be creative? (nautil.us)

Topics: UC Berkeley, Carnegie Mellon University, Carnegie mellon, Ray Kurzweil, University of California, new media, Marvin Minsky, human behavior

  • Since we were working in robotics and had robots in the lab we thought, “Why don’t we connect a robot to this Web and let people control it from anywhere in the world?” We got super excited about the idea of having a robot do something that was ironic.
  • Then I realized it was a very deep question because many hoaxes had been done on the web, and it’s not too hard to imagine the whole thing could have been faked.
  • He wasn’t a computer scientist but a philosopher saying, “No, you’re missing that a fundamental aspect of intelligence is experience and that requires embodiment.” He knew that to understand the world you needed to be inside the world, you needed to experience its behaviors and responses to you.
  • Right, it’s one thing to have a robot caretaker for an old codger like me, but when you turn that around and say the robot is taking care of my 3-year-old, it seems like there is something really wrong.
  • Dozens of times when I thought I had a new idea I would go on the Internet and probe around only to find that somebody else had exactly the same idea and did it.

For AI to Get Creative, It Must Learn the Rules–Then How to Break ‘Em (scientificamerican.com)

Topics: machine learning, neural network, University of california berkeley, artificial neural network, artificial intelligence, creative problem solving, human brain, University of California

  • Proponents say, however, the real beauty of training AI to be creative does not lie in the end product—but rather in the technology’s potential to expand on its own machine-learning education, and to solve problems by thinking outside the box far faster and better than humans can.
  • “We as humans still have to recognize the creativity or novelty.” His goal is to build artistic networks that can independently select and even tweet out their own best work based on the given theme.
  • GANs’ content-generating capabilities are a good start when it comes to developing AI that can solve real-world problems, says Ian Goodfellow, a staff research scientist at Google and lead author of the 2014 paper that first described the concept of GANs.
  • For example, researchers at the University of California, Berkeley, are using their “cycle-consistent adversarial network” ( CycleGAN ) to transform a video of horses into one of zebra s. The AI detects the basic shape of a horse in the first video and can play with the aesthetic on top of that image, immediately and seamlessly swapping a shiny brown coat of hair for one with black-and-white stripes while the image is moving.

There Will Always Be Limits to How Creative a Computer Can Be (hbr.org)

Topics: neural network, artificial intelligence, supercomputer, deep learning, computer simulations, computer program, computer system, human computer

  • Inevitably, “ artificial intelligence will soon be able to do the administrative tasks that consume much of managers’ time faster, better, and at a lower cost.” But, when it comes to more complex and creative tasks such as innovation, the question still remains whether AI can do the job better than humans.
  • Given these limitations, my co-author and I have proposed a solution: a new human-computer interface (HCI) that allows humans and computers to work together to counter each other’s weaknesses.
  • We shouldn’t concern ourselves, therefore, about whether computers will overtake humans; instead, we should focus on designing a system that allows humans and computers to easily collaborate together so each partner can build upon the other’s strengths and counter the other’s weaknesses.
  • After losing the first three games to Alphago, and witnessing the computer system make novel moves that players hadn’t seen before, Lee Sedol made a highly creative move of his own and ended up winning the game.

Job Automation: is AI capable of human creativity? (bigthink.com)

Topics: physical world, natural world, human being, human condition, technological progress, heat exchanger, creative people, interesting phenomenon

  • Andrew McAfee: Just about every time that I get involved in a discussion or a conversation about technological progress and how it can take away jobs from people and how it can automate away things that people used to do, one of the first things that people talk about this irreplaceable human skill is creativity, is coming up with some kind of eureka.
  • There’s a rapidly growing field called generative design and what that means is if you feed into a modern piece of technology the specifications that you want this building to be able to handle or this heat exchanger or the frame of a car or some kind of part out there in the physical world that has to meet some performance specifications or fit inside some performance envelope we’ve got software that will generate a part that will do that admirably.
  • And there’s an interesting phenomenon going on there: when people know in advance that they’re going to be listening to computer generated music they very often dismiss it as shallow or trivial or obviously not coming from a human composer’s mind and heart.
  • When listeners don’t know in advance that they’re listening to computer generated music they very often find it as evocative, as beautiful, as moving as anything a human being would come up with.

Artificial intelligence and creativity: If robots can make art, what’s left for us? (abc.net.au)

Topics: Artificial intelligence, Monash University, artistic practice, Australia, computer science, human condition, computer code, Eden

  • ” Art is one of the last domains in AI where there is an optimistic view on how humans and machines can work together,” says Dave King, founder of Move 37, a creative AI company.
  • ” If you have an algorithm that is working for you in the way you want it to it can source and discover lots and lots of different things.”
  • ” We’re naturally scared of anything where we take away something from people, particularly something as precious as being creative and art, which we associate with being the most fundamental human [trait] — that thing that differentiates us from every other species on the planet,” he said.
  • ” One of the oldest jobs on the planet, being a carpenter or an artisan, we will value most [in the future] because we will like to see an object carved or touched by the human hand, not a machine.”

Can robots truly be creative and use their imagination? (theguardian.com)

Topics: research associate, neuroscientist, University of California, professor emeritus, associate professor, computational creativity, machine learning, Artificial intelligence research

  • I’ve been working on writing novels computationally for well over 10 years now and I’m still trying it, although I believe that within the next two to three years I will have broken its back and will produce 100,000-word novels in half an hour or so, novels that I think most people would consider to be creative.
  • We are used to machines being used as tools that do not have a high level of cognitive ability, so it’s difficult for people to think of them as being able to exhibit truly creative behaviour.
  • Another problem is that it is difficult to automate the combination of ideas from many different sources that forms the source of much of human creativity: you might find inspiration from an interview with a neuroscientist in designing a new office layout.

Google wants artificial intelligence to be creative (businessinsider.com)

Topics: deep learning, creative thinking, practical applications, open source, artificial intelligence, use case, driverless cars, Google

  • Called Magenta, the group will use its AI system Tensorflow to see if AI can be trained to create its own art, music, and video.
  • Douglas Eck, a Google researcher who introduced Magenta at Moogfest, said one use case of the program is to have AI that can create music to counteract stress, Quartz reported.
  • If AI can master creative thinking, little by little, it inches closer to thinking like humans — or as we saw with Alphago, thinking in ways humans could potentially never dream of.

Can AI Be Creative? (barkleyus.com)

Topics: creative director, Google, creative professionals, creative work, clickbait, creative direction, customization, panel discussion

  • The creative direction it offered was to “convey ‘wild’ with a song in an urban tone, leaving an image of refreshment with a feeling of liberation.” Admittedly, that direction leaves a lot of room for interpretation, but it’s no less coherent than a lot of creative platforms I’ve seen.
  • It doesn’t matter one bit if Creative AI can’t produce something truly original or doesn’t have a soul or can’t make decisions about relevance.
  • I’m not placing bets just yet on when the robot creative revolution will happen, but I’m fairly confident it will happen before the end of my career; and possibly, of course, bring about the end of my career.

Who are the “godfathers of AI”?

If you follow latest news about Artificial Intelligence, you will frequently read quotes from the so-called ” Godfathers of AI”. In this post I collected and summarized the 10 most shared stories that made direct references to these godfathers.

The most frequently mentioned AI leaders were:

The most frequently mentioned topics were:

  • Human brain
  • University of Toronto
  • Deep learning
  • Ethics
  • Google
  • Neural network
  • Stanford University
  • Carnegie Mellon University
  • Large dataset
  • Facebook


The ” Father of Artificial Intelligence” Says Singularity is 30 years away (futurism.com) – February, 2018

Topics: artificial life, Ray Kurzweil, artificial intelligence, Kurzweil, global warming, world government, chemical evolution, human brains

  • It is that long-awaited point in time — likely, a point in our very near future — when advances in artificial intelligence lead to the creation of a machine (a technological form of life?)
  • Schmidhuber says it “is just 30 years away, if the trend doesn’t break, and there will be rather cheap computational devices that have as many connections as your brain but are much faster,” he said.
  • Of course, the development that he is referring to is the development of these artificial superintelligences, a thing that Schmidhuber says “is something that transcends humankind and life itself.”

AI Spotlight: Meet Professor Geoffrey Hinton, Godfather of AI (techstartups.com) –  December, 2017

Topics: Artificial Intelligence, machine learning, neural networks, deep learning, experimental psychology, data compression, pioneering work, graduate students

  • He continued his study at the University of Edinburgh where he was awarded a PhD in artificial intelligence in 1977 for research supervised by Christopher Longuet-Higgins.
  • Professor Hinton is currently the leading a government-backed research facility in Toronto who wants to create a partially automated factory that would mass-produce transplantable stem cells into disease-fighting cells using Artificial Intelligence (AI).
  • He and his two graduate students, Alex Krizhevsky and Ilya Sutskever, were hired by Google to develop Android voice search.

‘Godfather’ of deep learning is reimagining AI – Phys.org (phys.org)

Topics: University of Toronto, Toronto, New York University, Wired magazine, Google, Nicholas, Gary Marcus, large datasets

  • Hinton, a University Professor Emeritus at the University of Toronto, recently released two new papers that promise to improve the way machines understand the world through images or video – a technology with applications ranging from self-driving cars to making medical diagnoses.
  • ” This is a much more robust way to detect objects than what we have at present,” Hinton, who is also a fellow at Google‘s AI research arm, said today at a tech conference in Toronto.
  • With his new research, there’s little doubt Hinton is doing his part to move the AI ball forward – even if it draws on ideas he’s been contemplating for the past 40 years.

Noam Chomsky: Where Artificial Intelligence Went Wrong (theatlantic.com) – December, 2017

Topics: practical applications, unified theory, theoretical framework, biological system, cognitive system, information processing, statistical models, systems biology

  • The success of fields like personalized medicine and other offshoots of the sequencing revolution and the systems-biology approach hinge upon our ability to deal with what Chomsky called ” masses of unanalyzed data” —placing biology in the center of a debate similar to the one taking place in psychology and artificial intelligence since the 1960s.
  • One of the points he made was that AI and robotics got to the point where you could actually do things that were useful, so it turned to the practical applications and somewhat, maybe not abandoned, but put to the side, the more fundamental scientific questions, just caught up in the success of the technology and achieving specific goals.
  • Well, if success is defined as getting a fair approximation to a mass of chaotic unanalyzed data, then it’s way better to do it this way than to do it the way the physicists do, you know, no thought experiments about frictionless planes and so on and so forth.
  • One way to do it is okay I’ll get my statistical priors, if you like, there’s a high probability that tomorrow’s weather here will be the same as it was yesterday in Cleveland, so I’ll stick that in, and where the sun is will have some effect, so I’ll stick that in, and you get a bunch of assumptions like that, you run the experiment, you look at it over and over again, you correct it by Bayesian methods, you get better priors.
  • A very different approach, which I think is the right approach, is to try to see if you can understand what the fundamental principles are that deal with the core properties, and recognize that in the actual usage, there’s going to be a thousand other variables intervening—kind of like what’s happening outside the window, and you’ll sort of tack those on later on if you want better approximations, that’s a different approach.

The Godfather of AI Was Almost a Carpenter – Bloomberg (bloomberg.com) – December, 2017

Topics: artificial intelligence, Google, Emeritus Professor, University of Toronto, Geoffrey Hinton, Chief Scientific Adviser, preeminent expert, Vector Institute

  • He is an Engineering Fellow at Google, managing Brain Team Toronto, the Chief Scientific Adviser of the Vector Institute, and Emeritus Professor at the University of Toronto.
  • His name is Geoffrey Hinton, and this Bloomberg 50 profile of him takes a peek into the life of the world’s preeminent expert in artificial intelligence.

The godfather of artificial intelligence is a Torontonian (torontolife.com) – September, 2017

Topics: Google, Apple, Facebook, Uber, health care, neural networks, Toronto, government funding

  • He’s in charge of the Vector Institute, a U of T–affiliated organization that will apply AI to a range of fields, including finance, construction and health care.
  • Vector’s existence is the reason Google and Uber have created AI labs in Toronto—the institute only accepted international funding if its investors set up shop north of the border.

The True Father of Artificial Intelligence – OpenMind (bbvaopenmind.com) – September, 2016

Topics: Caltech, Marvin Minsky, Stanford university, California Institute of Technology, machine intelligence, artificial intelligence, personal computer, Ibm

  • However, McCarthy is enshrined as the father of artificial intelligence not only for managing to open the field and turn it into a new area of research, but also for continuing to provide evidence for its development for half a century.
  • “The speed and memory capacity of today’s computers may be insufficient to stimulate many of the more complex functions of the human brain, but the main obstacle is not the lack of capacity of the machines, but our inability to write programs that take full advantage of what we have,” he came to enunciate in those years.
  • However, despite his efforts, this system did not help McCarthy to achieve his true objective: that a computer would pass the Turing test, according to which a human asks questions through a computer screen, and if he cannot decide whether it’s another human or a machine that is responding, this is definitively intelligent.
  • Near the end of the research stage of his career, in 1978, McCarthy had to give up on his purist idea of ​​artificial intelligence: “To succeed, artificial intelligence needs 1.7 Einsteins, two Maxwells five Faradays and the funding of 0.3 Manhattan Projects,” he resignedly recognized.

Marvin Minsky, “father of artificial intelligence,” dies at 88 (news.mit.edu) – January, 2016

Topics: MIT Media Lab, Marvin Minsky, Harvard University, Seymour Papert, pioneering work, Artificial Intelligence, Princeton University, Nicholas Negroponte

  • Minsky joined the faculty of MIT’s Department of Electrical Engineering and Computer Science in 1958, and co-founded the artificial intelligence Laboratory (now the Computer Science and Artificial Intelligence Laboratory) the following year.
  • Minsky received the world’s top honors for his pioneering work and mentoring role in the field of artificial intelligence, including the A.M. Turing Award — the highest honor in computer science — in 1969.
  • Faculty Achievement Award; the Computer Pioneer Award from IEEE Computer Society; the Benjamin franklin Medal; and, in 2014, the Dan david Foundation Prize for the Future of Time Dimension titled “Artificial Intelligence: The Digital Mind,” and the BBVA Group’s BBVA Foundation Frontiers of Knowledge Lifetime Achievement Award.

Welcome to the AI Conspiracy: The ‘Canadian Mafia’ Behind Tech’s Latest Craze (recode.net) – July, 2015

Topics: big data, computer scientists, machine learning, deep learning, speech recognition, neural network, core technology, tech companies

  • In 2013, Hinton was hired as a distinguished researcher at Google, where he works on its expanding deep learning division; LeCun was tapped to lead Facebook’s AI efforts later that year; and last week, Ibm announced it was working with Bengio, a professor at the University of Montreal, to infuse Watson, its super-computer, with deep learning.
  • Fruits of their efforts are already starting to appear in front of consumers, with deep learning woven into products like the new Google Photos app and in the facial recognition technology infused in Facebook’s new app, Moments.
  • In 2012, the Google “Brain” team — a unit born in Google x with the audacious aim to build the largest artificial neural network, an AI brain — released a seminal finding: They sat the brain in front of millions of Youtube videos and, without input on feline features, it began spotting them.
  • The main reason it has been taking off in the last few years is scale,” said Andrew Ng, the scientist who launched the Google Brain team (called the deep learning team) and currently runs a similar team at Chinese search giant Baidu.
  • “There may or may not be products that come out of this for the next two or three or four or five years,” LeCun told Re/code.

How a Toronto professor’s research revolutionized artificial intelligence (thestar.com) – April, 2015

Topics: big data, Seymour Papert, speech recognition, computer vision, machine learning, DARPA, based systems, Silicon Valley

  • Hinton now spends three-quarters of his time at Google and the rest at U of T. Machine learning theories he always knew would work are not only being validated but are finding their way into applications used by millions.
  • He arrived that summer for what he describes as a trial run — he was hesitant to leave Toronto, where he has lived with his family for most of the past quarter-century — and the short-term stint didn’t have any other obvious job title.
  • The man who designed it claimed it would eventually be able to read and write, and the story said it would be the “first device to think as the human brainPerceptron will make mistakes at first, but will grow wiser as it gains experience.”
  • Ask anyone in machine learning what kept neural network research alive and they will probably mention one or all of these three names: Geoffrey Hinton, fellow Canadian Yoshua Bengio and Yann LeCun, of Facebook and New York University.
  • “It’s good to have some people considering the ethics and implications of this sort of thing, but it’s not something I’m worried about any time in the next, say, 40 years,” says Google senior fellow Jeff Dean.


John McCarthy: Computer scientist known as the father of AI (independent.co.uk) – November, 2011

Topics: Silicon Valley, Artificial Intelligence, Marvin Minsky, California Institute of Technology, Stanford, cloud computing, Steve Jobs, computer scientist

  • John McCarthy, an American computer scientist pioneer and inventor, was known as the father of Artificial Intelligence (AI) after playing a seminal role in defining the field devoted to the development of intelligent machines.
  • In 1958 he created the Lisp computer language, which became the standard AI programming language and continues to be used today, not only in robotics and other scientific applications but in a plethora of internet-based services, from credit-card fraud detection to airline scheduling; it also paved the way for voice recognition technology, including Siri, the personal assistant application on the latest iphone 4s.
  • Described as ” focused on the future,” McCarthy was ” always inventing, inventing, inventing,” and in the 1960s he conceived the idea of computer time-sharing or networking, which allowed users to share data by linking to a central computer; it ultimately lowered the cost of using computers.

What are the best machine learning blogs?

While building Frase‘s RSS feed reader, I had to follow blogs and publishers to test our feed monitoring platform. Being interested in AI, I started off by creating a list of machine learning blogs. I thought this list would be a good starting point for anyone trying to monitor news and insights from knowledgable AI-related sources. Most of them have RSS support and I plan on periodically updating this list.

Specialized Publishers

Non-profit Organizations

Academic Research and University-related


Large Companies

List was last updated: May, 16th 2018

Do robo-writers actually exist?

”Is Frase going to take my job?” – this is what writers frequently ask me while I am giving them a demo of the Frase Research Assistant. My answer is always the same: Frase doesn’t intend to replace the writer, but rather automate research tasks so you can focus on the creative aspects of content creation.

The closest Wikipedia entry for “robo-writer” is automated journalism:

”Also known as algorithmic journalism or robot journalism, where news articles are generated by computer programs. Through artificial intelligence (AI) software, stories are produced automatically by machines rather than human reporters. These programs interpret, organize, and present data in human-readable ways.”

Over the last five years, many stories have emerged about mysterious robo-writing projects, mostly in the context of how AI will radically disrupt journalism. Besides journalism, there is evidence about robo-writing being used in different contexts, from advertising copy to financial reporting. As I mentioned before, I I am personally in favor of robo-writing when automation can help writers become more creative and insightful.

The most shared stories on robo-writing:

For this post I’ve collected 12 of the most shared articles on the subject (ordered by published date), and then had Frase summarize them for me:

Robot Writing, AI, and Marketing: It’s the End of the World as We Know It (skyword.com) – January, 2018

Topics: big data, content marketing, lead generation, marketing automation, data mining, information architecture, internet of things, digital marketing

  • The definition of AI has merged and melded since the early days of sci-fi, so let’s clear this up: When we talk about AI today, we’re generally talking about “narrow AI,” or algorithms that are set up to do a very specific task, like trade a stock or make a widget.
  • Now, I’m here in Pittsburgh and there are self-driving Ubers everywhere on the road, so in a few years we employed a lot of folks, gave them a great livelihood and then all of those Uber drivers may potentially lose that livelihood.
  • If your job consists of a lot of repetitive marketing tasks like these, then you may want to consider how to evolve your role and skill set, as it will become more cost-effective and productive to have AI handle these types of activities.
  • “Truthfully,” he says, “I believe that where these jobs and these professions are going to shift to with the impact of AI is actually to the creation aspect; that’s where humans thankfully will still have a role.
  • AI does have a role in taking away a lot of the time that we spend looking at analytics and processing and crunching the data—all of that should be fulfilled by AI to free us up to do truly creative work.”

Coca-Cola chooses AI over brains to generate latest adverts (thedrum.com) – April, 2017

Topics: digital marketing, social media, Content creation, creative agencies, Coca cola, Adweek, Mobile World Congress, software algorithms

  • Coca-Cola is ditching flesh and blood creatives in favour of software algorithms in an experiment to see whether AI bots have what it takes to beat their human masters.
  • Mariano Bosaz, Coca-Cola’s global senior digital director, is spearheading the move as part of wider efforts to push the bounds of technology to see what they are capable of.
  • Bosaz added: “I don’t know if we can do it 100 percent with robots yet—maybe one day—but bots is the first expression of where that is going.”

Can Artificial Intelligence Replace The Content Writer? (digitalagencynetwork.com) – August, 2017

Topics: content marketing, customer experience, structured data, augmented reality, natural language processing, Gartner, data processing, artificial intelligence

  • While less likely to be automated, these areas are not exempt—roles such as data collection and data processing, once revered for their level of required expertise, are now 64% and 67% likely to be automated, respectively.
  • This can be seen in the above example: the second extract, written by Wordsmith, is more matter-of-fact and event-driven than its human counterpart (and the overuse of ‘season’ at the end particularly gets to me).
  • With the information already at hand, writers will have more time to focus on how articles are structured, how argument or opinion is built to leave the best impact on the reader.
  • Analyst giant Forrester have claimed that 16% of jobs in the U.S. will be lost to artificial intelligence by 2025.

Can robots truly be creative and use their imagination? (theguardian.com) – February, 2017

Topics: neuroscientist, research associate, professor emeritus, University of California, associate professor, machine learning, University of oxford, Artificial intelligence research

  • I’ve been working on writing novels computationally for well over 10 years now and I’m still trying it, although I believe that within the next two to three years I will have broken its back and will produce 100,000-word novels in half an hour or so, novels that I think most people would consider to be creative.
  • We are used to machines being used as tools that do not have a high level of cognitive ability, so it’s difficult for people to think of them as being able to exhibit truly creative behaviour.
  • Another problem is that it is difficult to automate the combination of ideas from many different sources that forms the source of much of human creativity: you might find inspiration from an interview with a neuroscientist in designing a new office layout.

What News-Writing Bots Mean for the Future of Journalism (wired.com) – February, 2017

Topics: BuzzFeed, Usa today, Fox News, Washington Post, Twitter, news articles, Megyn Kelly, Los Angeles Times

  • “Republicans retained control of the House and lost only a handful of seats from their commanding majority,” the article read, “a stunning reversal of fortune after many GOP leaders feared double-digit losses.” The dispatch came with the clarity and verve for which Post reporters are known, with one key difference: It was generated by Heliograf, a bot that made its debut on the Post’s website last year and marked the most sophisticated use of artificial intelligence in journalism to date.
  • It works like this: Editors create narrative templates for the stories, including key phrases that account for a variety of potential outcomes (from “Republicans retained control of the House” to “Democrats regained control of the House”), and then they hook Heliograf up to any source of structured data—in the case of the election, the data clearinghouse VoteSmart.org.
  • There may not be a wide audience for stories about the race for the Iowa 4th, but there is some audience, and, with local news outlets floundering, the Post can tap it.
  • “If we took someone like Dan Balz, who’s been covering politics for the Post for more than 30 years, and had him write a story that a template could write, that’s a crime,” Gilbert says.


New AI Can Write and Rewrite Its Own Code to Increase Its Intelligence. (futurism.com) – February, 2017

Topics: machine learning, deep learning, Google, program synthesis, MIT Technology Review, learning algorithms, large companies, mathematical framework

  • A company has developed a type of technology that allows a machine to effectively learn from fewer examples and refine its knowledge as further examples are provided.
  • For example, an AI system is fed data about how the sky is usually blue, which allows it to later recognize the sky in a series of images.
  • This form of probabilistic programming — a code that uses probabilities instead of specific variables — requires fewer examples to make a determination, such as, for example, that the sky is blue with patches of white clouds.


Automated content: Can algorithms write your content for you? (futurecontent.co) – September, 2016

Topics: online content, data science, Fast Company, Steve Jobs, Yahoo, business intelligence, Microsoft, Medill School of Journalism

  • French spent years scanning portions of two Susann books, Valley of the Dolls and Once Is Not Enough, into Hal’s databanks, and deconstructing the writer’s style into 100 different parameters which Hal then turned into the final prose.
  • Joseph Medill was an editor and journalist in the purest sense, and he created a publishing legacy both familial—three of his grandchildren went on to run newspapers—and professional, through the Medill School of Journalism (MSJ).
  • We know more about what readers read than ever before, but we also know how they interact with articles, what they look at next and what their responses mean.
  • Add to this vast improvements in other technology like facial recognition is it’s not beyond the realms of possibility that an AI will be able to pull in all these different data sets and create a story from scratch with all the nuances of a seasoned writer.


Artificial Intelligences Are Writing Poetry For A New Online Literary Magazine (popsci.com)

Topics: postmodern, artificial intelligence, human beings, average person, traditional sense, curation, common sense, good poetry

  • The project’s name is itself an apt title for the work done by humans for the site: the implementation of an artificial intelligence designed to write, and the curation of what it has written.
  • ” You can talk about what the creator was trained on, or how the creator works, but not the creator’s intent— maybe the algorithm writer’s intent, but it’s a step removed, which is more fun for the reader, I think.”
  • Still, reading an excerpt from ” Gimble” in a poem titled ” Madness,” makes one wonder how detached we can be from a machine’s ability to synthesize what reads as a believable abstraction on such an emotional and human subject:
  • i am all the world and the day that is the same and a day i had been


Google’s AI has read enough romance novels to write its own (thenextweb.com) – May, 2016

Topics: Google, final round, software engineer, annual event, Buzzfeed, Japan, deep learning, National Novel Writing Month

  • In an effort to make its apps more conversational, Google fed its AI engine a whopping 2,865 romance novels so it can improve its understanding of language.
  • After going through the massive trove of novels, the engine was tasked with writing sentences of its own based on what it had learned.
  • Given the rapid pace of development in the fields of AI and deep learning, it seems like the day isn’t far off when our next read will come not from a library shelf, but from a computer that tailors a custom book to your exact specifications.


Will Robots That Can Write Steal Your Creative Job? | Observer (observer.com) – April, 2016

Topics: machine learning, analytical tools, virtual world, artificial intelligence, blog post, Google, Ibm, computer program

  • In a way, it felt like going to see any other indie rock band, but I got the sense that the computer doesn’t have much sense for when a song should come to an end.
  • When I saw Narrative Science’s Kristian Hammond on a SXSW panel, he described how his company has built software that can turn piles of data into texts that can be read and digested by human brains.
  • For example, you might rely on some expert’s framework to devise a stock market portfolio for yourself, but you can’t call that expert up every night to ask how your picks are performing in light of his or her system.
  • Sometimes referred to as the economic singularity, it posits that soon so much more work will get done, and done better, by machines that hardly any people will be needed at all.


Japanese AI Writes Novel, Passes First Round for Literary … (digitaltrends.com) – March, 2016

Topics: Japan, artificial intelligence, press conference, writing process, award committee, Matsubara, literary award, human creativity

  • Yet, now that a Japanese AI program has co-authored a short-form novel that passed the first round of screening for a national literary prize, it seems that no occupation is safe.
  • One of the team’s two submissions to the competition made it past the first round of screening, despite a blind reading policy that prevents judges from knowing whether an AI was involved in the writing process.
  • But there are still some problems [to overcome] to win the prize, such as character descriptions,” said Satoshi Hase, a Japanese science fiction novelist who was part of the press conference surrounding the award.


This News-Writing Bot Is Now Free for Everyone | WIRED (wired.com) – October, 2015

Topics: Yahoo, news stories, beta version, structured data, mail merge, total sales, Associated Press, similar technology

  • Today Automated Insights has launched a beta version of its new free service based on Wordsmith, the technology it uses to generate stories for companies like the AP.
  • There are many rules—known as branches—that you can set, such as the ability to use one set of words when a variable happens to be greater than a certain number and a different set when it happens to be lower than that number.
  • You could create a template that will generate the text ” sales increased in quarter two” if the number in the spreadsheet cell containing the quarter’s total sales was bigger than the cell for quarter one.

How is NLP impacting content marketing?

Business leaders are starting to research more about technical topics like natural language processing and machine learning. In this post, we summarized the top 10 results on Google (as of May, 14th, 2018)  for the query “how is NLP impacting content marketing?”.  As a quick spoiler, these were the topics that got mentioned in at least 50% of the articles: unstructured data, business intelligence, search engine optimization, sentiment analysis, neural networks, chatbots, customer experience, automated content creation, digital transformation.

How can NLP technology be used for marketing (econsultancy.com)

Topics: business intelligence, unstructured data, structured data, search engine optimization, sentiment analysis, mobile apps, machine learning, scalability

  • The holy grail of NLP has been to convert ‘unstructured data‘ (text and multimedia) into structured data and we are inching closer towards that goal with advances in NLP and multimedia indexing technology.
  • Social prospecting solutions require NLP capabilities that can sift out passing mentions of a brand, and focus on those where this is an intent to purchase.
  • State of the art NLP systems can mine social media for such expressions of interest and return social handles of people matching the customer’s criteria.
  • The interaction of these two leads to compelling solutions, for example in leveraging socially trending topics (or brands) to promote customer content matching those topics or brands.

What Does Natural Language Processing Mean for Writers (portent.com)

Topics: natural language, machine learning, Natural Language Processing, neural networks, Artificial Intelligence, word processor, chatbot, language model

  • It’s not news that there are things computers are really good at that humans are bad at, and some things humans are really good at that computers can’t seem to manage.
  • Google made a splash in 2015 when the neural networks they’d trained on millions of images were able to generate pictures from images of random noise, something they called neural net “dreams.” And in 2016, they announced Project Magenta, which uses Google Brain to “create compelling art and music.”
  • “[Google’s algorithm] Hummingbird can find some patterns that can give it important clues as to what a text is about,” says Matthew, “but it can’t understand it the way a human can understand it.
  • Language takes root best in our “procedural memory,” which is the unconscious memory bank of culturally learned behaviors, rather than in our “declarative memory,” which is where you keep the things you’ve deliberately worked to “memorize.” Children can pick up other languages more easily than adults because they’re tapping into their procedural memory.
  • We can go into the idea that robots are going to take over the world and they just need to learn to speak first, and that’s kind of cool for a movie.

What do linguists make of AI and natural language processing? (econsultancy.com)

Topics: target language, natural language, machine translation, natural language processing, statistical machine translation, translation process, language professionals, foreign languages

  • One might assume that professional linguists would feel a constant need to stay updated on any impact such technology may have on their future careers.
  • — there is an apparent lack of understanding of the connection between AI, NLP, and advanced machine translation such as statistical machine translation (SMT) or neural machine translation (NMT).
  • Once the usage and purpose of statistical and neural machine translation are understood, most linguists start to think about how it could support them in their work instead of ignoring it as a threat to their livelihood.
  • It’s difficult to say how many linguists would admit to using SMTs for their translation projects, as despite improvements in the technology, its acceptance amongst clients is still low.

Can Artificial Intelligence Replace The Content Writer? (digitalagencynetwork.com)

Topics: content marketing, customer experience, structured data, augmented reality, natural language processing, Gartner, data processing, artificial intelligence

  • While less likely to be automated, these areas are not exempt—roles such as data collection and data processing, once revered for their level of required expertise, are now 64% and 67% likely to be automated, respectively.
  • This can be seen in the above example: the second extract, written by Wordsmith, is more matter-of-fact and event-driven than its human counterpart (and the overuse of ‘season’ at the end particularly gets to me).
  • With the information already at hand, writers will have more time to focus on how articles are structured, how argument or opinion is built to leave the best impact on the reader.

Artificial Intelligence for Content Marketing and Content Creation (techemergence.com)

Topics: content marketing, best practices, healthcare industry, content creation, search engines, Google, customer service, deep learning

  • With improvements in AI technology, newer NLP platforms can augment human researchers by creating multi-page summarized articles for even open-domain questions like ‘what is the future of AI technology going to look like?’ or ‘How is AI affecting the healthcare industry?’
  • Summarizing content through the use of NLP can be either “extractive” (where the system distills text into just the most relevant parts, cutting out the rest) or “abstractive” (which is machine learning based, and involves AI coming up with it’s own “wording” for summarizing a given text).
  • As opposed to what Google does in summarizing fact based questions, more advanced contextual summarizing would involve condensing information from the top 50 search results, finding meaningful relationships between these results and then extracting the most appropriate sentences from within those relationships.
  • Essentially, the platform offers a word processor that can learn from things that you write and can research contextual topics in the background and give you links for the topic etc.

Can Neuro-Linguistic Programming really Influence Sales? (maximizer.com)

Topics: emotional responses, skill sets, empirical support, Virginia Satir, human psychology, desired outcomes, specific goals, basic concepts

  • Stephen Covey’s habit of “begin(ing) with the end in mind” applies, a clear vision of a path to a successful sale, looking for visual clues that monitor how successful your sales pitch is, or whether your attempts at conversation are working well or failing miserably.
  • While wholesale subscription to NLP principles like anchoring and reframing may be less than ideal, managers should be aware of some general concepts that could impact an individual’s performance and potential as a team member.
  • This works on both sides of marketing equations, where sales prospects are less wary of the dogma of high-pressure salespeople that immediately put them on edge, and sales pros are less likely to be intimidated by negative feedback and clients that are unimpressed by their products or services.

How NLP Systems Help Marketers | Centric Digital (centricdigital.com)

Topics: natural language processing, natural language, target audiences, actionable insights, digital communication, digital transformation, social media, computational linguistics

  • The ability to keep a finger on the pulse of consumers’ reactions to a product on social media is immensely valuable, since, in the words of Apiumhub, “It has the potential to turn all of Twitter or Facebook into one giant focus group.” Countless SaaS offerings accomplish this by scraping the text away from social or publishing platforms and providing actionable insights based on interpreting those statements.
  • In fact, countless components of a potential consumer journey can be mined with the same datafied approach, creating a thoroughly personalized marketing touch point and driving sales on the basis of an overall sense of each consumer’s style and preferences.
  • As previously stated, the simple Amazon tracking that occurs when someone views a product page is a one-size-fits-all operation: The ad engine isn’t quite sure whether the consumer bought the item or not, and so it surfaces an ad for that item repeatedly in the middle of almost every subsequent site they visit.
  • If marketing content could make meaningful use of expressed language from the user’s entire web journey, including search terms, social media posts and reviews written customers would receive information in their preferred medium.
  • For example, when the user searches for a “baggy sweater” with the intention of finding a sweater that fits loosely, and instead the search returns sweaters made of a bag-like material, or sweaters with tote bags depicted on them — this is not product marketing on par with a premium ecommerce shopping experience.


Artificial Intelligence–the Next Frontier In Content Marketing  (blog.marketo.com)

Topics: unstructured data, machine learning, natural language processing, neural networks, deep learning, marketing automation, natural language generation, seamless integration

  • With the ability to process an enormous amount of unstructured data and decipher natural language, AI is used to extract insights and make recommendations based on previously established criteria.
  • This capability allows marketers to fully leverage the power of personalization and marketing automation technologies to deliver targeted content to each prospect or customer and increase the ROI of their content marketing efforts.
  • You can use AI to identify trending topics by using algorithms to track conversations on the Internet, such as those occurring on social media and within published content, to help you stay ahead of the trends and create content that will lead the conversations.
  • Leveraging the AI features in your current tools will not only give you a great starting point to familiarize yourself with the technology, but you’ll also be able to take advantage of the seamless integration that’ll allow you to get up and running faster and more cost-efficiently.

Artificial Intelligence in digital marketing (besttechie.com)

Topics: big data, search engine optimization, social media marketing, content marketing, digital marketing, search queries, search engines, Whatsapp

  • One of the most popular and common words that are consistently used in the world of marketing is “segmentation.” There are plenty of people with a variety of needs and interests, and the companies need to categorize them as segments to increase their sales and retain their customers successfully.
  • There are plenty of benefits that are brought by AI in the world of digital marketing, but there are many challenges that are faced by the digital marketers due to the entrance of AI.
  • AI based marketing tools can assist the digital marketers in developing an effective marketing strategy, but the thing is that these tools can be easily available to all digital marketers, which forces the digital marketers to increase their skills and constantly be aware of new tools and services in an effort to stay on top of their game.
  • For now, the digital marketers should keep on learning the new digital marketing skills, and they should keep themselves updated with the recent updates on artificial intelligence to be competitive in the world of digital marketing.

Content Intelligence: Will AI-powered Content Marketing be a game changer? (martechadvisor.com)

Topics: email marketing, content management, user experience, marketing automation, Content Marketing, customer engagement, data management, individual users

  • -A better breakdown of audience data by using data management platforms like Lotame can tell a content marketer what type of reader is consuming their content, what the reader’s other category interests are, location, etc.
  • Like several other companies, Wayblazer leverages the Ibm Watson technology for the travel industry by focusing on the language recognition API to analyze triggers from a traveler’s search so that they are further be able to share personalized hotel recommendations.
  • Also gone are the days of waiting for the RJ to play your favorite track what with platforms like Spotify, (a world digital music service) using AI and deep learning to provide music recommendations to users based on their past listening preferences.
  • To answer the glaring question as to why content marketers are increasingly looking at AI to supplement their content marketing efforts, Mark Schmukler, CEO and Co-founder, Sagefrog says, “Personalization has a big effect on the success of content marketing and can be well-executed through artificial intelligence and marketing automation tools.
  • A marketing agency with expertise in content marketing and marketing automation can help marketers find the tools and tactics that together create the best possible user experience throughout a content marketing campaign.”

How is voice search impacting SEO?

Voice search and its implications in SEO is a big topic. In this post, we look at the statistics mentioned in the top 20 articles for the search query: “impact of voice in SEO”.  Most of them discuss the growth of voice search and its adoption by demographic.

Share of voice search:

The takeaway: most sources agree that today’s share of voice search is about 20% and will grow to 50% by 2020.

  • ComScore expect 50% of online searches will be made by voice by 2020. (searchrise.co.uk)
  • A few resources reveal the fact that approximately 20% of the users are voice searchers and the percentage is keep on increasing continuously. (techcolite.com)
  • Around 20% of queries that are asked on Google’s mobile app, Android devices are voice searches. (dotzweb.com)
  • Depending on which survey or expert you accept, local searches represent somewhere between 20 percent of 30 percent of all voice search queries. (practicalecommerce.com)

Frequency of usage

The takeaway: the rate of daily usage is increasing, and 41% of new voice users only started 6 months ago, which shows how emerging this market is.

  • Google reports that 55 percent of teens and 40 percent of adults use voice search daily; and, according to Google’s Behshad Behzadi, the ratio of voice search is growing faster than type search. (searchengineland.com)
  • Nearly 27% of 2,000 people surveyed use voice search at least once a week, and approximately 22% of these use it every single day – a number expected to grow as AI continues to improve and new devices reach the market every year. (jeffbullas.com)
  • In fact, 40% of adults perform a voice search once per day.” (digidiscuss.wordpress.com)
  • “41 percent of all adults have used voice-activated search on their mobile devices at least once (most frequently, to get directions), and more than 55 percent of mobile phone users between the ages of 13 and 18 use voice-activated at least once per day (most frequently, to initiate a phone call).” (propeller.co.uk)
  • It was also stated that 41% of people using voice search last year had only started in the previous 6 months, and that 1 in 5 searches on an Android app in the USA were through speech. (sozodesign.co.uk)


The takeaway: voice search seems more prevalent among teens, but adults are feeling more comfortable with it and make them feel tech savvy.

  • A study from Northstar Research stated that 55% of teens (13-18) and 41% of adults used voice search more than once a day. (translatemedia.com)
  • The study also concluded that people generally use voice search while multi-tasking with 54 per cent of teens using voice search socializing with friends and 23% of adults searching hands-free while cooking. (translatemedia.com)
  • There’s very little data right now on personal assistant usage, but one report suggests over half of adults (56%) are using tools like Siri on their mobile devices. (geonetric.com)
  • Adults are feeling more comfortable with it: 41% talk to their phones every day and 56% say it makes them “feels tech savvy.” (glean.info)

Local-based search

The takeaway: hyperlocal and personalized experience.

  • Mobile voice search is three times more likely to be local-based than text search. (moz.com)
  • It is three times more likely that a voice search will produce local results than it is for a text search. (creativeclickmedia.com)

Technology improvement

The takeaway: it looks like voice technology is ready to go mainstream.

  • Today, Google’s speech recognition error rate is only 8%, down from ~25% just two short years ago. (moz.com)
  • In fact, two years ago word error rate was over 20%, but current speech recognition word error rate is as low as 8%—a huge leap in a short amount of time. (searchenginejournal.com)
  • A report by USA Today shows that the word error rates had come down to roughly 8 percent from 25 percent a few years ago. (resultfirst.com)
  • With Google‘s assistant, nearly 70% of requests are “natural” or made in “conversational” language, and technology has evolved enough that voice recognition accuracy is now at about 95%. (sozodesign.co.uk)
  • With time, word error rate has decreased to 8% and getting better at picking up speech quirks. (aimteck.wordpress.com)

Schema markup

The takeaway: there will be plenty of optimization work to be done by digital marketers.

  • With only 0.3% of websites using schema markup today, you’ll be capitalizing on a huge source of SEO. (jeffbullas.com)

Mobile vs. Desktop

The takeaway:  search results from desktop and mobile are dramatically different, which also requires different SEO strategies for both worlds.

  • BrightEdge (my company) research found that 79 percent of keywords return different results across mobile and desktop, which points to the fact that users expect different content depending on their context. (searchengineland.com)