Artificial Intelligence: Threat or Opportunity?

‘Humans’ has returned to our TVs, and brought with it a realisation of the fear that Artificial Intelligence (AI) will one day gain consciousness and run riot. In it we see the ‘Synths’ fighting for their rights, vulnerable in their humanity. But we also see the darker, riskier side in Niska, who distrusts humans but is more volatile in her reactions. We know she has killed, and it’s her representation of AI that we really fear.

Humans isn’t the first to bring this conscious AI danger to life; for me, the most haunting representation is Ex Machina, where (spoiler alert!) Ava successfully manipulates Caleb into falling in love with her and setting her free, then leaves him stranded. That’s where the film ends; nobody knows what she goes on to do, and our imaginations run riot. Will she destroy the planet? Usher in an age of AI overlords? Or simply blend seamlessly into human society?

When we hear the words Artificial Intelligence, this is what we picture: humanised robots taking over the world. But many people don’t associate it with the products, programmes and apps that we use every day. It’s generally recognised that Siri is a product of AI. Facial recognition in your camera and on your Facebook? That’s AI. Music, video or clothing suggestions based on your purchase or listening history? That’s AI. Jerry Kaplan, an entrepreneur from the famed Silicon Valley argues that AI has been tied to the intelligent uprising of robots, when it’s really a “continuation of longstanding efforts to automate tasks, dating back at least to the start of the industrial revolution.”

There’s a fear that AI will render entire professions redundant, putting hundreds, thousands or even millions out of jobs. However, McKinsey suggests no jobs will become redundant over the next decade due to AI, but it will affect “portions of almost all jobs” (its research involved over 2000 work activities that make up 800 job roles). In other words, don’t worry – you’re still likely to have a job, but the more laborious elements of it can be automated, allowing you to focus on adding value that only you can add. The present reality that AI offers in a business environment is a more productive workforce.

The McKinsey research found that about 60% of all occupations could be affected by automation, with 30% or more of their job role automated. While automation could affect office-based workers the most positively, it doesn’t necessarily mean those in roles that require “predictable physical work” (i.e. in a factory setting) will be replaced by machines; the cost of automating manual tasks can often outweigh the cost of workers, meaning people are more cost-effective. Additionally, automation might eventually replace some roles that require repetitive, manual work only, but could open up new jobs maintaining the ‘robots’.

Digital behemoth Google is looking to expand its AI offering, as well as gearing up to work more closely with APIs. Recently, Google announced its acquisition of API.AI. In its announcement, Google said “API.AI has a proven track record for helping developers design, build and continuously improve their conversational interfaces.” It cited well-known communication platforms Facebook Messenger, Slack and Kik as the company’s successes – certainly some big players. API.AI has achieved a lot in its 2 short years in existence.

API.AI focuses primarily on text recognition, speech recognition and natural language processing, all available in 14 languages. Its aim is to “help machines understand people better”, and it focuses on translating human requests into actionable commands. As the company name suggests, API and artificial intelligence (AI) go hand in hand – you can read more about how APIs are the glue that holds all of our favourite programmes together on our blog. Creating an artificial intelligence agent through API.AI is surprisingly easy, which illustrates how far we’ve come with AI.

Automation tools rely heavily on artificial intelligence; after all, automation is simply the use of technology to perform the mundane tasks that humans have historically had no choice but to carry out. For example, we use a tool called Qlik Sense for data analytics, and it is one of the most intelligent and innovative data tools around. Users can upload data from various sources and Qlik will bring it all together consistently. Qlik Sense is also proof that it isn’t only low-paid, low-skilled jobs that will be impacted by automation. Its standout feature is that it will “spot hidden patterns and trends” in your data that you might not have otherwise identified. It has automated a process that only skilled data analysts could realistically carry out, yet it doesn’t make the data analyst unnecessary – it adds extra value to their role and frees up their time for what really matters. Time is also saved because developers and analysts no longer need to recode the data to present different views; the developer can now spend their time coding new website functionality, for example.

In a similar vein, it’s expected that around 80% of data entry and processing activities can be automated. That doesn’t spell the end for roles in that area; most office-based roles will involve some level of data entry or processing. For example, marketing often requires data input in the form of logging new customers. With automation, that process happens automatically thanks to intelligent data mapping. This is good news for many businesses, for whom data is more prevalent and important than ever.

It’s not just logging financial or customer data that can benefit from automation. TermSet is a new cloud-based solution that scans your documents for keywords, and applies taxonomy and tags to your files intelligently based on the content. This is done through a technology called Natural Language Processing – much like API.AI. This means your workers don’t need to spend time organising, sharing, tagging or duplicating documents – they’re all highly searchable and they’re effectively categorised. No more dithering over whether a document needs to sit with campaigns or events; now it can do both!

So how do you make the most of AI and automation in your business? It’s recommended that first, you document all of your business processes to gain an understanding of what activities your workers are doing on a daily basis. It’s important to consider “technical feasibility”, according to McKinsey; i.e., what can you realistically automate? This will likely include manual data tasks, sharing information, and many other tasks that don’t require “thought, judgment or cognition”, according to Redwood.

Once you’ve decided which processes you can automate, you need to consider cost. Do you need new hardware or software? What is the cost of developing and deploying those programmes? If you’re on Office 365 or you’re thinking of upgrading soon, you’ll get automation tool Flow in the cost of your subscription; automation doesn’t always have to be expensive. Or, if you’re looking for a powerful and sophisticated automation tool, Nintex is the answer. Its innovative drag-and-drop functionality allows you to create workflows that are as simple or as complicated as you like. You can automate something basic like email replies, or end-to-end processes like logging new customer queries and assigning them to relevant colleagues, and even company-wide processes. Sign up to our Nintex events and bring us your most challenging workflow; we’re confident that we’ll be able to solve it using Nintex.

Additional considerations include the maintenance of the automation software; will your existing people oversee it, or will you need to hire staff knowledgeable about automation tools? Human considerations are also important, not just in terms of labour costs and job roles, but acceptance. There’s been a lot of talk of robot nurses (and little robot dogs) being deployed to look after the elderly. Just because it’s possible, it doesn’t mean you should do it – some roles, like nurses, should always be human. But that doesn’t mean that you can’t automate the systems that host the patient data so nurses can spend less time updating them.

Using AI to automate laborious and time-consuming tasks could make your workforce significantly more productive, without making any job roles redundant – at present. You could think of it as increasing the ROI of your existing staff – with less time spent on manual, repetitive tasks, they can put their minds to adding value that only humans can through creativity and learned skills and knowledge. If you haven’t taken the first step on the journey to automation, now is the time. You’ll need to spend time mapping your business processes and identify what could be automated, but the benefits are potentially enormous.