AI sparks unease. We need barriers for bots

In the normal scheme of things, I would prefer natural to artificial any day, but with the latest buzz about Artificial Intelligence (AI), it looks like one may have to make a fresh evaluation.

All this talk about Chat GPT seems to have stirred up a hornet’s nest with academics wondering how they would evaluate the intelligence of a student, who might have accessed the Open Source network, to submit a term paper or an assignment. Anxiety levels have also gone up about whether companies would make use of these artificial aids to make certain jobs redundant, thus putting further pressure on a workspace that has already shrunk with the advent of digital technology.

In the midst of all this, the details of a two-hour chat shared by a New York Times columnist and Bing’s AI Chat Bot seemed to take one back to sci-fi films that involved robots. Who would have ever imagined that what one viewed in these films would one day come to pass with even more dangerous consequences for the world that we dwell in?

Also Read | ChatGPT AI text tool: A fad or promise?

The reporter who had the conversation with the chatbot named “Sydney” reveals a sense of unease that creeps in as the exchange progresses, and the same will probably be felt by anyone who reads the conversation. What hit me was the bot’s programming revealing a desire to do harm and cause chaos, which the reporter called the “shadow self.”

The fact that this “shadow self” sounded so much like real people, who lurked online and in the real world did send an awful chill and the reporter confessed to feeling the same.

What brought relief though is that there are checks and balances that prevented ‘Sydney’ from proceeding further, with the harmful and often bizarre suggestions. Even more of a relief is that, within a week of introducing this new version of the Bing search engine, Microsoft has realised that they seem to have ventured into tricky territory.

Also Read | How to take help from ChatGPT

Human users have used open-ended questions to probe the bot with strange and harmful requests, while some have confessed to a “surprising creepiness” when it came to personal conversations. The company is hence considering further tools to create barriers to prevent “more alarming and strangely humanlike responses.”The telling word here is “humanlike,” because coincidentally, in our very own Karnataka, the state president of the ruling party, Nalin Kumar Kateel, decides to invoke Tipu Sultan and link him with love jihad while spewing hateful words against the minorities. You wonder where the checks and balances are on this kind of negativity from those who are positioned above him in the hierarchy.

Even if his words cannot be retracted, as in the case of an AI bot, at least there should be a public condemnation of his statements. But unfortunately, even in the TV debate run by a national channel, those who represented the party, instead of sounding embarrassed, continued to speak in the same vein about love jihad without being able to substantiate the claim with any data, even when asked for it.

Also Read | AI is stepping into the next level

Perhaps the sleepless night of the NY Times reporter and many others stems from the fact that if the technology to create these chatbots were to fall into the hands of people who spew hatred in the real world, there would be no mechanism to abort or curtail the hateful words that had the capacity to incite violence and xenophobia.

Even today, social media, especially WhatsApp groups, are being used to spread misinformation in a post-truth society. It is hard to understand how people are willing to swallow all the falsehoods that show up on their phones and other digital devices.

A world where reality merges with the digital world can only result in total chaos. Imagine the Kateels of the world merging and expanding their messaging with similar or worse avatars, with nothing to hold them back.

Dystopia versus Utopia is a choice that is more in the hands of people than machines. Is it possible to ask for responsibility from those at the helm of affairs? Probably the only time that politicians approach the people is when they come seeking votes.

It is an opportunity that should not be passed up, as finally, it can only be natural intelligence that can win the day by asking for accountability from those who profess democracy and claim that they seek power for the good of the people.

(The author is a Bengaluru- based independent writer.)