We hear a lot about how smart machine learning-based tools and platforms are going to change the world of work. Some commentators, like Stephen Hawking, have made apocalyptic prophecies about how the world as we know it will end as the machines take over, and there are no more jobs. Loath as I am to disagree with anyone so eminent, I think the picture is likely to be slightly more nuanced.
Yes, I think that machine learning and artificial intelligence will change the world of work, and will certainly make some jobs or parts of jobs redundant. But I don’t see it as a wholesale shutting down of human input. Instead, I think we are likely to see new jobs emerging. Perhaps more importantly, I think we will see much more job integration. Indeed, there are already glimpses of how this might work, as smart tools and platforms become more ubiquitous.
Defining and exploring job integration
It is worth saying a bit about what we might mean by ‘job integration’. It sounds obvious enough, that it means bringing together parts of different jobs, and combining them into a whole, to make a single new job. This is going to be not just possible, but essential, as artificial intelligence systems start to do more of the ‘grunt’ work. For example, data scientists now have to do quite a lot of the basic work around data preparation. In future, much could be done by algorithms, freeing up their time to do other work, such as governance.
Of course, it could be argued that this is just a way to cut jobs and therefore costs. And to a certain extent, that would be true, until we start to look at the second area of integration, or perhaps more accurately, partnership: between humans and algorithms. I think this is likely to be the real game-changer.
The premise here is that machines, alone, are not all that powerful. They can learn, but sometimes they learn the wrong things. You only have to think of Microsoft’s Tay to understand that. This means that AI systems, even the most intelligent, need to be overseen by humans to make sure that they stay on track, and continue to do the ‘right’ thing. This may, in some cases, mean being ready to take over when required. To avoid problems with customer experience as a result, this ‘takeover’ needs to be as seamless as possible, requiring close integration between human and machine processes.
Consider call centres. Chatbots are increasingly seen as the way to handle some of the more routine enquiries: for banks, perhaps balance enquiries, questions about overdrafts and so on. We could think of this as FAQs, but with a personal answer. But once the conversation gets more complicated, or starts to go round in circles, rapid human input is vital to avoid a poor customer experience. The key issue is to figure out what it is that algorithms cannot do, and make sure that humans are there to do them instead, at the right time.
The role of smart tools and platforms
Smart tools and platforms can make job integration significantly easier. An example makes this clearer: a combined analytics/data management platform that is shared and used across the whole organisation, which allows individuals to use their preferred programming language and also provide visual output. It sounds simple, but in one stroke, it means that there is a shared language and system across the whole organisation, from data scientists right up to the CEO, and resources and data can easily be shared. This makes governance much easier, but also means that staff at different levels can cooperate with each other to generate and use models. Staff can also move between teams and managerial levels without worrying about having to learn about new tools.
Expertise is shared, instead of siloed. Job integration is suddenly a reality, and not just a concept.
A spectrum of change
What is clear is that job integration is not a sudden necessity, brought on by the rise in machine learning. Instead, machine learning and smart platforms simply facilitate what was already happening: people working together to share tools and expertise, and get the job done. These platforms, systems and tools are just that: tools. In the same way that ploughs, and then combine harvesters, helped farmers to work more efficiently, so these tools can help us to do more, freeing up time to concentrate on more interesting problems.
Different jobs, yes. But perhaps not the end of the world.
Take a few minutes to explore the following
Open Source vs Proprietary: What organisations need to know which highlights some of the costs/challenges of open source (as well as the perceived benefits) and how a hybrid approach can give organisations the best of both worlds. The findings are based on a survey of 300 senior IT and business decision-makers across the UK and Ireland.
Download a White Paper: SAS in the Open Ecosystem
