As any company nowadays knows, artificial intelligence (AI) is a tool you need to leverage if you want to move your business into the future. In 2014, skeeled was already way ahead of the game, which might sound good but it doesn't come without its setbacks as well.
AI has been called the future of technology for decades now, but that future is yet to come. Although the advancements made in the past years have been exponential, the human mind needs time to adjust to this reality as well.
A mere 200 years ago the industrial revolution was coming to an end and now we are on the brink of a new revolution, which might displace many individuals from the biggest human capital heavy industries in the world (e.g. drivers and factory workers). As such, it is important to understand that coming too early to market is as much of an issue as coming too late.
The start
I came back to this 2014 HBR article many times over the years, where it is stated and I quote:
Our analysis of 17 studies of applicant evaluations shows that a simple equation outperforms human decisions by at least 25%.
If you think 6 years back, to 2014, to be a truly AI-powered company was a title only big players such as IBM, Facebook and Google could claim. At the same time, skeeled was already dreaming of automated matching and ranking, based on human-defined criteria, which a trained AI tool could learn and manage on its own.
But this was just the start. After the groundwork had been laid down, skeeled could move on to more exciting tools, such as a talent pool, taking the real power of data and the speed of an automated matching to provide real-time suggestions to recruiters, saving them hours of grunt work with one click.
Even small developments using natural language processing (NLP) can save time. The time a recruiter would spend sifting through hundreds of occupations and education standards, such as ONET, ESCO and ISCED, to find the ones that are best adjusted to the profile they are looking for, can now be automated and trained in the context of our wider knowledge base, to be done alone.
One of the trends we see today that will propel AI to the next level, is Transparent and Responsible AI.
The setbacks
However, the adoption of AI in recruitment still faces some resistance from those who don’t understand how it really works. The first setback I've encountered was in a meeting with individuals that are simply afraid of AI. It is hard for me as a tech guy to explain, in an “elevator pitch” kind of speech, why there's no need to be afraid, and that AI can only save you the mindless work, letting you focus on more human tasks. Ideally, it wouldn't be a matter of trust. AI can be leveraged in a way that will show you its thought process.
And that is where AI is going to win the hearts and brains of people. No one trusts robots, and soon, there will be no need to, given that many companies like skeeled are developing ways of providing you feedback on the automated decisions taken by AI, in order to promote its transparent and responsible use.
We do understand the importance of unbiased and unprejudiced decision-making, and we simply don't feed our algorithm with race, gender, age, religion or any data that isn't objectively necessary for the assessment of a person's skills, soft or hard.
The second major setback I have had to deal with came from an experience I had not long ago, where a mostly non-technical person asked me if we used neural networks (NN), and the answer is no, we don't currently do. The reaction was instant and bold: “Then it's not AI.”
I also do understand the eagerness of achieving fully automated AI, sometimes described as level 5 or general AI, and building a machine able to pass Turing's test, making it indistinguishable from a human. But I have to disagree with such an assertive comment though.
AI has many components, such as machine learning, NLP and vision. Some lean on others to achieve a common goal, as for example NLP can be supported by optical character recognition (vision) to develop a truly remarkable real-time translation system (e.g. Google Lens' Real-Time Translation).
The future
As for the future of AI, or the general direction this technology is going, I think it is having a similar path to what computer use had. Most jobs in any industry 50 years ago didn't require knowledge of how to use a computer, and now, most do. From office workers, mechanics, construction, medicine, everyone is using a computer nowadays.
AI is already infiltrating into all these areas, with smaller or bigger impact. In the future, there won't exist a distinction between software using AI or not using AI, as it will be just another tool in the engineers’ belt to improve the performance and accuracy of the software.
The future of AI lies in using every known technique collectively, and making them interact with each other in a way that makes it seamless and at the same time transparent.
It is not an easy task, but so much has been achieved in the 200 years since the industrial revolution, that maybe in 200 more we will get there.
This text is part of an amazing e-book by skeeled that I definitely recommend you check! How AI is Impacting the Recruitment Process.
Also, we are hiring Software Engineers for our R&D team, feel free to follow the process on that link, it'll be 100% online! Be safe #stayhome
More notes on engineering, AI and the systems around the work.
Back to all writing