According to a new Forbes survey, a significant portion of Americans still have more faith in people than in artificial intelligence. According to the survey, people would be better at creating laws, picking gifts, and even giving medication, to mention a few tasks. Based on their knowledge and relevant research, the College of Engineering faculty members have their own perspectives.
To find out more about how AI affects us, we wanted to speak with some of the experts in the field. Experts in computer science, electrical engineering, aeronautical engineering, and even building construction are among these academic members. The positive, negative, and (perhaps) frightening things they said about AI are listed here.
The New AI Technology is Indeed a Blessing:
In a recent preprint study, a group of Microsoft researchers who were evaluating OpenAI’s GPT-4 which, in my opinion, is the best of the new advanced chatbots on the market right now—said it contained “sparks of advanced general intelligence.”
GPT-4 outperformed 90% of human test takers in the Uniform Bar Exam, a standardized exam that several states use to certify attorneys for practice. Compared to the prior GPT-3.5 version, which was trained on a smaller data set, that percentage increased from just 10%. Numerous additional standardized tests showed comparable gains.
The Good: Improved Accessibility and Quality of Life!
“Autotrophy with physical disabilities can gain new possibilities through AI and robotics.” Self-driving vehicles have been promising greater mobility, rehabilitation robots have enabled children to walk, assistive robot arms and motorized wheelchairs have restored the independence of older adults. Regular consumers of this technology hold out the promise of enhancing their quality of life.
The Bad: Potential Bias from Incomplete Data!
“AI is a powerful tool that is easily misuse. AI and learning algorithms tend to extrapolate on the data they are given. The resulting AI systems become unfair and biased if the creators do not provide representative data.
A human recognition algorithm may not be able to recognize a person with brown hair, for example, if it is train on only images of individuals with blonde hair (i.e., brown hair = not a human). In the real world, rushed AI implementations have created systems that display gender and racial biases. AI’s weakness is that it does not treat all users equally”.
The Scary: Artificial Intelligence is Influencing Our Decision Making!
“The detrimental effects of AI are already affecting us. Think of streaming services’ recommendation engines, which dictate the types of shows you watch based on what an artificial agent recommends to you. More generally, contemporary AI systems influence human decision-making at many levels, from viewing habits to buying habits, from social mores to political ideologies.
To claim that the impact of AI will be a problem for coming generations ignores the reality that the impact is already being experience in our daily lives. The situation with artificial intelligence today is fundamentally uncontrol and unregular. Companies and organizations can build algorithms that maximize their revenue, engagement, and reach.”
The Good: Improved Communication with Machines!
Our relationship with technology is being transform by large language models, or LLMs. More interactive dialogue with computers is now possible due to their capacity for parsing and generating text that approximates human writing. These models are far from mere job automation; they are adaptive support tools that individuals can utilize for Emotional Support, brainstorming, or even practicing difficult interactions.
Think of having a resource—a not exactly a friend, but a helpful tool—available to assist whenever you require new information or a different perspective. While it’s important to remember that these models are still tools and not substitutes, they are starting to bridge the gaps in domains that we have traditionally reserved for human touch.
The Bad: Increased Dependence on Technology Reduces Critical Thinking!
Lowers Critical Thinking:
Controlling our reliance on LLMs is a challenge that accompanies their power. Over-reliant users of AI-created content without questioning risk losing their powers of critical thinking. Besides, such models can inadvertently perpetuate biases in their training data since they are train on massive amounts of online content. Therefore, it is important that we apply LLMs with an open-minded perspective, taking cognizance of the risks and prejudices they contain, and ensuring that they augment and not diminish human intellect.
The Scary: Potential Loss of Human Connection!
The potential to degrade actual human connection is among the deeper challenges of LLMs in human-AI interactions. Of course, there’s also fear for the authenticity of our experiences with AI as we increasingly begin engaging with them more regularly.
Will we at some point cherish the predictable, personalized responses of an LLM over the unbridled, sloppy, spontaneous yet genuine nature of human interactions? In addition, since AI is capable of creating engaging stories, there is the moral danger that it will be employee to deceive or manipulate.
Hence, it is crucial that we discuss how to set boundaries and ethical guidelines for the usage and implementation of LLMs so that they are use to enhance our lives instead of devaluing interpersonal relationships. While LLMs are challenging, they also bring unprecedented possibilities. It is up to us to judiciously leverage their advantages.
The Dangers of AI: Conscious or Not, the Risk Remains!
Others assert that such LLMs are actually dumb automation devices, with the belief being that they would be less likely to break out of their programming if they’re not conscious. Whether or not such language models are unconscious entirely, now or even in the future, doesn’t matter. To be on record, while I’m still open to new data as it becomes public, I agree that it’s unlikely they’re truly conscious as of now.
In either event, millions of people can be kill by a nuclear explosion without any awareness whatsoever. So too can artificial intelligence kill millions of people without awareness in numerous ways, including the use of nuclear bombs directly (much less probable) or indirectly (more probable).
