Subscribe

The machines are here

Artificial Intelligence (AI) is set to change the world. The question isn't when, but how...

By Tamsin Oxford
Johannesburg, 12 Jul 2017
Nick Agyros, GotBot.
Nick Agyros, GotBot.

Earlier in 2017, there was a conversation taking place between Elon Musk and Demis Hassabis. It was a conversation recorded by Wired magazine and it highlighted the stark contrasts around how people and organisations view Artificial Intelligence (AI). Hassabis is the co-founder of DeepMind, a Google-owned laboratory in London that focuses on algorithms and AI. Musk is seeking space travel and is concerned about the consequences of AI. One may well be needed to escape the other.

"Most of the fears the market has regarding AI are fed by mainstream media, which portrays the common scenario of machines taking over," says Nick Agyros, director of GotBot. "AI today, however, stands to revolutionise the business sector with opportunities for customer personalisation and communication in real-time, 24 hours a day. We are currently seeing huge interest and research going into AI and its capabilities across different industries."

One thing that scientists, think tanks, experts and technology gurus all believe is that AI is set to change the world. The applications for AI are limitless. Think bots that communicate endlessly with consumers on numerous platforms; think Alexa, Amazon's voice assistant that is now being further developed through the Alexa Fund in an attempt to win the AI race; think Facebook and its FAIR (Facebook AI Research) programme that is working with everyone from developers to academic institutions to find the answers to true AI; think the future and how it is here, man.

"AI can no longer be considered the stuff of science fiction," says Vaughan Mason, senior manager, Technology Advisory D&A lead, KPMG South Africa.

"If you grew up with Star Trek, you would remember Star Fleet officers using a Tricorder device to scout unfamiliar terrain, provide detailed analysis of star ship data or medical diagnostic data. Today, you could quite credibly replicate most of these functions using any high-end smartphone," says Hans Zachar, Technology MD at Accenture.

AI transformation

There has never been so much data at the proverbial fingertip. There has never been greater internal and external pressure to analyse this data to reduce risk, ensure compliance, drive efficiencies and, perhaps most importantly, engage consumers. AI is an opportunity that business leaders cannot afford to ignore. Already there are organisations deploying the technology to deliver highly complex processes. Of course, they are equally being used for the less glamorous tasks - the repetitive and standardised processes that define tedium, but require accuracy.

"There is a good chance that AI and machine learning will change the 80/20 rule: the idea that 80% of effects come from 20% of causes," says Simon Robinson, owner of Suncircle Group. "With access to a great variety of data, algorithms will get smarter and process complex patterns in smarter ways. AI will drive the advantage, and the early adopters will lead the race. When viewed for what it is - an indispensable tool for supporting humans in virtually every aspect of life - it is clear that AI is transforming business."

The AI that's relevant to the business is being used by Apple and Google to power their virtual assistants; by Facebook and Google Photos to help identify and tag people, places and things; in videogames, to make characters more realistic and personalities more engaging; in healthcare to assist doctors in making accurate diagnoses.

"AI is able to do things like learn, hold conversations, make decisions and complete complex tasks. These abilities are central to creating systems that can augment human intelligence and eventually help us with everything from cooking and doing laundry to driving and performing lifesaving tasks in an operating room," says Cliff de Wit, chief innovation officer at Microsoft South Africa.

Hello, Dave

If you're not convinced that the development of AI is placing humanity on the cusp of digital slavery, then this is all very good news indeed. The thing is, AI isn't actually even at the point of holding an intelligible conversation, much less massing its giant intelligence for an incursion.

AI can no longer be considered the stuff of science fiction - Vaughan Mason, KPMG South Africa.

Says Dale Connock, chief risk officer at Strate: "The common definition of AI has become horribly tainted. People are using AI to describe a broad spectrum of computer-based activity that is clearly muddying the waters. Much of what we are encountering today really fits into the category of tech that's designed to perform a very narrow task, something it can do repetitively at a more effective rate than a human could."

The reality is that the traditional concept of AI as a machine that can learn and think isn't on the horizon. It is on the agenda, and it's this that concerns Musk, Wozniak and Hawkings, et al. It's the idea that humanity creates a system that could outsmart us, very likely to our detriment.

"There is a potential dark side behind every new disruptive technology - we are certainly in uncharted territory with AI - and it's clear that the AI train is moving, and it's a get on or miss out situation," says Zachar. "It's ill advised to rush headlong into these new technologies without carefully considering the business and even ethical and moral implications that AI can bring."

Google, Facebook, Microsoft, IBM and many other such impressive names are joining forces across global and competitive borders. Their goal is to create ethical frameworks within which they operate and to remain aware of the potential pitfalls and negative impacts that AI may bring. It is a bold future, but it is one that can offer both businesses and people opportunities that neither knew were possible.

Looking long-term

The inability of today's AI to replicate the creepy Mother of Alien remains a balm to many a concerned mind, but what of tomorrow? What are the long-term applications of AI for the business? We ask some of the experts...

Claus Jepsen, chief architect, Unit4, says: "The convergence of conversational user experiences with augmented solutions that self-learn will fundamentally change how enterprise software serves people. The possibilities for delivering new enterprise software experiences are exciting. For the first time, we can create true self-driving software."

For Tom Wells, executive director, Synthesis Software Technologies, the real long-term value may lie in the ability to experiment with ideas, and test those within the sandbox of the AI model. It can support the business in removing a lot of guesswork around product development without impacting on the brand or its budget.

Ryan Falkenberg, co-CEO, Clevva, adds: "AI will fundamentally change how business functions, especially the role of people in business. With the onset of the new economy, we are moving away from the simple capture and management of information (decision-making maps) to the capture and scaling of logic. With AI, we can fundamentally change how we support people in how they perform their jobs. Businesses should apply AI where it can add value - help existing staff perform better, use robotic process automation to remove basic administrative data capture, and leverage data to move towards improved customer service through cognitive computing."

Dario Fanucchi, co-founder, Isazi Consulting, adds, that most tasks that require repetitive work will be automated in the future so the business will look to the employee for creative thinking and decision-making, and the small company will start making big differences thanks to the ubiquity of AI.

This article was first published in the July 2017 edition of ITWeb Brainstorm magazine. To read more, go to the Brainstorm website.

Share