Augmented Intelligence (IA)
Several decision-making factors limit us today: speed accuracy trade-off, knowledge-base, cognitive-emotional interactions and prediction.
Speed Accuracy Trade-Off (SAT)
Decisions often require a trade-off between speed and accuracy (SAT)[1].When faced with a task, such as reading CVs under a time constraint the human brain struggles- fast decisions are more error prone while careful decisions take longer. According to BeHiring, companies receive applications within 200 seconds after a job is posted, and an average of 250 CVs are received for each job position[4]. This leads to one of the biggest challenges recruiters must face. Resume Overload. For instance, Hays Recruitment receives about 30,000 CVs daily, and Google take in 100,000 job applications monthly. This demand requires recruiters to read through CVs quickly as competition for talent increases. In 2012, TheLadders conducted the first ever formal, quantitative study of recruiters’ on-the-job behaviour, where eye-tracking was used for 10 weeks concluding that recruiters only spend a mere 6 seconds reviewing a candidates CV. In 2015, research shows that just 8.8 seconds is spent studying any one person’s CV in a process that has become ‘Tinderised’. Nevertheless, quantity is affecting speed, thus trumping accuracy.
Knowledge Base
Laurence Kim Peek was born in 1951 with an extraordinary gift- the ability to memorise things from the age of one (inspiring the movie Rain Man). He had an exceptional recall due to his unusual neural connections. Laurence could read any book and never forget it unlike ‘normal’ human information storage capability. The human brain consists of one billion neurons with a trillion connections. These neurons communicate with one another to recall and store information, resulting in a 2.5 petabytes (or a million gigabytes) human brain’s memory storage capacity[5], in comparison to a computer or Laurence Peek. A recruiter does not have the unlimited knowledge proficiency to know everything on a CV and determine whether there are any errors. ResumeDoctors.com found that just below 50% of CVs included inaccurate information with regards to candidates including education. With the Big Data hype taking a stance, the realm of information is changing and growing at an alarming rate; with the immense volume, velocity and variety of data it would be impossible to retrieve such information to make the right choice.
Emotional Interactions
Logical, rational calculations form the basis of sound decisions but emotion is one neurological factor that interferes with decision making. If emotion is fundamental to a task then it is beneficial, but if an emotion is unrelated to a task, such as shortlisting candidates then it can be disruptive to the candidate shortlisting process. Personal preferences can impede in hiring decisions, as bad hiring choices are a result of unconscious biases[6]. In 2012, the EEOC (U.S. Equal Employment Opportunity Commission) received nearly 100,000 cases of discrimination in hiring [7]. Recruiters and hiring managers are subconsciously drawn to candidates that are similar to them, rather than those that of right cultural fit or those with different experiences and abilities to bring to a company. One key difference between a computer and person is the ability to eliminate ‘unconscious prejudice’ and be rational by trying to maximise benefits and minimise cost.
PREDICTION
A study published in Nature stated that ‘brain activity predicts decisions before they are consciously made’- ten seconds to be precise[8]. This implies that some choices are made in advance and our brain is constantly trying to make predictions about future events. The human brain is always attempting to reduce the discrepancies between expectation and actual experience, i.e. by reducing the prediction error, though in recruitment some hiring decisions and their future implications cannot be accounted for. This high uncertainty is the reason for firm’s reluctance to hire or open job vacancies. Furthermore, one in four employers stated they weren’t sure why they made a bad hire and said sometimes you just make a mistake[9]. This comes down to our inability to predict how hiring a specific candidate is likely to affect employee turnover, company vision and values etc.
Closing Remarks
Did you notice something? Machine don't have any of these limitations:
Machines can read any text at a fraction of the speed and store all this information; Big Data
A machine doesn't have emotion (yet), therefore not clouding judgement
A machine has a degree of freedom to potentially explore the algorithmic search space (via heuristics) and probabilistically compute the possible scenarios and outputs certain recruitments would have locally and globally on a business
The troubles we are facing today in making efficient and optimised choices are beyond a neural capability. As IBM Watson states 'the entity that’s going to solve the problem is the interaction of humans and machines working together to make an integrated intelligence'. NASA confirmed that allocating roles and functions between humans and machines is essential to defining an effective architecture[10]. Through technology we can make data-driven approaches and thus intelligent decision-making. I believe the new ecosystem will capitalise on the strengths of humans and machines while compensating for weaknesses. As you recently seen Elon Musk say "Over time I think we will probably see a closer merger of biological intelligence and digital intelligence.” He added that “it's mostly about the bandwidth, the speed of the connection between your brain and the digital version of yourself, particularly output."
Being a computational neuroscientist and spending several years building a cognition/AI startup with the aim of replicating human decision-making, I have learnt the potential is huge: fintech, space, recruitment, healthcare and more. Where does this take us? Welcome to my world: Cognition!