Ever since conversational AI chatbots and other forms of generative artificial intelligence shot to fame in the latter half of 2022, there has been an increasing amount of concern, suspicion, and even fear surrounding the impending boom of AI technology. Needless to say, there have been discussions over a pause in AI development and the furthering of this technology. Led by demands and letters signed by leading tech entrepreneurs like Elon Musk and several other individuals joining in, the clamor around the temporary cessation of AI development has grown louder. However, it is important to assess if putting a stop to the development of technology is even possible and viable. Apart from economic motivators, scientific advancements such as the promotion and active production of advanced artificial intelligence also require tangible reasons based on facts if a pause is indeed what is necessary.
Despite the numerous concerns and worries that surround the current pace and nature of AI technologies, understanding whether or not it is practical to pause an entire industry’s progress is key, alongside decoding why there is concern surrounding future developments, to begin with. While certain sectors like education have put forth legitimate concerns surrounding the dangers of AI to both student learning and critical thinking, there has also been a considerable amount of fearmongering and speculation among the masses, especially around concerns of automation and economic setbacks. The below portions of this article discuss whether or not it is prudent to impose restrictions on the further proliferation of artificial intelligence technologies, while also trying to shed light on the existing concerns that have warranted such demands.
Why are the Risks of Artificial Intelligence Being Outlined?
Speculative risks often take center stage despite there being several real-world concerns related to artificial intelligence.
While the release of conversational and other AI chatbots saw considerable backlash from educational institutions and pedagogical experts, the recent letter, supposedly signed by entrepreneurs and industry experts, cites several risks to society. To understand them, a brief background on the development of AI is essential. Artificial intelligence and its development began right with the creation of the first programmable computer in the early 1940s during the period of the second world war. As computing became a discipline of its own, artificial intelligence, based on the abstract concepts of human thought and mathematical reasoning, slowly became established by the latter half of the 1950s. Subsequent efforts at developing artificial intelligence saw periods of both progress and decline, given that the construction of intelligent machines was far more complex than initially thought. However, successful implementations of machine learning algorithms in both the academic and business space rekindled interest in advanced AI in the early years of the current century. The recent successful testing of conversational AI chatbots like ChatGPT has further shed light on the success of deep learning and language models that have made it possible for machines to understand human language through statistical association and processing. That being said, AI has also had numerous detractors since its initial stages of development, with several risks being associated with its proliferation.
The open letter calling for a pause on AI development was made possible by the Future of Life Institute, a non-profit organization that bases its focus on mitigating risks that arise from modern and transformative technologies. Primarily, the letter emphasizes that AI development must be taken to the next step only after it becomes clear that their effects remain positive, with manageable risks. Specifically mentioning GPT-4, the institute’s letter called for a pause on the development of AI more powerful than the most recent iteration of the GPT series. Apart from the frequently-mentioned risks of automation arising from AI, it also alleges that the risks of artificial intelligence include outnumbering, outclassing, and making humans obsolete. Fairly speculative, it cites several research papers to draw support for its claims and suggests AI development be paused for six months to assess the progression of technology and to evaluate the dangers of AI. The primary motivation for the letter has been the growing race to develop conversational and generative artificial intelligence in what is clearly a rapidly growing market.
An AI Future: The Practicality of Pausing Development
Pausing the progression of technology is never a straightforward process.
Despite having several famous names as signatories on its roster, the seemingly compelling open letter has drawn a considerable degree of flak, even from the researchers the letter has cited. Primarily, criticism has come due to the alleged disingenuity that has motivated the letter, often obfuscating the cited research material to fit the agendas of the non-profit organization. Following this, several signatories also took back their signatures, making the intent behind the letter dubious. Most quoted researchers that have come out against the attempt to pause AI’s development have stated their concerns with growing artificial intelligence technologies to be entirely different from those mentioned. Risks such as AI bias, threats to academic integrity, racism being elicited by artificial intelligence, and offensive content are more pressing issues to the development and sustenance of AI when compared to speculative risks. Researchers also cited more immediate problems such as AI’s influence on decisions related to climate emergencies, war, and existential hazards that are of greater importance than those being propagated through the medium of the open letter.
Most importantly, however, it is important to understand whether or not it remains prudent or even conceivable to put restrictions on the development of artificial intelligence. Apart from the business-centric will to develop AI, it must also be noted that public demand is a determining factor, and AI remains no exception to this trend. In the immediate aftermath of the release of ChatGPT, the application saw a deluge of subscribers that necessitated the expansion of systems and created an urgency in other companies to develop their own editions of conversational artificial intelligence. This was observed in the famed run-up to the release of Google Bard following Microsoft’s integration of the GPT language models with their Bing search engine, which now has the potential to spiral into a rivalry. Clearly, public demand shows no signs of dropping. Also, imposing a pause on development might unfairly penalize companies and developers that are just starting to create their versions of generative artificial intelligence. While there have been limited levels of regulation on the use of artificial intelligence, especially in the educational space, imposing industry-wide diktats might not be easy due to the rather diverse nature of tech businesses and the lack of consensus.
Addressing AI Safety
Securing artificial intelligence and mitigating risks will be key to future development in the niche.
Despite being largely speculative, the open letter from the Future of Life Institute does indeed make AI’s stakeholders approach the notion of AI safety once more. While efforts remain in place to constantly bolster the security of conversational and generative artificial intelligence, it is known that these systems can be vulnerable. Apart from obvious risks such as data security and identity theft, researchers and developers must pay an equal amount of attention to the impact of artificial intelligence on human beings, their behavior, and on the potential outcomes of its usage. Despite the seeming impracticality of imposing restrictions on AI development and unfairly barring companies from creating potentially ingenious technologies, regulatory bodies must pay close watch to the extent to which artificial intelligence can impact and influence human roles in society, following which a more structured approach can be undertaken.