Would you work with a robot?

I was thrilled to attend Book Machine’s event Talking Marcomms: AI and the future of Publishing, where a panel of AI specialists and publishing professionals discussed the tech phenomenon that is AI, and how publishing can further implement it to reach new heights. A bot won’t write the next literary masterpiece, but they do have other uses in publishing! 

Luka – a robot reading companion for kids (photo from lukareads.com)

Alexa, what is AI? 

AI stands for artificial intelligence. 

Fortunately, the type of AI used in the workplace today isn’t anything you might see in an episode of Black Mirror. 

Hannah Marcus, Associate Director at Discover.ai and panelist at Talking Marcomms, says a more descriptive term for this is “machine learning”. We enter a dataset; an algorithm takes a set of rules from the dataset; and the computer learns these rules in order to produce the desired result. Pretty simple stuff right…? It is best not to think too much about how it works or your human brain might explode.

What benefits will AI bring to publishing?

ALEKS – AI programme created by McGraw Hill Education (photo from aleks.com)

Beth Hardie, Marketing Director at McGraw Hill (a major educational publisher) discussed their AI program ALEKS that developed bespoke learning solutions to kids with different ability levels in subjects such as maths. This bespoke element is what makes AI so appealing to a multifaceted industry like publishing – it could be employed in areas such as identifying market trends, finding areas of interests for commissions, editorial tasks, checking for plagiarism, collating sales trends and anything where a large enough data set can be mined. Non-fiction is a genre that can greatly benefit, Beth says.

Zeta, a customer lifecycle management marketing company, measured that personalised marketing emails increased by opening rate of up to 63.22% when created by AI. The facts cannot be ignored!

Won’t people lose their jobs?

The panel discussed how this is a major misconception, Hannah compared it to anxieties following the industrial revolution, when this period mostly led to an increase in general quality of life. The idea behind growing AI in publishing is to create a synergy of man and machine, while best making use of metadata. Got a task you find fiddly or boring? There’s a programme for that. Want more time to work on the more creative or fulfilling elements of your book marketing campaigns? AI can do it for you. 

Alright, what’s the catch?

There are some challenges in implementing AI that will be tricky to overcome for publishing. Bigger companies such as Penguin Random House have the means to pay for software or employ specialists in-house, while smaller independent publishers won’t be able to afford the investment of time and money to make AI effective, creating a gap of opportunity. 

There are ethical concerns, namely algorithmic bias, it is important to ensure the data fed to programmes is accurate and that a large number of diverse voices go into developing programmes. There is also the issue of navigating GDPR legislation, and the data of individuals must also be managed carefully. These hurdles that can be overcome, but they are there nonetheless.

The bottom line is that AI can let publishers get more familiar with their markets, enhance their productivity and ultimately make more sales, but a strong moral compass is needed and risks must be taken for success.

(Featured image from Tara Winstead from pexels.com.)