The use of AI to increase the efficiency of mental workers
The new capabilities of AI, which is able to recognize the context, concepts, and meaning of concepts, open up new, sometimes unexpected, ways of working together between mental workers and machines. Experts are able to contribute to training, quality control and fine-tuning the results of AI. Machines can complement the knowledge of their fellow humans and sometimes help educate new experts. These systems, more likely to mimic the human mind, are more reliable than their data-dependent predecessors. And they can have a significant impact on mental workers, accounting for 48% of the workforce.in the United States - and over 230 million brainworkers worldwide. But to take full advantage of this smarter AI, companies will need to redefine the workflow and jobs.
Brain workers - people who make decisions at work, reason, create and apply ideas in cognitive processes not related to routine - for the most part agree with this. Of the more than 150 experts taken from a large international study on the use of AI in large companies, almost 60% say that their old job descriptions are quickly becoming obsolete due to their new collaboration with AI. About 70% say they need retraining due to new requirements related to working with AI. 85% agree that senior management should also participate in common attempts to change roles and processes of mental workers. And starting to fulfill the task of rethinking the use of mental labor in combination with AI, they can apply some of the following principles:
Let expert people tell AI what is important to them. Take medical diagnostics, in which AI is likely to be everywhere. Often, when the AI issues a diagnosis, the logic of the algorithm is not clear to the doctor, who must somehow explain the solution to the patient - this is a black box problem. Now, Google Brain has developed a system that can open a black box and translate its work into human language. For example, a doctor evaluating the diagnosis of AI “cancer” would like to know to what extent the model took into account various important factors - the patient’s age, previous chemotherapy, etc.
The tool from Google also allows medical experts to introduce into the system concepts that they consider important and test their own hypotheses. For example, an expert may want to see if the diagnosis will change the introduction of a new factor previously unaccounted for by the system. Bin Kim, the developer of this system, says: “In many cases, when working with applications on which a lot depends, experts in a particular field already have a list of concepts that are important to them. We at Google Brain are constantly confronted with this in the medical applications of AI. They don’t need a set of concepts - they want to provide models of concepts that are interesting to them. ”
Make models that match common sense. With the accumulation of concerns about cybersecurity, organizations are increasingly using data collection tools at different points in the network to analyze threats. However, many of these data-based techniques do not integrate data from multiple sources. They also do not include the common sense of cybersecurity experts, who understand the spectrum and variety of motives of attackers, understand typical internal and external threats and the degree of risk for the enterprise.
Researchers at the Alan Turing Institute , a British state institute that studies the science of data and AI, are trying to change this. Their approach uses a Bayesian model- a probabilistic analysis method that takes into account the complex interdependence of risk factors and combines data with estimates. In the cybersecurity of enterprise networks, among these complex factors there are a large number of devices connected to the network, and their types, and the knowledge of the organization’s experts about hackers, risk, and much more. And although many AI-based cybersecurity systems include the ability to make decisions at the last stage, researchers at the institute are looking for ways to incorporate expertise at all levels of the system. For example, an expert’s knowledge of the motivation and behavior associated with an attack through IP theft — and how they differ, say, from a DDOS attack — are directly programmed into the system from the very beginning.
Use AI to help beginners become recognized experts. AI is able to quickly turn beginners into experts. HP demonstrated this by using a cognitive computer platform from the AI lab to analyze customer call data over two years. The call center used a queue-based system to distribute calls, which caused customers to wait a long time for an answer, and the quality of user support was poor. The cognitive computer platform was able to identify the unique "micro-skills" of each specialist - knowledge about the special types of user requests obtained from previous calls. Now it is used to redirect calls to agents who successfully handled similar situations earlier. As a result, the support center improved by 40% indicators for resolving the situation on the first call,
With the training of AI support specialists, they automatically update their knowledge, eliminating the need to do this manually in their profile. Moreover, the more knowledge a specialist receives, the more complex tasks the software redirects him. Meanwhile, software is constantly improving its knowledge, and AI conclusions about micro-skills improve the efficiency with which an expert trains software. It is worth noting that several companies are working on this retraining task; for example, ASAPP startup provides real-time offerings to service support specialists.
Use AI technology that effectively uses these to mark up your expert workflow. Experts in many types of knowledge are quite rare, and do not produce a large amount of data suitable for training. But deep learning and machine learning, on which many breakthroughs in the field of AI are based, require a huge amount of data to build systems from the bottom up. In the future, we will see more systems created from top to bottom, which will require much less data to create and train, which will allow them to perceive and take into account the special knowledge of workers.
Take a recent competitionorganized by the medical imaging laboratory at the University Hospital of Brest and the Faculty of Medicine and Telecommunications in Brittany. Participants competed in the greatest accuracy of medical imaging systems, which were supposed to report which instruments the surgeon used at each moment in a minimally invasive cataract operation. The winner was a machine vision system, which was trained for six weeks with just 50 videos of surgical operations, 48 of which were conducted by experienced surgeons, one was a surgeon with one year of experience, and one was an intern. Accurate tool recognition systems allow doctors to analyze surgical operations in detail and look for ways to improve them. Such systems can potentially be used in generating reports,
All of these examples indicate that engineers and pioneers in various disciplines are developing AIs that can be easier to train and evaluate, and also include extremely valuable and often rare expert experience. To begin to take advantage of these new capabilities, organizations need to review their AI budgets. And in order to get the most out of both these systems and of mental workers, they need to reconsider the interaction of specialists and machines. As today's MO systems complement the capabilities of ordinary workers, tomorrow's systems will raise the effectiveness of knowledge workers to previously unattainable heights of universal perfection