Stephen Hawking Cautioned About The Dangers of Artificial Intelligence– yet AI Offered Him a Voice
The argument over how artificial intelligence can assist humans featured a significant contribution from the late Stephen Hawking. Hawking publicly expressed his concern that intelligent machines might someday take over. Even further, he stated that potential advancements in AI “Could spell the end of the human race.”
However, Hawking’s relationship with AI was much more nuanced than this frequently quoted soundbite suggests. The fundamental concerns he highlighted were regarding superhuman AI, the stage that is, at best, decades away, if it ever occurs at all, where AI systems not only copy human intelligence processes but also continue to evolve them without our support. And yet, fundamental AI technology became essential to Hawking’s capacity to convey these anxieties as well as all of his other concepts.
The tumultuous Hawking-AI relationship
My colleagues and I research the consequences of developing technologies, such as those Stephen Hawking was concerned about, at the intellectual property and health law centers at DePaul University. At its foundation, the idea of artificial intelligence (AI) incorporates computer technologies intended to give robots the ability to think strategically, emulating and ultimately surpassing human thought processes.
A severe type of AI, in which reasoning computers “take off” on their own, changing themselves and independently creating and constructing ever-more-capable systems, was something Hawking warned against. The slow speed of biological evolution would severely limit human intelligence.
AI as a danger to humanity?
AI can be used for evil purposes well before it develops to the level of superhuman technology. Scholars and commentators are already concerned that self-flying drones could pave the way for deadly autonomous robots.
Other ethical and practical issues are also brought up by today’s developing AI. Most AI systems are built on secret algorithms that produce results that even their creators may find difficult to explain. Computational errors can happen and fundamental mathematical models can be skewed. AI might gradually replace human expertise and raise unemployment. Furthermore, a lack of access to AI can worsen existing inequalities.
Some of these worries were emphasized in The One Hundred Year Study on Artificial Intelligence, which Stanford University issued in 2014. However, it has not yet found any proof of the “imminent threat” Hawking said AI would bring to humanity.
However, Hawking’s opinions on AI are more complex and less alarming than he typically is given credit for. They essentially speak to the necessity of comprehending and governing developing technology. Frequently urged further investigation into the advantages and risks of AI. He also thought that non-superhuman AI systems could contribute to the eradication of conflict, poverty, and illness.
Hawking talks
This apparent contrast between excitement about AI’s benefits and concern that it will eventually replace humanity may have been inspired by his personal experiences: Hawking began to interact with the environment using AI.
He has been unable to speak since 1985, he has employed a variety of communication devices to help him converse and write, leading to the now-famous computer that is controlled by a single muscle in his right cheek.
The computer program’s initial iteration was excruciatingly slow and prone to mistakes. Simple AI altered it. His word choice was considerably accelerated using an open-source application. More crucially, it employed artificial intelligence to examine Hawking’s own words before using the results to aid in the expression of fresh ideas. The system improved so much as a result of processing Hawking’s books, journals, and lecture notes that he no longer even needed to type the phrase that is most commonly associated with him, “the black hole.” When he clicked “the,” “black” would propose that he do so by itself, and “black” would prompt the word “hole” onto the display.
AI makes people’s health better.
Stephen Hawking’s encounter with such rudimentary AI serves as an example of how non-superhuman AI can, in fact, improve people’s lives. He had a crippling neurological condition, but speech prediction enabled him to get along. Already, other AI-based systems are assisting in the prevention, treatment, and reduction of disease.
A patient’s likelihood of developing a serious blood infection, for example, can be predicted using AI analysis of medical sensors and other health data. Studies showed that it was far more precise than alternative methods and gave much more prior notice.
To comb through 700,000 patients’ electronic health records, another team of investigators developed an AI program. The program referred to as “Deep Patient” discovered connections that had previously eluded medical professionals, revealing new risk patterns for specific malignancies, diabetes, and mental health conditions.
Even in a procedure on pigs that is remarkably identical to a specific kind of operation on human patients, a robotic surgery system driven by AI surpassed human surgeons.
AI holds so much potential for enhancing human health that gathering medical data has become essential to both software development and public-health initiatives in the United States. For instance, the Obama White House started a study project to gather DNA samples from at least one million Americans. The information will be made accessible for AI systems to review when researching novel medical therapies, potentially enhancing patient recovery and diagnostic accuracy.
These advantages of AI are all currently present, and more are in progress. They do show that superhuman AI systems may be highly effective, despite Elon Musk and fellow technology enthusiast Stephen Hawking’s concerns. It might never happen. As Hawking was aware, there is much to be acquired in the interim. He appealed for both research and moderation using AI, which gave him a greater and more effective voice than his body could.
