AI-written articles will force humans to improve the quality of their writing

by Ross Dawson

This year has seen massive advances in article writing by AI. The Open AI Institute released its GPT-2 article generator last month after earlier in the year saying it was too dangerously powerful to release to the public. You can try it with small text seeds here.

Another article generator introduced in May, Grover, is described as ‘A State-of-the-Art Defense against Neural Fake News’ which can detect whether AI wrote articles, however it also swiftly generates fairly good articles. Below I have put an article generated by Grover given only the title and my website as a reference (I do like how it calls me Professor!). Have a read or generate your own article. It isn’t high quality writing and it goes a bit off point, but it is arguably better than a lot of content available on the web.

There are of course many implications of anyone being able to readily generate a profusion of content on any topic, possibly injected with a specific ideological slant or implicit marketing message.

One of them is that many people currently earning money from writing will need to lift their game. This is most obvious for those who write primarily for search engine optimization. Yet there are plenty of others who will need to do better to kep ahead of the machines. 

This is one example of how improving AI will raise the bar on human performance. AI won’t be a true substitute for human writing for considerably longer than the foreseeable future. Yet human writing will have to, and will, get better. 

AI-generated article: 

How to create a positive future for AI in business and society

By Professor Ross Dawson

Artificial intelligence technologies can provide us with significant opportunities. The technology makes our lives easier in areas such as information and communication management.

The technology can also provide opportunities for targeted and relevant service in areas such as logistics and financial services.

A company like JPMorgan Chase has now implemented AI technology to enhance their customer service. Instead of fielding a 100 person customer support team to call in, they can now issue a 10 person call team for less than a third of the cost.

The potential applications are quite broad. It could even go from powering entertainment services, to helping diagnose and detect disease.

But not everyone is convinced by the potential for such technology. While it can certainly enhance and empower people, there are some concern that AI and related technologies could have negative consequences.

In its founding statement, the Cambridge AI foundation acknowledges concerns, stating that, “Artificial intelligence is already playing a major role in society, and there are fears that it might go too far and create social disruption.”

A poll by Emap asked people if they believed AI would make their job more complicated or more manageable in the future. The clear majority said it will make things more difficult, whilst only 30 percent said it will make things more manageable.

Gathering the views of 2,000 people in the last week, the company discovered that people were worried that automated jobs will lead to the loss of jobs, and that this could spark social unrest.

Some people didn’t feel confident that AI would actually create more jobs overall. The majority of people said AI is not a jobs creator and will actually lead to job losses.

People were also unhappy with the fact that it is unclear whether AI will be restricted to use in specific sectors, or will be able to operate across a variety of industries.

However, the survey did conclude that the public were keen for a more active role in shaping the technology. Many people were eager to receive AI/machine learning education and assistance, and were able to identify positive and negative impacts the technology will have.

This information has been gathered during the Emap AI study conducted by YouGov.

Read more:

So how can we, as a society, harness the potential of AI? What do we need to do to ensure that its benefits are to be harvested for the benefit of the whole?

These are the things I believe are needed:

Listen to people. Use AI/machine learning technologies, if it does pose risks to security and the environment, with responsibility. Develop strong language around the ethics of use. Come up with a design philosophy for both the technology and its applications.

Putting these in place will ensure that not only will we see technology to improve lives, but that we will be putting in place a way that AI can also make society better.

As an academic, my research focuses on addressing societal issues in artificial intelligence with skills based learning and computer modelling. In my years of research, I have found that it is at a very early stage. I believe there is great promise for what AI is capable of, but that there is still a lot of work to be done.

In 2016, at the International conference on Machine Learning and AI, scientists and academics all agreed that there was an alarming lack of consensus about how AI would develop, and how humans would interact with it.

It is clear that AI is going to play a key role in our future. There are already many AI related innovations that are being developed.

Just to name a few, the development of the Alexa voice assistant, the use of Google Now for reading and phone calls, the development of the Routines in Apple’s Carplay, and in set-top box Apple TV are just a few examples.

This technology promises to change the way people think about the world. The future could be either a place of personalised service, enabled by AI, or could be dominated by algorithms used to make decisions regarding many aspects of our lives.

I believe there is a real opportunity for the UK and the EU to lead the way in establishing ethical and social requirements for AI. The EU Digital Single Market is an excellent vehicle to achieve this goal, and it is important that the UK play a leading role in this.

We can also work together in order to nurture and support the UK’s AI/machine learning sector. And finally, we must work together to achieve societal benefits.

Working with others in academia, industry and government will help to make the UK the best place in the world to develop AI technology.