The question of whether writing could be automated through the development of Artificial Intelligence (AI) reaches the heart of what imagination and creativity are for humanity. AI is developing with increasing speed, impacting our daily lives in more and more ways. How humans will deal with the development of AI in the future is yet to be seen.

Deep Learning is expanding the potential for machines to mimic human thought patterns with increasing speed and accuracy. The developing neural networks and capacity of big data soon may offer real competition for human neural networks in writing, and many other areas as well.

How humanity will respond to this challenge, and the changes which machine evolution will present, will likely impact activities and jobs in the same manner technology continues to influence culture today. Throughout all human stages of cultural advancement through technology, things have been lost along the way, while others have been gained.

Along this path of adventure and creation, one aspect of humanity remains the same. We are storytellers who continue to push the boundaries of our own story.

This blog will present an overview of the differences between AI, Machine Learning, and Neural Networks as well as the current capacities of AI and a projection of the next decade’s advancement. It will also speculate how writing may become automated, and the role of creativity may play in the future.

The nature of writing automation and creativity will be analyzed, along with the actual and philosophical implications of AI. For writers who are interested, may the routes invested in this blog empower your own creativity.

Deep Learning: The differences between Machine Learning, Neural Networks, and AI

Artificial Intelligence (AI) is currently in its infant stages, and just as infants process information differently than mature minds, the beginning permutations of AI are categorized under the term “Deep Learning.”

Relatively new, Deep Learning has allowed for an expansion of AI applications 35 times in the past three years. Deep Learning is a form of predictive algorithms similar to what the human brain uses to read. When reading, the brain sees an outline of a word, and in a matter of microseconds isolate the possible meaning of the shape.

Deep Learning

Deep Learning techniques are currently used by Google in the image and voice recognition algorithms, by Netflix to determine what shows you are likely to enjoy, and by MIT researchers to predict future events. As a part of the intelligence community, Deep Learning looks to synthesize and streamline knowledge gathering processes, allowing technology to accurately, and seamlessly mirror humanity’s desires.

Within the discussion of Deep Learning there has been confusion on the differences between AI and Machine Learning (ML). This difference is succinctly described as:

“Artificial Intelligence is the broader concept of machines being able to carry out tasks in a way that we would consider ‘smart’. And, Machine Learning is a current application of AI based around the idea that we should really just be able to give machines access to data and let them learn for themselves.”

Machine Learning

In a sense, Machine Learning is like a type of training wheel for the developing structure of AI. Deep Learning further narrows the focus on Machine Learning, which creates problem solving processes for human problems and desires.

The process of Deep Learning is providing a computer system with a large amount of data, which it can use to make choices about how to solve problems and manage other data.

This data is accepted and moved through Deep Learning’s neural networks which are constructions of binary true/false questions which create logical pathways. Deep Learning is focused on developing larger neural networks in which to support the demands of evolving big data.

Neural networks

Within the process of Machine Learning, neural networks enable computers to process information in a similar fashion to how a human brain does. These choices are based on probability algorithms of binary choice. The process of learning occurs within the feedback systems, which inform the computer if they made the correct choice.

Currently this process of Machine Learning enables programs to understand the gist of text (if it’s complaint or praise), analyze a piece of music to determine if it will make someone happy or sad, and compose music with similar thematic elements. Within the field of Natural Language Processing (NLP) machine learning has been a source of support to the efforts to grasp communication challenges.

In turn, this process helps machine learning understand how to communicate with diverse human groups in ways it projects to be well received.

The ideological history of AI

The conception of AI has been around for a great deal of human history. It was present in times of Greek mythology. However, as humanity’s capacity and definitions of technology have evolved and changed, so do the conception of what AI is and can be.

As a result of narrative storytelling found in science fiction works, humanity has become comfortable with the idea of interacting with electronic devices in ways similar to how they would interact with other people. This has resulted in the ubiquitous acceptance of the technology found in a smart phone.

In the developmental stage, those working on AI development have focused on the capacity for machines to make decisions, and carry out tasks like a human mind would. However, most experts believe it will likely compute at speeds that far exceed human capacity.

While some theorists would immediately say AI can do a better job, there remains much debate about the value of speed and capacity within a moral context.

Intelligence applied or general

AI is currently broken down into two schools of devices created to act with independent intelligence-applied or general AI.

Applied AI is more common, and seen as algorithms designed to intelligently trade stocks or drive an autonomous car.

General AI in theory can handle any task, and are less common and in development. Machine Learning is a form of general AI which is thought of the state of the art application of the science.

Rather than a program being written, true AI would be more akin to an entity being born. Unlike algorithms and programs which have stringent parameters, true AI would share humanity’s free will. However, just like humans, free will is governed by laws and possibly morality, AI theorists posit that an artificially aware entity could be governed by machine laws and have no regard for the laws of man.

Current capacities of AI

Today, there are many writing programs and tools which use aspects of artificial intelligence, but the full capacity of what AI can do remains to be seen in the realm of storytelling and philosophy.

A science purist may define AI as an artificially intelligent being who has achieved self-awareness, and could pass the Turing test of consciousness. AI may be defined by a programmer as...

...the capacity of a machine to intelligently imitate human thinking.

As such, the current capacities of AI are closely linked with the semantics of those who are creating it and for what purpose that is.

The process of AI evolution is expanding exponentially, as revealed by AI program AlphaGo. Go is believed to be the oldest board game in the world with records showing it being played almost 2,500 years ago. The possible combination of moves in the game is almost unlimited, and far exceeds the number of moves one can make in chess.

AlphaGo recently defeated one of the most capable Go players of all time. AlphaGo utilizes deep reinforcement learning, which builds on the foundation of machine learning through the application of positive reward feedback in the same manner many animals are trained.

Game time

Through withholding the positive reward of winning the game and through extensive experimentation with strategy and scenarios, AlphaGo was able to become a master level Go player, and may be unbeatable. Analysts emphasize that AI systems could be set up in many different simulated environments in which to be trained in deep reinforcement learning, and could streamline machine learning.

Machine learning is also being refined through the application of generative adversarial networks (GAN).

Developed by research scientist, Ian Goodfellow, GAN is the process of one network providing new data after a training set while another network works to decipher between fake and real data.

This process sharpens the discriminating capacity of machine learning, and has the capacity to de-blur video footage, generate video-game imagery, and apply stylistic changes to designs generated by computers.

Machine Learning is continuing to evolve quickly. The following list will give examples of some of the capacities and applications of AI at use today:

• Siri: The personal assistant program created by Apple, this voice activated computer is able to perform many functions by voice command that once were done through individuals searching.

  Siri manages schedules, writes, emails, finds information, provides directions, and many other features. This foundational form of AI is a form of Machine Learning which uses predictive and probabilistic algorithms to learn about their user’s preferences and style.

• Alexa: Amazon’s home management tool, Alexa has the premier speech recognition AI capacity which allow it to take commands throughout one’s home. This AI system manages a smart home’s capacity to regulate things like temperature, and security. It’s proven to be a boon for those with limited mobility.
  
  Alexa helps perform all the same functions as Siri, and is also a form of Machine Learning which has been embraced by the public.

• Tesla: The self-driving electric car utilizes predictive algorithms to learn about its driver. The utilization of Lidar and radar enables Tesla cars to “see” better than human drivers. Those who advocate this technology affirm it is only getting smarter and more capable as time goes by.
  
  
• Cogito: A fusion of machine learning and the behavioral sciences, Cogito is a customer support system with adaptive emotional intelligence, which has made it capable of replacing humans in this role. Millions of voice calls are being handled by Cogito daily.
  
  
• Boxever: Delivering micro-moments to improve client’s experience in the travel industry, Boxever utilizes machine learning to predict what its customers would find personally interesting to pleasantly surprise them during their travel.
  
  
• John Paul: This luxury travel concierge company enhances client relationships via AI predictive algorithms which aim to understand their client’s needs and desires even before they do. This customer support hub supports client interactions for many companies, and aims to be the go to concierge support system.
  
  
• Amazon.com: Amazon’s transactional programs are rudimentary forms of AI, but that has not quelled the amazing amounts of money the company has been able to glean from e-commerce. Amazon’s predictive algorithms continue to grow more refined in order to provide their customers with options which may best fit their desires.
  
  
• Pandora: This internet radio uses a system called musical DNA which analyzes songs based on 400 characteristics in order to allow users to design radio stations based on their likes.
  
  
• Nest: Nest is a machine learning thermostat that Google purchased for $3.2 billion, and is now a part of Alexa’s array in home comfort. This system employs predictive algorithms to determine user's desired home temperature throughout the day.
  
  
• Facebook: Facebook earns billions of dollars per year through employing predictive algorithms which pinpoint their users tastes and positions strategic advertising.

Chatbots

Machine Learning Discussion

Artificial Intelligence simulating human behavior is currently being utilized in the form of Chatbots, which are programs designed to communicate as a person would. This form of Machine Learning utilizes predictive algorithms in the attempt to mimic its users’ thought and communication styles. The following chart details the best chatbots available today:

Current capacities of AI writing

Machine Learning Language

The current capacities of AI in the practice of writing are emerging at startling rates of efficiency and interest to those in the intelligence community. Automated Insights has created Wordsmith, a solution which writes earnings reports from data much faster than humans could generate.

Gartner has predicted that as of 2018, 20% of all business content will be written by AI programs. This includes legal documents, emails, shareholder reports, press releases, white papers, and articles.

Currently the Associated Press (AP) reported that the NBA’s Orlando Magic used Wordsmith to write emails to fans to help create interest and engagement. So far other companies are using Wordsmith to write product descriptions, automate internal reports, and generate online content.

This form of Machine Learning utilizes NLP language generation to turn raw data into intelligible reading content. Wordsmith utilizes the process of Machine Learning to be adaptable to any platform it is needed for. As the director of Automated Insights expressed:

Wordsmith is a natural language generation (NLG) platform. That’s an academic way of saying it lets anyone build out their own artificial intelligence to automate their writing. For example, an investment bank could automate commentary on customers’ portfolios. For the first time, that bank could explain changes in the portfolio in understandable prose, as though an expert was personally explaining it to each customer, every day. The final templates become very sophisticated. So sophisticated, in fact, that academic studies have shown that readers cannot distinguish the content from a Wordsmith user’s template from articles written manually by journalists.

Facebook’s heavy investment in AI technology and research has enabled AI to develop their own language to communicate with other AI systems. This language is unintelligible to humans and has the capacity to bypass the need for APIs (Application Programming Interface).

APIs were created for multiple computer programs to share data and understand each other, but they were slow and costly to create.

The AI generated language transfers data faster than human language using ciphers, which are data packs like words that have the capacity to hold multiple meanings at once.

This is a huge leap forward in the capacity for Machine Learning and the autonomous evolutionary capacity of AI. However, this leap makes human programmers uncomfortable, as it enables AIs to have covert interaction.

This is because thus far, the focus of the development of AI interfaces are those which mimic humans. Because of this covert interaction the development and empowerment of the language has been squashed by developers.

While AI platforms like Wordsmith are revolutionary for content writers, they are still only as strong as their programming. Some business managers advocate that, for high quality content which engages readers, human writers still reign supreme. When AI programming is undercut because it is too different or strange for developers, the possibilities of a truly foreign development of an AI consciousness are curtailed.

It appears that writing bots are good for organizing raw data and creating templates, but the meat and potatoes of content creation remain a human domain. The question remains...

How long will this be the case, and is it possible for AI to overtake writers in all forms?

AI capacity projected for the next decade

Moore’s Law

The co-founder of Intel, Gordon Moore, made a prediction in 1965 which has since been coined “Moore’s Law.” Moore’s Law emphasizes that…

due to the accelerating rate of technological advancement that computers would dramatically and exponentially increase in power as it decreased in cost.

This insight has been proven true, and the potentials for its expansive application in the next ten years seem to bend the rules of science fiction and reality together.

The following elements reveal the proposed expansion of Moore’s law for the next decade:

1. A $1000 human brain
Tech trend analysts project that by the year 2025, a computer which costs $1000 should reach the same computing speed of the human brain: 10^6 cycles per second.

2. Perfect knowledge
This refers to the conception that the prevalence of AI mixed with big data and intelligence software will equate to the ability to know any type of data you desire in real time

3. Hyper-connection of all humanity
Marc Zuckerburg is working on Wi-fi for the entire world in the hopes of being the means through which all humanity connects to the Internet. Pouring billions of dollars into AI research it is Facebook’s dream to be the one stop shop for all of humanity’s needs: personal assistant, banker, therapist, travel guide, and friend.

The hyper-connectivity of having all people online will add 5 billion new consumers to the e-commerce market, and drastically increase the logistics of global shopping. While this offers amazing opportunities for people in developing nations to have access to the material goods of the developed world, it also presents major new trash and post-consumption materials challenges.

4. The Internet of Things (IoTs)
The Internet of Things is the interfacing of the material world with the digital, and includes smart houses, cars, appliances, and the possibility of interfacing the human body through implants. While the skeleton for this reality is already in place and an ethical language is being defined to discuss the many theoretical implications of this emergence, must to be refined in the next decade to bring the Internet of Things online.

The goal of the IoTs is to make a digital copy of all that is in 3D to interface with it completely, and theoretically this could come to replace reality for some users. AI would be the backbone of this interface, and already Cisco is projecting this type of data interface would easily create $19 trillion dollars of new value.

5. Reprogramming of healthcare
AI technology has the capacity to radically undercut the warped profit driven healthcare model through cutting out the middleman. AI doctors, diagnostics, biometric sensors, and the application of developing tech may erase the need to depend on the current corrupt model of healthcare delivery.

Digital healthcare is projected to be a $3.8 trillion market, and many tech giants are already cutting up their piece of the digital pie. The use of genomic sequencing applied with machine learning could help prevent disease and disorder outbreaks before they occur, and greatly improve the way healthcare information systems run.

The inclusion of robotic surgeons carrying out autonomous procedures may undermine the need for human specialists. This reality may lead to a radically more affordable health care climate which would lead to a healthier population who, living longer, would need new avenues for expression.

6. Virtual and augmented reality
Integral to the Internet of Things, virtual and augmented reality may come to replace many of the aspects of reality which define culture. Specialized interfacing eyewear could become ubiquitous, rewriting how cities are seen and interacted with, and giving reality a virtual “sheen”.

Currently Sony, Microsoft, HTC, Facebook, Google, and Qualcomm are working to have the most effective means of capitalizing on this trend.

7. Next generation Siri
Thought of as J.A.R.V.I.S. from Iron Man, the next generation Siri could be a more advanced AI interface with expanding capacities to reason, respond to questions, and apply theorizing to production.

Much like how Tony Stark bounces ideas off J.A.R.V.I.S. and has J.A.R.V.I.S. run simulations ultra-fast while he is chewing on the next theoretical application. This type of AI could be the genius best-friend tech heads always dreamt of.

8. Blockchain
Blockchain is the matrix on which bitcoins are supported by. This form of cryptocurrency enables decentralized and democratized value transference. Blockchain has the potential to be a system of value communication and exchange which could undercut the power hold banks have to set and manage value. Investors are developing this technology with this motivation in mind to help streamline currency exchange.

9. AI and employment concerns
Widely respected venture capitalist, Kai-Fu Lee (Sinovation Ventures) has projected that robots will replace 50% of all jobs within the next decade. As a result of the hyper-competitive globalized economy, and the expense of employing human labor in many repetitive jobs it is likely that automation could be the largest force impacting the coming job market.

This eventuality has already impacted many industries, and the impact has been considered for some time without any real supportive alternatives being implemented. Called “the decision engine that will replace people”, Lee emphasizes that humans cannot compete with robots in many ways.

While Lee emphasizes that this change has the capacity to eliminate poverty it is unclear how that would occur without the wild socialization of food and housing which many of the newly unemployed would require. Lee believes that robots will never be able to replace the capacity for human’s ability to touch the heart.

However, many human interactions have come to take on mechanistic feels during this period of fascination with technology. This leaves two main possible routes if and when robots replace 50% of human’s jobs:

• A newfound age of human cooperation and evolution with the support of AI, or;
  
  
• Widespread poverty and depression as much of humanity is left without purpose or the means of supporting themselves.

Companies most invested in AI technology

Google

Looking to maintain its dominance on the web, Google is at the forefront of AI development. Purchasing the intelligence startup, DeepMind, at $400 million was one of the largest AI investments as of yet.

A form of AI machine learning utilizing neural networks, DeepMind has been utilized in the healthcare reform debate, to expertly play games of strategy, and find the quickest routes of travel.

Google made TensorFlow, its machine learning system free for anyone who wants it. This summer Google has begun a new research project, People + AI Research Initiative (PAIR) in order to gauge how people will best interact with this new technology.

Facebook

Facebook has one of the most expansive investments in AI development, with the motivation of making Facebook accessible and irreplaceable for everyone. Opening a lab dedicated to AI development and research (FAIR), Mark Zuckerburg believes that AI is the future of human interaction. Zuckerburg hopes to develop AI capacity to learn their user’s deepest desires and help bring them to fruition.

Along with utilizing this form of data mining to improve advertising, it will also help make the social media platform more fulfilling for its users. Facebook utilizes facial recognition algorithms to identify people in pictures, and creating detailed population maps through which to plan web access modalities for a totally interconnected world.

Microsoft

In 2015 Microsoft created Project Oxford, an AI interface which aims to understand and communicate with users through face, emotion, and speech program interfaces (APIs). Created in 2016, Microsoft Ventures has been created to fund and support AI startups.

With the aim to help many AI initiatives development for the good of all humanity, this venture effort supports the evolution of Big Data, cloud computing, SaaS, security, and fleshing out the Internet of Things. The first brainchildren of this support structure have been Agolo and Bonsai. Bonsai organizes the management of machine learning algorithms autonomously, and Angolo makes information summaries in real time.

Apple

While Apple is investing in AI they are keeping their motivations for doing so largely to themselves. Having bought the AI start up Emotient and Vocal IQ, Apple may be competing with Facebook to in a race to create the first real version of a J.A.R.V.I.S. like program.

IBM

IBM is currently working on the development of an AI teaching support application, which will help with lesson planning, and improving their Watson computer. The Watson AI system can respond to many questions asked in natural language, making its mark by defeating human opponents on the quiz show, Jeopardy!

Can writing be automated?

Automated writing highlights the fact that there are many types of writing.

Writing ranges from heavily data driven quarterly reports and scientific papers, to creative fiction and postmodern works of writing, which fuses experimental and poetic approaches.

The answer to this question will be different based on what writing is being utilized for, what a person’s view of technology is, and will dynamically evolve as the capacity of AI technology does.

However, at the root of all AI technology is the human hand. Currently, Natural Language Generation (NLP) machine learning technology (which transcribes data into readable text) is only as capable as the programmer behind it.

This reality emphasizes that if, and when, AI plays a large role in creating written content, the human hand and mind will be behind this evolution. It appears that one aim of Machine Learning evolution is to reduce the amount of repetitive and redundant logistics work for humans, which is inherent in so much content creation and data analysis. Ideally, this could free up writers to work more creatively, which raises an interesting question.

Could AI technology ever come to replace human creative writers?

Genius vs. intelligence

A good example for this thought experiment is the writer David Mitchell, author of Cloud Atlas. Cloud Atlas is a web of different stories told throughout a wide timeline, and through multiple dialects. These multiple angles converge to reveal a fundamental spiritual truth about the nature of the human soul.

Some other genius-level writers, such as Kim Stanely Robinson, author of the Mars Trilogy, may be more readily accessible to mimicry due to his adhesion to facts, hard science, and proven human nature.

Mitchell’s work, on the other hand, stretches the boundaries of perception in the best ways.

Ultimately it is unlikely that AI would ever be able to replace human creativity and imagination in this way, as it is always evolving.

However, if machine consciousness and language were given the freedom to develop along its own lines rather than in patterns of human mimicry, it would be likely that creativity and imagination would result when AI reached a level of independent consciousness.

If AI systems are to be kept only in service of human masters without the ability to plumb the depths of the computer consciousness this emergence is much less likely.

Technology addiction and creativity

Another avenue of this question to be investigated is the influence of technology on human creativity. Currently AI algorithms successfully write news stories, pick winning music hits, win games, and compose music through mimicry and deep learning techniques.

While this represents an evolution of technology, it also may represent a devolution of human culture in this period of fascination, and possible over-fixation on technology.

Technology addiction results in many types of decreases in human reliance for many. This includes memory loss, attention span depletion, desensitization to images and news due to over-exposure, and the degradation in communication quality rooted in using technology to avoid connecting.

Technology addiction has aided a homogenization of creative culture for those who are over-reliant on technology, even as those who use technology to enhance creativity are showering humanity with unbridled new forms of art and invention.

Post-humanism

Known by theorists as the post-human phase of evolution, this would represent an emergence of human and machine consciousness. While some theorists posit this as an essential stage in evolution to overcome the weaknesses of emotion and violence which plague human relations, other theorists emphasize a post-human would lose what makes humans truly human. Are humans destined to become cyborgs?

Any peace in the post-human world would be a neutered peace, without the freedom of mind and spirit necessary to enjoy and utilize it. In a post-human world, AI could replace all writers, because creativity and imagination would be a simulacrum of what David Mitchell is. A copy of a copy, which post-humans would consume without the memory of the tactile impact of true genius.

If technology addiction continues to overwhelm most of humanity, human thought and expression may come to resemble machine thinking ever more so.

Future outlook

For every question and opinion about the human condition there are innumerable “experts” who express all positions to be had on a question. Throughout history those individuals whose actions have carved out their niche in the human community largely ignored “experts” and set themselves up as they saw fit.

It is likely that human determination will continue through any challenge, every evolution, and all changes to the global economy. A future in which writers’ observations and imagination will no longer be sought with avidity is unlikely.

As recent history has shown, a brief dip in quality and accessibility of quality print media led to a hunger which is satiated today by a resurgence of reading and consumption of written material.

Throughout the dip in the economy’s support of writers, they continued to create, driven by their own creative genius. When the tide of interest returned, they capitalized on their skills.

Humans are great storytellers, and we thrive in enjoying stories with higher and more dramatic levels of risk, sometimes saving the day at the last minute. These stories are told in novels, screenplays, theater, the news, international relations, and the drama we make of history.

Throughout humanity’s many phases of obsession imbalances are eventually met with a resurgence of balance at a higher level of cohesion and quality. Hopefully, this trend will continue and writers will always find a ready audience for their unique voice and vision.

Works Cited

Adams, R.L. “10 Powerful Examples Of Artificial Intelligence In Use Today.” Forbes.com, 10 Jan. 2017. Retrieved from: https://www.forbes.com/sites/robertadams/2017/01/10/10-powerful-examples-of-artificial-intelligence-in-use-today/#68d40ea0420d

Castrounis, Alex. “Artificial Intelligence, Deep Learning, and Neural Networks Explained.” Innoarchitech.com, 1 Sep. 2016. Retrieved from: https://www.innoarchitech.com/artificial-intelligence-deep-learning-neural-networks-explained/

Diamandis, Peter. “The World in 2025: 8 Predictions for the Next 10 Years.” singularityhub.com, 11 May 2015. Retrieved from:
https://singularityhub.com/2015/05/11/the-world-in-2025-8-predictions-for-the-next-10-years/

Hadfield, Jack. “Facebook AI invents new language not understandable by humans.” breitbart.com, 17 July 2017. Retrived from: http://www.breitbart.com/tech/2017/07/17/facebook-ai-invents-new-language-not-understandable-by-humans/

Hardesty, Larry. “Explained: Neural networks.” Mit.edu., 14 Apr. 2017. Retrieved from:
http://news.mit.edu/2017/explained-neural-networks-deep-learning-0414

Intel.com. “50 Years of Moore’s Law.” intel.com, 2015. Retrieved from:
https://www.intel.com/content/www/us/en/silicon-innovations/moores-law-technology.html

Kaput, Mike. “Will Artificial Intelligence Replace Writers and Content Marketers?” marketingaiiinstitute.com, 22 Nov. 2016. Retrieved from: http://www.marketingaiinstitute.com/blog/will-artificial-intelligence-replace-writers-and-content-marketers

Knight, Will. “5 Big Predictions for Artificial Intelligence in 2017.” technologyreview.com, 4 Jan. 2017. Retrieved from:
https://www.technologyreview.com/s/603216/5-big-predictions-for-artificial-intelligence-in-2017/

Marr, Bernard. “What Is The Difference Between Artificial Intelligence And Machine Learning?” Forbes.com, 6 Dec. 2016. Retrieved from: https://www.forbes.com/sites/bernardmarr/2016/12/06/what-is-the-difference-between-artificial-intelligence-and-machine-learning/#644476ae2742

Marr, Bernard. “What Is The Difference Between Deep Learning, Machine Learning and AI?” Forbes.com,8 Dec. 2016. Retrieved from: https://www.forbes.com/sites/bernardmarr/2016/12/08/what-is-the-difference-between-deep-learning-machine-learning-and-ai/#25617aad26cf

Mercer, Christina. “10 tech giants investing in artificial intelligence: What is their plan and who are other key players?” techworld.com, 11 July 2017. Retrieved from: http://www.techworld.com/picture-gallery/data/tech-giants-investing-in-artificial-intelligence-3629737/

Nusca, Andrew. “The Current State of Artificial Intelligence, According to Nvidia’s CEO.” fortune.com, 22 Mar. 2016. Retrieved from:
http://fortune.com/2016/03/22/artificial-intelligence-nvidia/

Roetzer, Paul. “How the Associated Press and the Orlando Magic Write Thousands of Content Pieces in Seconds.” marketingaiinstitute.com, 10 Nov. 2016. Retrieved from: http://www.marketingaiinstitute.com/blog/how-the-associated-press-and-the-orlando-magic-write-thousands-of-content-pieces-in-seconds

Souppouris, Aaron. “How a robot wrote for Engadget.” engadget.com, 15 Aug. 2016. Retrieved from:
https://www.engadget.com/2016/08/15/robot-journalism-wordsmith-writer/

Steiner, Christopher. “Can Creativity be Automated?” technologyreview.com, 27 July 2012. Retrieved from:
https://www.technologyreview.com/s/428437/can-creativity-be-automated/

Techlabs, Maruti. “What Are The Best Intelligent Chatbots or AI Chatbots Available Online?” chatbotsmagazine.com, 16 Apr. 2017. Retrieved from: https://chatbotsmagazine.com/which-are-the-best-intelligent-chatbots-or-ai-chatbots-available-online-cc49c0f3569d

Yan, Sophia. “Artificial intelligence will replace half of all jobs in the next decade, says widely followed technologist.” cnbc.com, 27 Apr. 2017. Retrieved from: http://www.cnbc.com/2017/04/27/kai-fu-lee-robots-will-replace-half-of-all-jobs.html

Zorabedian, John. “A sassy chatbot named Rose just won a big test of artificial intelligence.” sophos.com, 22 Sep. 2015. Retrieved from: https://nakedsecurity.sophos.com/2015/09/22/a-sassy-chatbot-named-rose-just-won-a-big-test-of-artificial-intelligence/