The occurrences at OpenAI in the last six days could be described from a multitude of perspectives: a spicy boardroom drama combined with a duel between competing visions of AI.
However, the return of Sam Altman on Tuesday to OpenAI may signify that the battle appears to be over. Team Speed Up won. Team Slown Down lost.
Sam Altman, the CEO and co-founder of OpenAI, a non-profit artificial intelligence research and development company, was fired by the board on November 18, 2023. After a dramatic turn of events, with displeased investors and angry investors, and an entirely new board, Altman was rehired to lead the company. The new board will, initially, be composed of three members: Adam D’Angelo, the chief executive of Quora, Bret Taylor, former Salesforce co-CEO, and Larry Summers, former US Treasury Secretary. These members are expected to provide Altman with the support needed to revolutionize AI through current obstacles and ambitious goals.
Nevertheless, the new structure of the board has drawn criticism as the reconstituted OpenAI board will not have veto power as the old board did. Additionally, their preferences will be balanced alongside others’, similar to the company’s executives and investors. In addition, since OpenAI’s controlling shareholder is a non-profit, the company is not accountable to public shareholders. The lack of investor oversight has led to concerns about over-assertation of power, in not only OpenAI but other AI companies. For that reason, it is imperative to consider not only the benefits of AI but also its possible implications in employment and ethical dilemmas, seen in the potential increase in automation of jobs as well as discrimination within AI tools.
Employment
The worry over the capacity of OpenAI in conjunction with other AI companies has led to criticism of the growth of AI in the future. Many fear jobs being given to machines, making humans less necessary for certain tasks due to dependence on AI-driving machinery. Goldman Sachs reports, cited by BBC, that “AI could replace the equivalent of 300 million jobs”, and could “replace a quarter of work jobs in the US and Europe.” This includes, but is not limited to, jobs in the technological sector, such as software and web developers, programmers, and data scientists, with work amenable to AI technologies. AI can also affect media jobs, such as advertising, journalism, and roles involving content creation. Like media roles, jobs in the legal industry, such as paralegals and legal assistants, can be automated due to their responsibility of consuming and synthesizing large information.
Ethical Dilemmas
AI has historically perpetuated biases in society. In 2015, Google had to apologize for labeling African Americans as “gorillas”, exemplifying failures in facial recognition. To add, a study found that there was a 35% misclassification in facial recognition of darker-skinned women. These concerns are some of the many that led the Center for AI Safety, a non-profit organization, to release a statement, declaring, “Mitigating the risk of extinction from AI should be a global priority alongside other societal-scale risks such as pandemics and nuclear war.” Over 350 AI professionals signed this statement. When looking at the 2024 election, deepfakes, synthetic media that is digitically manipulated through AI, pose an ethical risk. Through the usage of deepfakes in advertisements, and in this case, election campaigns, deepfakes can cause voter manipulation and microtargetting. In fact, this has already happened in the Republican Party nomination. A University of Virginia analysis describes how Florida Gov. Ron DeSantis’ presidential campaign “War Room” released a video of former President Donald Trump hugging Dr. Anthony S. Fauci, former director of the National Institute of Allergy and Infectious Disease. DeSantis’ campaign admitted that the video was fake.
But despite this, AI has already, and will likely continue to, come to fruition in critical aspects of society that we will cover in this article: education and healthcare.
Healthcare
One of AI’s biggest potential benefits is that the algorithms can help people stay healthy so that they don’t need a doctor, or at least not as often. Technological applications and patient-driven applications enable proactive management of a healthy lifestyle. AI is also being used to detect diseases more accurately and specifically. For example, a study found that AI-supported mammograms increased breast cancer detection by 20%. The proliferation of AI-powered wearable medical devices is applied to oversee early-stage heart disease, enabling physicians to better monitor and detect potentially life-threatening episodes. More specifically, Google’s DeepMind health works with healthcare organizations and clinicians to unlock better diagnoses and real-world healthcare problems. The group that utilized AI for mammograms found an additional benefit: a reduced workload reading of 44%. The reduction of repetitive tasks such as billing, scheduling, and managing electronic health records can allow for a decline in healthcare burnout, and an increase in focus on patient care. Moreover, AI can increase access to healthcare, through telehealth, by providing patients 24/7 access to medical advice and support that they may have been barred due to geographic, language, and financial barriers.
Education
By the time the Stanford Institute for Human-Centered AI held the inaugural AI+Education Summit on February 15th, 2023, ChatGPT had over 100 million users, and 30% of college students had used it for assignments. This makes ChatGPT one of the fastest-growing applications ever adopted, causing educators to wrestle with responses to this technology, as it could hinder critical thinking skills and compromise students’ abilities to think independently. However, AI will be able to transform teaching and learning, when used appropriately. More specifically, AI technology and algorithms can increase accessibility to children with disabilities in autism, dyslexia, visual impairment, and hearing loss. For example, LuxAI, a Luxembourg-based company, has been selling the QTrobot, designed for autism. This robot uses facial expressions, gestures, and games to teach autistic children about communication, emotions, and social skills. Similar to healthcare, AI can alleviate the roots of burnout in educators, such as the tasks of grading, lesson planning, and general administration being outsourced to technology.
Seen in four critical points of our world today: employment, ethical dilemmas, healthcare, and education, AI can pose major fruition in the future, as can it cause devastating failures. AI lies in the hands of the researchers, developers, and consumers, as to how it will redefine the future for the good, or impair it.