Concerns are mounting over artificial intelligence taking over roles traditionally held by humans, spanning from warfare to the workplace.
A recent essay suggested that drones might soon take over warfare, entrusting AI with the decision-making on whom to target. Companies are already replacing human workers with machine intelligence.
"AI hit the technology scene in a huge way over the past year and is already heavily being embraced - or companies are racing to get involved so they don't fall behind," Kev Breen, the director of Cyber Threat Research at Immersive Labs, told Tech Times in an interview. "But among all its popularity, we're simultaneously seeing a significant amount of FUD - fear, uncertainty and doubt - and misunderstanding. People still don't fully understand the risks and vision of AI, which lends itself to paranoia and fears of unfounded massive cybersecurity attacks by AI."
Increasing Concerns around AI
The fear that AI could dominate humans stems from the recent rapid advancements in the technology's capabilities. Concerns range from AI systems making critical decisions without ethical considerations to the potential for job displacement across various sectors. The idea of autonomous weapons and surveillance systems operating beyond human oversight adds another layer of anxiety. This apprehension is not just about losing jobs but also about the profound implications for privacy, security, and the very fabric of human society.
SAP, a big technology company, recently announced that is cutting jobs while spending over $2 billion on AI. They're doing this as part of a significant change in the company. At the same time, SAP said it would change 8,000 jobs. This news means some people will lose their jobs, but others will get training to work with AI.
There is "definitely" a risk of advanced AI causing widespread harm to humanity, said Darren McKee, the author of a book about AI risk and safety titled "Uncontrollable." Reasons to be concerned include rapidly advancing AI capabilities, confusion over exactly how advanced AI systems work, whether they will do what we want, and whether we can control them if we can't.
"But even without definitive proof, we should acknowledge the risk in uncertainty. Would we rather be overcautious or caught off-guard?" he added. "If a large number of experts said that there was a 10% chance your plane would crash, you wouldn't get on that flight. Why wouldn't it be prudent when AI experts are warning of severe risks from advanced AI?"
Experts say many jobs are at risk from machines, but which ones are in the crosshairs is still too early to say. Technological advancements, especially in AI and robotics, are rapidly evolving, potentially leading to significant automation in various sectors, Mayank Jindal, an engineer at Amazon who focuses on building AI-based software, said in an interview. While some fields, especially those requiring repetitive or predictable tasks, are prime candidates for automation, others that demand creativity, empathy, and complex decision-making resist easy substitution.
"Stories of AI surpassing human capabilities in specific tasks feed into narratives of a future dominated by autonomous machines," he added. "However, these fears can be overblown when overlooking the current limitations of AI, including its reliance on human oversight and the complexity of replicating human-like understanding and empathy."
Not Everyone Sees AI as a Doomsday Event
Some experts say fears about AI are overblown. There are real risks associated with AI, and we can expect vastly more sophisticated fraud, election disinformation, unfair treatment of people, corporate scandals, etc., due to the intentional and unintentional misuse of AI, Kjell Carlsson, head of AI strategy for Domino Data Lab said in an interview.
"Unfortunately, the fear-mongering about the largely imaginary fears of calamitous unemployment or the destruction of humanity serves only to make it harder to mitigate the real-world risks we face with AI today," he added. "Ultimately, the biggest threat we face with AI today is people, and we need the understanding, processes, platforms, and regulation to ensure that people develop and use AI responsibly and effectively."
While there should be regulation that protects people from the dangerous effects of "bad AI" like deep fakes, for example, the prevalence of these types of malicious AI use cases is vastly overblown, Dan O'Connell, Chief AI & Strategy Officer at Dialpad, told Tech Times. He said that most AI can be defined as "good AI," or AI that enhances human productivity, and it's not as scary or all-encompassing as people fear.
"As we've seen time and time again, it cannot be relied upon to produce consistently accurate results without ever-present human oversight," O'Connell said. "Think of generative AI like an editor or a copywriter; it's a tool that makes you faster and better at your job. Government regulation has proven to slow down innovation, and I worry about forms of regulation where there are too many restrictions that stop good people from collaborating quickly and freely. Generative AI is impressive and amazing, but it won't be the magic pill for everything you need. Just like with the invention of the first computer or the internet, AI will make us more efficient and better at our jobs and create new start-up growth."
AI will serve as a co-pilot to assist human workers in improving decision-making, reducing administrative and manual tasks, and elevating their jobs, argued Gautam Kunapuli, engineering manager at Motive. His company's research found that 96% of physical operations leaders agree that AI will impact their jobs, but only 12% have concerns about AI eliminating their jobs. Rather, 1 in 4 physical operations leaders use AI daily to support their work, and 71% say they want to use generative AI.
"Consider jobs in the physical economy like construction, trucking, field service, etc, that work in a variety of weather conditions and dangerous situations and require far more complex analysis and decision-making capabilities than the decision-making required for driving a car," he said. "Drivers are still needed for pickup and delivery (e.g., delivering gasoline via a hose to a pump in the ground), navigating bad weather, construction, or a tire or traction problem, getting off an exit, driving through cities, and even platooning, where automated trucks follow a lead truck driven by a driver."
Most occupations such as programming, document review, and writing won't disappear, but they will take a different shape, said Or Hiltch, Chief Data and AI Architect at JLL Technologies.
"AI will take it to the next level; computers will still generate code for us - AI is like a compiler from natural language to machine code in that sense," he added. "So, most of the human work would be to review the code generated by the AI."
