Vigilance against AI is on the rise, with leaders in various fields expressing concerns about both the growing power of AI and its role in society.
Thousands of CEOs, Technicians, Researchers, Academics and More Call for Pause of AI Adoption in Early 2023, Despite Millions of People Starting to Use ChatGPT and Other Generative AI Systems signed an open letter.
Specifically, it called on “all AI laboratories to immediately stop training AI systems stronger than GPT-4 for at least six months.”
The March letter cried out AI’s “grave risks to society and humanity” and called it “an uncontrollable race to develop and deploy an ever more powerful digital mind that no one, not even its creators, can do.” It starts by blaming the AI labs that are participating. – Can be understood, predicted or controlled with certainty. ”
They’re not the only ones concerned about AI. In a February 2023 Monmouth University poll, 46% of his respondents said AI would do as much harm as good, while 41% thought technology would do more harm. I’m here. Only 9% of respondents believe computer scientists can develop AI that benefits society.
In reality, AI has many potential advantages and disadvantages. In fact, the signatories of the open letter acknowledged both the downsides and upsides of the technology, stating that “Humanity will enjoy a prosperous future with AI.”
Top 5 benefits of AI
AI has the following advantages:
1. Available 24 hours a day, 365 days a year
One of the biggest and most-cited benefits of AI is its 24/7 availability. Other computer technologies he has been running 24 hours a day, and companies have benefited from the high availability of such systems, but only if humans can operate the systems. Jordan Ray Kelly, senior managing director and head of cybersecurity for the Americas at FTI Consulting, believes that AI’s ability to make decisions and take action without human involvement in many business situations means that technology can become independent. It means that it can function as a system and ensure continuous operation.
2. Scalability
Sreekar Krishna, Principal and National Leader for AI at KPMG US, says AI not only works continuously, but can scale almost infinitely. He cited the personalized recommendations that companies such as Amazon and Netflix provide to their customers. A frequent customer-facing clerk might be able to extend such a service to that same individual with enough interaction, but AI can do it to hundreds of thousands of customers at the same time. AI is demonstrating this scalability in the financial industry as well, with institutions using the technology to verify and verify millions of transactions instantly and monitor for potential fraud on a daily basis. . “Humans alone cannot scale to that extent. We need automation, and AI is essential for that automation,” said Krishna.
3. Improved accuracy and reduced error rate
Unlike humans, AI systems don’t get tired or distracted. Because they can process infinitely more information and consistently follow the rules to analyze data and make decisions, they are much more likely to get accurate results almost all the time. “I’m not saying these platforms are perfect,” Kelly warned. To achieve such accuracy, AI models must be built on good algorithms without unintended bias, trained on sufficient high-quality data, and monitored to prevent drift.
In a February 2023 Monmouth University poll, 46% of respondents said AI would do as much harm as good, and 41% believed technology would do more harm.
4. Improved safety
An example of AI’s ability to improve safety is General Motors’ Super Cruise feature that ensures drivers pay attention to the road. According to information on the company’s website, “While in operation, Super Cruise uses a driver attention system to monitor the system’s status and detect the user’s head and eye position to keep them alert and alert on the road. It reminds the user to manually steer accordingly.” Other automakers are also offering AI-powered features such as lane departure warnings to improve driver safety.
But AI’s contribution to safety extends beyond the road. They are also used on production lines to keep workers safe, such as stopping machines if they get too close to a certain area. And they are used within robots to handle a variety of dangerous tasks, from defusing bombs to accessing burning buildings, and preventing humans from performing those life-threatening tasks.
5. Perform routine, repetitive tasks
Experts also appreciate AI handling repetitive tasks in both work and personal lives of humans. “It takes the burden off the tedious work that humans have had to do,” Krishna said. As more computer systems incorporate AI into their operations, AI will be able to perform low-level, tedious tasks that can eat up an individual’s time. Everyday examples of AI handling everyday tasks include robotic vacuum cleaners in the home and data collection in the office. As a result, humans can spend their time on higher-value tasks.
Top 5 flaws of AI
Commonly cited drawbacks of AI include:
1. Lack of creativity
AI is responsible for creating everything from computer code to visual art, but it lacks a mind of its own. “You can only know what you know. You can’t think outside the box. No kidding,” Kelly said. “There are limits to what you can eat.”
AI basically makes predictions based on an algorithm and given training data. Machine learning algorithms help machines learn over time, but they lack the creativity, inspiration, and capacity for new thinking that humans have. “This is not a replacement for critical thinking. It’s just another arrow in our quiver,” said Chaim Mazar, chief security officer at cybersecurity technology maker Gigamon. The World Economic Forum’s February 2023 report notes that while AI can support and enable human creativity, “the prevailing opinion is that AI cannot generate radically new ideas on its own. ”. Whether AI can develop such capabilities is a matter of debate, even though a Microsoft researcher claims in his April 2023 paper that AI has evolved to reason like humans. subject to debate.
2. Lack of empathy
AI can be taught to recognize human emotions such as frustration, but machines cannot empathize and have no ability to feel. Humans can do that, which gives them a huge advantage over emotionless AI systems in many areas, including the workplace. Consider the fact that the service sector dominates the US economy. The Brookings Institution predicts that by 2023, “four of her five private sector workers in the United States will be employed in the service economy, in everything from providing care in hospitals and nursing homes to providing health care. I’m doing business.” We provide food and ensure that products get from the port to the store shelves and into the hands of the consumer. Krishna said some roles, and other aspects of the role, can be automated using AI, but many of those roles “need empathy and touchpoints.”
3. LOSS OF HUMAN SKILLS
Experts usually cite AI’s ability to free people from repetitive, mundane tasks as a positive, but some believe that this particular advantage comes with a downside—the loss of people’s skills. increase. People often first learn and master simple, repetitive tasks to improve not only their knowledge, but also their personal and professional skills, which those tasks must accomplish in order to achieve their goals. You’ll be able to understand how it fits into the larger piece of work that shouldn’t be. But as AI takes over these entry-level jobs, there are also concerns that people may lose their ability to know and understand how to perform those tasks. It can hinder your ability to truly master a profession or trade. It may also lose the necessary ability to intervene and perform work if the AI fails.
4. Overreliance on technology and potential for increased human laziness
Similarly, some thought leaders say they fear AI will enable human laziness. “We’re all worried that it’s going to make us less alert, less thinking, less thinking,” said Bill Wong, chief research director at Infotech Research Group. He and others noted that some users appear to assume that the AI technology is working and use it without double-checking the results. AI is far from perfect, and the online site AI Incident Database “indexes the collective history of harm, or near-harm, that has been realized in the real world by the introduction of artificial intelligence systems.”
5. Unemployment and displacement
300 million full-time jobs could be lost to automation, according to a Goldman Sachs Research April 2023 report. The authors also estimate that “roughly two-thirds of U.S. occupations are exposed to some degree of automation by AI.” However, the story is complicated. Economists and researchers say AI will eliminate many jobs, but it will also shift some workers to higher-value jobs and create new types of jobs. Predicting. Existing and future workers will need to be prepared by learning new skills, experts say, including the ability to use AI to complement human capabilities. The Goldman Sachs Research authors conclude that “Although the impact of AI on the labor market is likely to be significant, most jobs and industries are only partially exposed to automation and will likely be replaced by AI.” It is more likely that it will be complemented rather than augmented.”
Examples of good use cases for AI
According to experts, AI has brought about major advances in many areas of society. The following use cases demonstrate the positive aspects of this technology.
New vehicles employ AI-enabled systems to improve road safety by monitoring blind spots, alerting drivers when they are distracted, and taking preventative measures such as automatic braking to avoid collisions. to ensure
Swiss researchers announced in the spring of 2023 that they had helped a paralyzed man to walk for the first time in 12 years using AI as part of a medical plan.
Scientists have always used the latest tools to advance their research, and this is exactly what happened during the COVID-19 pandemic. Case in point: Researchers have created AI models that can predict which novel coronavirus variants will prevail and when and where a spike in cases will occur.
Examples of bad use cases for AI
Although AI itself is neutral, we can see that in some situations AI’s use can be limited and potentially harmful to others. These real-life examples show how AI can be used inappropriately.
In the late spring of 2023, a New York attorney came under judicial scrutiny for submitting to court citing a fictitious case fabricated by ChatGPT. The attorney admitted to using ChatGPT to prepare the documents and told the federal judge he had no idea the tool could make such mistakes.
In 2019, a video circulated online showing California Democrat Nancy Pelosi, then Speaker of the U.S. House of Representatives, apparently drunk. The video was a deepfake (a media format altered using AI). The veracity of that Pelosi video and others that followed have raised alarm bells about how AI-generated content can be used to distort the truth and spread misinformation. rice field.
In 2016, when Microsoft released a chatbot on Twitter, there was another well-known example of an AI use case going poorly. Microsoft engineers designed the bot to behave like her teenage girl, and the name given to the bot, her Tay, while engaging online like any other teenage girl. I expected to learn to be But Tay apparently didn’t have the guardrails to block racist, misogynistic, and anti-Semitic language, becoming an aggressive and hostile bot that Microsoft was forced to shut down.
