Despite the fact that the AI phenomenon has emerged in the last few years, we are still not entirely clear about it.
Computer Science enthusiasts, as always, are going gaga about it like anything else new computer science or IT related. It’s understandable that AI has a lot of positives and its negativities are also getting crystallized gradually. First and foremost, a lot of people regard AI as some kind of extension of automation that replaces human labor. But not many can explicitly explain what we will achieve by replacing humans so much. Increased unemployment can't be any sane ambition.
A lot of people, however, could argue that science and innovation cannot consider socioeconomy; they will go by their natural flow. But it still raises the question of some regulation, at least, in some areas of innovation and their application especially when the innovation has great potential to do harm alongside the good.
The issue is actually quite complex. Let's start with some simple examples. Clearly, AI generated fake videos and photos commonly referred to as “AI Slop”, fake sounds including impersonation. AI generated art and other creative work, etc. are doing more bad than good. The more difficult debate involves the question of AI replacing humans. As of now, such things are happening largely to low end white collar jobs. However, the data on this matter is kind of inconclusive. For example, in the US some correlation has been found between expansion of AI in the areas of human labor and increase of unemployment but to the surprise of many as of now the negative effect is modest.
One explanation for the modest negative effect is that even if one employs AI to replace humans, that AI’s work has to be supervised by humans at one level. Also with its expansion, AI related innovation and maintenance require additional manpower. Therefore, the layoffs at one level can be absorbed at another with some skill upgrades. However, the effect of AI- good or bad- can vary between developed world countries and the developing world ones. In populous countries like Bangladesh, India, Indonesia, Nigeria, etc. AI replacing humans at a mass scale may create social problems and even unrest. These kinds of countries have little to no capacity to absorb the redundant workers.
But the issue is even if we want, from some standpoint, to regulate AI we can't really stop it. There is no way we can stop it neither is necessary. Countries and authorities should rather focus on how to regulate it efficiently in a way that it doesn't impede the benefits AI brings while stopping its abuses and misuses. This is a fine yet complex line. It's also a daunting task for governments and their bureaucrats who make policies. They need qualified people to help create such regulations and keep upgrading them because the AI world is a dynamic and ever-changing one.
We must acknowledge the perks of AI too. We know about AI’s contribution in aiding various systems, operations and also further research and innovation. For common people, it has brought great ease in finding and learning about things. One simple example could be the social media algorithm. Just one search for a thing you need will allow a series of prompts making your work almost effortless. We also need to recognize the harm that AI is doing at common people’s level. Fields of education and research are impacted both positively and negatively. It has become very difficult to give writing assignments to students to assess their originality. While in case of research papers, AI can aid with searches, content and summaries, AI is also misused for writing a research paper itself.
AI’s negative impact on creating and spreading misinformation and disinformation is also considerable. For example, we often find fake photocards with news headlines from well known media houses. People had to invent fact check mechanisms to debunk the misinformation and disinformation. Interestingly AI has come to rescue in fact checking as well exemplifying both the perks and pitfalls.
It's true that we are kind of fearful of AI because it's hard to comprehend the full gamut of AI. In that state of mind we tend to forget the immense contribution of AI in the areas of boosting productivity, automation, personalization, decision making, advanced data analysis, cost reduction, automating repetitive tasks, minimizing errors and predictive maintenance.That’s how AI is already deeply embedded in our lives and that's why AI is irreversible.
The AI regulations also need to address the issue of crime, AI generated fake audios, videos, documents, etc. are being used to commit financial crimes, tarnishing reputations and making abuses. The regulators have to find effective ways to outsmart the criminals and abusers and stop misuse of AI.
AI is irreversible. We have a lot of interest and enthusiasm about AI including some skepticism. The key thing is to learn to live with AI, extract its goods and prevent its evils.
The writer is a contributor.
