Turing, AI, and modern recruitment: how is it all connected?

The recruitment profession involves a lot of communication and routine. Perhaps it’s time to ‘invite artificial intelligence (AI) to an interview’. But to keep expectations from being too high, you need to listen not only to marketers, but also… to mathematicians and IT specialists. This will help you predict the outcome of ‘cooperation’ with AI in recruitment.

So, can a machine think like a human? After the Second World War, cryptographer and computer scientist Alan Turing proposed a test to determine whether a machine could demonstrate intelligence and behaviour that could not be distinguished from human beings.

Later, Eliza was created, a virtual companion developed by scientist Joseph Weizenbaum. She imitated the behaviour of a psychotherapist, getting people to talk more about themselves and then asking clarifying questions. Moreover, this first chatbot was better at convincing people that it was a human than the GPT-3.5 chatbot of 2022.

Other chatbots could not pass the test ‘honestly’ because they were unable to find and switch to a new topic, as well as lacking a personality.

‘Rise of the Machines’ is cancelled for now

After the Turing test, experts tried to test machines for creativity, visual perception, and even a sense of humour. The problem that everyone involved in AI faced was that machines were not able to encompass intelligence from all sides. They could perform one specific task.

In most cases, this situation still exists today, so it is important to distinguish between the following areas of AI. Among them:

Expert systems. They appeared back in the 60s and operate on the principle of ‘upload data – get results’, which partially replaces an expert in quite specific areas (medical, legal, financial). The main problem is that they need a carefully created database. And there were few highly specialised experts who could devote time to this. In addition, they did not always enter commonly known information, which led to errors.

Machine Learning also has different manifestations. For example:

• Computer Vision – computer vision, image analysis and search for matches or specific objects. It is used in video surveillance cameras, for recognising number plates on vehicles, face ID in phones, and self-driving cars.
• Natural language processing– speech recognition by ear, machine translation, classification of texts into thematic categories. It is used in voice assistants such as Siri, Alice, and Google Assistant.
• Predictive analytics– building forecasts based on available data, assisting in business planning. For example, Walmart has long been using a system that receives information from cash registers and predicts which products will be in demand and which will not.

Neural networks. Machine Learning cannot be considered outside the context of neural networks. A neural network can take in and ‘react’ to certain information or actions. For example, we show it a picture of a dog, and it tells us whether it is a dog based on certain data collected by machine learning.

It may seem like the possibilities of neural networks and machine learning are endless. We can upload photos of all human emotions to them and know who is lying and who is not. Or let’s introduce a video interview system that will evaluate candidates, compare them, and choose the best ones. But… a neural network will give an average of the analysed data. It doesn’t have the intuition or humanity to come up with special meanings. Therefore, there is always a risk of an erroneous conclusion.

For example, at one time, there was a fuss about Amazon’s recruitment system, which selected men more often than women. Therefore, it is a misconception to say that any area of AI can replace a recruiter. But it can make their work easier, more efficient, and faster, especially the routine part of it.

How AI elements can strengthen a recruiter

Among the tasks that AI can help with, sourcing, working with candidates at the initial stage, stands out. These are also routine operations that are often repeated and performed manually.

For example, writing job descriptions and working with creatives. For this, recruiters turn to the same virtual assistant with generative AI – ChatGPT. It can also handle the preparation of questions for interviews and screenings, help with the simulation of technical interviews, the creation and evaluation of test tasks, checklists for skill checks, etc.

In turn, automated recruitment software with built-in AI elements can perform more strategic tasks. This may include labour market analysis, selection of candidates according to specified criteria, increasing search speed, processing large amounts of data (especially in mass hiring), optimising selection, automated screening of candidate profiles and resumes, etc.

With the help of AI, a recruiter can communicate better with a large number of candidates, and has the potential to receive good recommendations for interviews, hiring, and assessing the suitability of candidates for vacancies. So, in the end, we see this potential:

• speeds up recruitment and reduces recruitment costs,
• the quality of selection is improved by taking into account new important factors for a large number of candidates,
• objective assessment of candidates based on clear criteria (without bias and human factor),
• automation of routine tasks to focus on more strategic ones.

AI in recruitment: successful cases

With the advent of AI-powered tools, classic recruitment has not only changed, but transformed into a new, adaptive and modern format. Recruiters and recruitment agencies use AI assistants not only to search for candidates but also to format CVs, roadmapping, and track the entire process.

Hilton. One of the first successful projects to apply AI in recruitment is the Hilton hotel company, which has long developed a chatbot that can answer candidates’ questions, schedule interviews, and provide personalised feedback. Hilton also uses AI to analyse candidate data and predict their suitability for the position. Thanks to this, the company has reduced the time for hiring several times.

Amazon: After the scandal of gender bias in recruitment, Amazon assures that new AI models in their recruitment are protected from such shortcomings, and a new automated candidate assessment system is already in place. It examines candidates’ CVs, comparing their achievements and background with the profiles of Amazon employees in similar positions, and issues recommendations.

Delta Air Lines. The American airline Delta Air Lines also actively uses AI to improve the candidate experience and simplify hiring. The company has developed a chatbot that is able to respond to applicants’ queries and provide individual feedback. In addition, Delta Air Lines uses AI technology to identify possible gaps in employees’ skills, as well as to provide the most personalised training and development.

Instead of conclusions

The use of AI elements in recruitment opens up many opportunities to speed up and facilitate routine tasks, improve the objectivity and quality of selection, and reduce recruitment costs. Another significant advantage is the ability to use AI at all stages, from job description to onboarding. This helps recruiters focus on making informed decisions, human communication, and more strategic tasks. However, it is important to maintain a balance between automation and human feelings, and then the effectiveness of AI-enhanced recruitment will be quite achievable and predictable.

Its implemented elements allow you to significantly reduce the time required to select candidates by automating routine and repetitive operations and focusing on the most promising candidates.

Very soon, we will share with you even more positive and even revolutionary information regarding the implementation of AI elements in our software.