SURJIT FLORA
It is important to consider that the world operates according to a specific structure, and it is crucial to avoid imposing our personal biases, as they may not always align with what is objectively true. If a specific trait is discovered to be associated with lower performance or any other outcome, then so be it. ML is indeed a highly unbiased form in this context. There seems to be a lack of understanding regarding the concepts of race and gender. These symbols may not hold much significance, as they merely calculate the connections they observe. It can be argued that this is one of the most impartial methods of hiring, unless specific rules or criteria are hardcoded. We shouldn’t hard-code political prejudices into our systems. Leave them to find the world’s truth, not what we want. The only prejudice is the one you wish existed (more “equality”) but doesn’t. The prejudice is yours, not the algorithms or data. Ignore them. Google is predominantly composed of individuals with questionable intentions, seeking to exert control over every aspect of your existence, reminiscent of a dystopian narrative. Google once had a written policy of avoiding unethical behavior. It has been removed quite times ago. The American left has shifted significantly towards a more extreme position, raising concerns about potential parallels with historical political movements. It revolves around government control and their desire to obtain as much as possible, regardless of the consequences. Google opposes free speech to control you. Google opposes firearms, so you can’t fight. Google fosters the concept that only the left-wing of a large government can feed and protect you since everyone is racist and misogynistic. Where leftists established the laws, self-defense was outlawed. Everything they can make illegal. The future of humanity can be greatly influenced by AI in numerous ways. It has the potential to enhance efficiency, aid in problem-solving, and optimize decision-making processes. Artificial intelligence is not synonymous with true intelligence. And it’s not “artificial consciousness”. There is a common concern among many individuals regarding the potential for AI to unexpectedly become sentient, become displeased, and assume control of the planet. Or that AI will eventually replace human workers. This is highly likely to occur. However, without the “wake up” component. Nevertheless, it is crucial to take into account the ethical and societal consequences of the progress made in the field of artificial intelligence. The future impact of AI on humanity will be determined by the way it is developed, regulated, and incorporated into different areas of life. The concept of AI surpassing humans as the dominant species is widely discussed. As we delve deeper into the intricacies of AI, it becomes increasingly evident that it falls short in comparison to the effortless capabilities of humans. In my opinion, AI will primarily serve to support and enhance human capabilities in our everyday activities, rather than supplanting us. Allow me to explain… Humans have the remarkable ability to adapt. We are naturally equipped with a brain that has the remarkable ability to learn and adapt to new situations as we progress into adulthood. AI lacks comparable capabilities. Training an AI system requires a substantial amount of examples. Once trained, altering its path or thought process necessitates reprogramming it entirely. Even with extensive training and instruction in communicating human ideas to an AI, there is no absolute assurance that the software will interpret the information accurately. In his book “Mindful Universe,” Steven Hawkins suggests that reality is a construct of our minds. According to him, our brains shape our perception of reality, making it feel real to us. Essentially, it is difficult to determine whether AI will comprehend our commands and requests without gaining a comprehensive understanding of the underlying system that enables it to comprehend human communication. Even with extensive expertise in artificial intelligence, there is still no guarantee that the system will fully understand and be able to execute our requests as commands or functions in an operating system or programming language. Because when you consider it – even among individuals with a shared understanding of morality, this is not always the situation. There are numerous ethical considerations involved in societal communication that are not solely determined by written laws, although written laws do contribute to the establishment of social norms. Last year, ‘prompt’ made the Oxford Word of the Year shortlist. This shows that generative AI is always in the news. The prophetic perspective is that AI will conquer the world. Will AI rule be a success or a disaster? Consider this question. ChatGPT demands stubbornness. Despite global discussions on tolerance and diversity, the future AI on the verge of replacing humans maintains its strict worldview. AI has a clear view of gender, nationality, and color, unlike the modern world. Rest of the World, a tech news site, reported that Midjourney, a popular image-generation AI application, has engaged in extensive racial stereotyping. “An Indian person” is usually a bearded elderly guy because of beard on their face. “A Mexican person” is frequently a sombrero-wearing dude. Many New Delhi streets are filthy and littered. Most impoverished people are black. Although skewed, these assertions represent scary insights from neural cues provided to Midjourney. Bias may have devastating repercussions. Given the hype, companies are using AI to automate recruiting, security, loan provisions, etc. An AI-powered recruiting tool can screen out competent candidates based on race or gender. Imagine no more-this has occurred. Amazon had to withdraw their AI-based recruiting system because it was biased against women. In 2022, US authorities mistakenly detained Robert Williams owing to prejudice in face recognition technology. These facial recognition technologies also misidentify black and Asian minorities, according to studies. In addition to recruiting, AI biases have affected wealth exchange. Algorithmic discrimination in Chicago denies minority neighborhoods financing. AI systems that automatically accept loans may exclude marginalized people due to past biases in the training material. ChatGPT responded to an impressively sharp request from an X (previously Twitter) user by saying smart scientists are “white” and “male.”. Another study by the study by the University of Michigan found that Open AI’s CLIP misrepresents low-income and non-Western lifestyles. Also found was geographical prejudice against African nations. Overall, the report card for AI is a bit mixed. However, it’s worth noting that the field is relatively young, with only 60 years of history. Physics has made significant progress over the past 400 years, but there is still much to learn. One of the mysteries that continues to baffle scientists is the concept of “dark matter,” which accounts for a large portion of the universe’s mass.
Ultimately, AI has the potential to assist physicists in overcoming their challenges, once they enhance their problem-solving abilities. However, is it possible for us to develop an automated version of Stephen Hawking? It will take some time. The team at Allen Institute for AI is facing challenges in getting machines to successfully pass question answer tests designed for preschool level. We still have a lot of progress to make.
(The writer is a veteran journalist and freelance writer).
