Eps 1: Explained: How AI got everyone fired
— Forklart
| Host image: | StyleGAN neural net |
|---|---|
| Content creation: | GPT-3.5, |
Host
Delores Steeves
Podcast Content
The companys experimental hiring tool uses AI to assign a job candidates rating between one and five stars -- a lot like how shoppers rank products on Amazon, some of those individuals said. It has since been used to screen job applicants for roles from software engineering engineers to tech program managers, opening up the possibility for its eventual widespread use throughout Amazon. Amazon is now using a much-watered-down version of the companys experimental hiring tool to help it do some of its most basic work, including scraping duplicate applicant profiles from databases, said a person with knowledge of the effort.
Amazon built the artificial intelligence hiring tech in the mid-2010s, but stopped using its system after it showed a bias toward women. AI has been increasingly used for hiring across industries over recent years, but questions persist over its role in instilling or amplifying biases that can arise during hiring processes.
About 55% of American HR managers said artificial intelligence, or AI, will become a common part of their job within the next five years, according to a 2017 survey from CareerBuilder, a talent software company. In an economy in which data is changing the way companies create value--and compete--experts project that using artificial intelligence on a wide scale will add up to $15.7 trillion to the worlds economy by 2030.
The world is on the verge of transforming many industries with artificial intelligence, but how AI systems are developed needs to be better understood because of the significant implications that these technologies will have for society at large. Our main task as businesspeople is to foresee what artificial intelligence means with respect to how humans think and operate, and to work towards the ambitious, strategic integration of the new technologies in our organizations.
As AI transforms how companies operate, many people think who does this work will also transform--organizations will start replacing human employees with smart machines. According to the research of MITs CCI, we are a long way away from artificial intelligence being at the level of human intelligence, and theoretically, AI could potentially replace human employees. Rather than calling for human jobs to disappear, the study predicted AI would still spur significant innovation, leading to new jobs.
An important piece in a really smart kind of Future of Intelligent Work, though, means we are really expanding a workforce, with humans and machines alike becoming a part of it, but one designed to enhance human quality of life, as well as being more efficient at doing our jobs. In theory, automation and AI are supposed to liberate humans from dangerous or tedious tasks, so that they can engage in more intellectually stimulating tasks, making companies more productive and raising workers wages. To some, AI is technology that will augment, not replace, workers.
IBM and the tech firm LivePerson claim that in building AI, they are freeing humans up to perform more complex tasks. Amazon has invested heavily over the years to try and automate various types of jobs. Amazon has continued exploring other ways of automating Amazons warehouses, as well as the introduction of new robots, partly because the company has been rotating through so many first-line workers that at times it has worried about running out of employees in certain U.S. regions.
The company, which employs over 1.4 million workers, denies its algorithms are used exclusively for firing workers. Now, Recode has seen the classified internal documents, raising the question of whether the companys new AI tech, which it began testing last year, is the kind of thing that could someday replace a few hundred of its human recruiters.
Amy Loomis, a research director at market research firm IDCs Worldwide Future of Work research service, said that although using AI/ML for hiring and firing workers may sound surprising, it is pretty widespread today, in circles, in greater and lesser degrees. A recent report by the market research firm IDC predicted that by 2024, 80% of global-scale companies would be using digital managers empowered with AI/ML to hire, fire, and train workers for jobs measured in continual improvement -- but that only 1 in 5 companies would realize real value from this transition, with no human input. The use of Artificial Intelligence and Machine Learning has boomed at a time when managers and leaders are adjusting to mixed-mode work environments and struggling to monitor telecommuters.
Humans possess generalized intelligence, including problem-solving, abstract reasoning, and critical-judgment skills, that will continue to be valuable in the commercial world. For humans to harmonize human and machine strengths and weaknesses, they will have to educate themselves on how AI works, what it can be used for, and to determine--through the judging capabilities of their true intelligence--how best it can be used to promote work serving human interests. When robots are as intelligent and as capable as humans, humans will have little left to do, as machines will be as powerful and intelligent as humans.
If, for instance, a business requires applicants with college degrees, then algorithms will automatically exclude them from a potential list of applicants, even if their work experience would qualify them to work.
The assessment is made possible because the computer programs are trained to review applicants, by looking at patterns in the resumes submitted to Amazon over the course of 10 years.