Artificial intelligence and machine learning may be ideal for picking up the day-to-day tasks of running enterprises, but still fall flat when it comes to innovation or reacting to unforeseen or one-off events. While enterprise-grade AI is still a ways off, it’s incumbent on business and IT leaders to start piloting and exploring the advantages AI potentially offers.
That’s the word coming out of a recent report from the MIT Task Force on the Work of the Future, which looked at AI as part of a broad range of changes sweeping the employment scene and workplace. “We are a long way from AI systems that can read the news, re-plan supply chains in response to anticipated events like Brexit or trade disputes, and adapt production tasks to new sources of parts and materials,” state the report’s authors, David Autor of the National Bureau of Economic Research, along with David Mindell and Elisabeth Reynolds, both with MIT.
For starters, data – the fuel that propels AI decision-making – is not ready for the leap. Most successful AI initiatives to date are based on machine learning (ML) systems, which depend on large data sets. There are many areas where AI and related technologies show great promise for “enabling innovations in design, measurement, and materials—creating new products and new methods of production,” Autor and his co-authors point out. “Still, our interviews find many companies consider themselves at the early stages of adoption of these techniques, figuring out how to collect and structure data such that they can apply greater insights to their existing operations. Doing so requires integrating multiple data sources, often for hundreds to thousands of machines in larger companies. It requires merging expertise from both operations and information technology, while ensuring continuous improvement in the production system.”
The most prominent ML applications being pressed into service at this time include image classification, face recognition, and machine translation, as well as document analysis, customer service, and data forecasting. However, ML systems “still face challenges with respect to robustness and explicability,” Autor and his team state. “The industries that use ML are slowly learning that the data used to train ML systems must be as unbiased and trusted as the systems themselves need to be—crucial challenges in an era of hacking and cyber-warfare.
Explainability is also a challenge, as ML systems “tend to be black boxes that offer no insight into how they make their decisions,” they add. “Explainability is essential for systems that must be robust to failure, interact with humans, and aid in significant decisions with legal or life-critical implications.”
The notion of 100-percent automation is also a fallacy, as the co-authors also dismissed the notion that AI can be employed to completely automate facilities such as factories. “’Lights out’ factories, with no human input, have long been a utopian- dystopian vision for the future,” they state. “The vision may make sense for some situations where the product or process is mature and highly stable. But even the most automated electronics or assembly plants still require a large number of workers to set up, maintain, and repair production equipment. A typical mobile phone—a stable and uniform product made in very high volumes—is touched by dozens of human hands during production. “
Instead, AI and ML are being directed to “specific pain points and automating activities such as materials transport or inspection,” they write. The technologies are augmenting, but not replacing, labor. Jobs are not being eliminated, but the tasks that make up a job are being automated. “ML applies at the task level – ideally to tasks with easily measurable results – and does not fully automate particular occupations in any case of which we are aware, though all occupations have some exposure.”
For example, AI is taking on some of the heavy lifting of “workers labeling data, paralegals doing document discovery in law firms, or production workers performing quality inspection on factory lines,” Autor and his colleagues observe. “We also see cases where AI and ML tools are deployed to make existing employees more effective, by aiding call center responses, for example, or speeding document retrieval and summary. Some applications in engineering involve using AI to search physical models and design spaces to propose alternatives to human designers—enabling people to come up with entirely novel designs. “
To achieve the optimum balance between AI and human initiative, the MIT team urges organizations to “redesign workflow and rethink the division of tasks between workers and machines, akin to what occurred as Amazon deployed robotics in its warehouses. The resulting changes in work design will alter the nature of many jobs, in some cases profoundly.”
Still, there is no clear-cut answer on what impact AI and ML are having on employment prospects, they conclude. “While it seems unlikely that AI has greatly impacted the labor market so far—beyond spurring increased demand for computer and data scientists—we have no definitive evidence on this topic to date,” Autor and his co-authors state. “The implications for specific skill groups are as yet uncertain and will in part depend on managerial and organizational choices, not on technologies alone.”
Credit: Google News