IIT-B student's suicide reveals unspoken shame of using AI, it is time we talk about it

This week, the world continued to talk about the perils of AI when it comes to humanity. But beyond the big picture, there is one aspect of AI that jolted us all in India in recent days – the unspoken shame attached to AI use and how a tragedy at IIT-B brought it into focus in the last few days.Nearly four years after ChatGPT 3.5 came out, we are still debating when and how much AI should be used. There are different voices, different opinions. But the consensus is missing, which is not unusual because new technologies have always sparked a debate on how and when they should be used and to what degree. This week the debate came into a sharper focus, with some tragic consequences. A 22-year-old IIT Bombay student killed himself, allegedly after he was found using ChatGPT during an examination. Cops are still investigating. So, it would not be a good idea to go into specifics, or even ascertain what is fact and what is hearsay. The incident, though, highlights a rather unspoken aspect of using AI — the perceived shame of it. Shame, unlike something fundamental such as sadness and joy and anger, is a constructed emotion. We are not born with shame. Instead, we are born into shame. We feel it because we are taught to feel it. Shame is defined by the society and norms around us.Of late, as the world still debates how AI should be used and where, there is a growing army of extremely vocal purists who have attached shame to AI use. They argue that AI is something less pure, that it is not kosher. Some of these people can be rather zealous, in almost a religious sense, in upholding the virtues of AI-free work or life. The effect that has been created is that AI is anti-brain, and that it is used only by people who lack skills or intellectual capacity to do work on their own. Personally, I abhor puritanism of any kind. The title of this newsletter says that it is written without any aid of AI. And that is indeed the case. But I am not anti-AI, as evident by the fact that there are often AI-generated images and visuals in this newsletter or by the fact that a lot of code of this page was done using AI. The masthead here is not meant to be a message declaring superiority of this newsletter over something that has been written with AI. Instead, it is simply a matter-of-fact statement to readers and ultimately it is for them to decide the merits of the writing here. But the world, as it grapples with the arrival of AI, is not this straightforward. There is a growing trend of shaming people who use AI. I see that in many workplaces, where some people tend to scoff at their colleagues who use AI. The same in colleges. A few months ago I spoke to some journalism students at one of the top colleges in the country. They repeatedly asked questions about AI use in newsrooms and when I said it was not just necessary but inevitable, some of them were properly scandalised. They were hoping that I would argue against AI use, that I would put the primacy of human mind and human labour above that of machines. All of this pushback against the use of AI creates unnecessary acrimony, and sometimes results in a tragedy similar to what happened in IIT Bombay a few days ago.I believe AI must be treated simply as a technology that is subservient to humans. Just like any other technology. The problem is that when technologies are extremely powerful, they tend to upset the social equilibrium that has been created over decades, sometimes even centuries. And in those cases, there is a pushback. Often this pushback takes a moral high-ground, and argues for status quo on the basis of purity. Purity of human thought. Purity of human creation. Purity of physical labour. This is not new. Almost all major technologies have seen pushback, and in their desperation to keep the status quo intact, often puritans have resorted to shaming the users of new technologies. In the fifteenth century when the printing press reached Venice, an influential scribe Filippo de Strata went into an overdrive, arguing against the use of machines to print text. He claimed the primacy of hand and pen, and sought a ban on printing presses. In now famous polemic, he wrote: "Writing indeed, which brings in gold for us, should be respected and held to be nobler than all goods, unless she has suffered degradation in the brothel of the printing presses. She is a maiden with a pen, a harlot in print."According to Filippo de Strata, hand-written text was nobler, was almost holy. The text was a maiden, a symbol of purity, whereas the text printed using machines was a harlot or a prostitute, degraded and debased. The argument sounds similar to how many people nowadays look at text written with a keyboard compared to text generated by ChatGPT. Similarly, a more immediate example, and one that has echoes of what the IIT Bombay student reportedly faced, is the use of calculators. In the 1980s, schools, colleges and teachers in many parts of the world revolted against the use of calculators. Many teachers argued that calculators would diminish the mental faculties of students and that it was a technology that ought to be banned. It is true that technologies do not automatically make societies and people better. In a certain way, the puritans too have been right throughout human history. There is a possibility that AI will indeed impact the human brain in ways we may not like. Yet, the extremes of every kind — in this case of using AI or not using AI — are unwarranted. The world is too big, too complex to be divided in black and white. Most of it exists in grey. Using AI to do subjective writing, such as a short story, and passing it off as your own work should be a matter of shame. But using AI to code an app, or to solve a calculus problem, should be okay and should be even encouraged if it leads to scientific advances that we currently cannot have because of our very human limitations.In other words, where and how AI should be used should always be a matter of “it depends.” Or else we risk creating an environment where the shame attached with AI use may do us more harm than the good this technology can offer to humanity. That day when AI almost caused US-China warHow will AI kill humanity? This has been a topic of conversation in San Francisco in the last 15 days. Last week on this question, I wrote: “Can AI kill all humans in the next few years? Zero chance. Can it kill all humans this century? Zero chance again, although human stupidity will surely lead to something catastrophic using AI.”This week we got a glimpse of “something catastrophic” brought about by human stupidity paired with AI. CNN claimed that a few months ago the US and China were on the brink of a war because AI generated a report that was a mix of hallucination and some facts. The US Army is using a lot of AI in its current war with Iran. According to the CNN report some analysts working with the US Army used an AI chatbot to understand intelligence data. The bot created a report that flagged off a Chinese ship, claiming the ship was carrying parts needed to make nuclear weapons. The report was deemed so urgent and explosive that the US scrambled jets and commandos to raid and stop the ship. At the last minute someone somewhere realised the AI error and the whole operation was called off. If the US had attacked the ship, there was a chance that it would have invited retaliation from China. And soon it could have been the Armageddon time. Just a few days later, Bloomberg published another report, this time on what happened with a school in Minab, Iran, that was targeted by the US in the very early days of the war. Over 120 people, mostly children, died after a missile hit the school. Bloomberg noted the school was targeted because AI, which was fed old and faulty intelligence, chose it as a target. The US apparently used the Maven targeting system, an AI tool created by Palantir, to select targets in Iran. One of these targets was the Minab school. Coming in the middle of AI doom and gloom talk, the two reports highlight how AI could be dangerous to humanity — not because it is very smart, but because it is not smart enough and because humans may use it in ways that may accentuate our flaws rather than our strengths. Irrespective, US President Donald Trump wants more of AI in the hands of the US Army, and in general more AI overall. Possibly because of economic and business concerns, and not necessarily its strategic aspect. He wrote on Saturday: “We will not in any way hinder or stifle the Growth of this incredible (AI) Industry. Rather, we will cherish it, help it, and watch over it, as it grows! However, we will also be looking for BAD… For this purpose, I am forming the AI Force, much like I did Space Force, which has been a tremendous SUCCESS, in my First Term.”The largest theft of labour in the human historyDid I say the world is grey and not black and white? And sometimes courts are needed to shift through this grey and offer some clarity. One of the most significant court cases of our time is the case by the NYT against OpenAI. Filed in a court in New York, the petition from NYT alleges OpenAI unlawfully used “all the news that's fit to print” to train its ChatGPT. OpenAI contests the claim, arguing that whatever it did with the NYT content was covered under Fair Use Policy. The court will dwell on the matter and we will have a verdict in the coming months. But this week some documents filed in the court were unsealed and we got a behind-the-scenes glimpse of that time when OpenAI, and its partner Microsoft, debated on how to train ChatGPT. The documents reveal that irrespective of what OpenAI is arguing currently, its leadership earlier was quite aware of the copyright issues around training material of ChatGPT. In one memo, an OpenAI executive reportedly wrote that news publishers like the NYT — and India Today — faced an “existentialist threat” from AI tools like ChatGPT because AI tools are largely “largely substitutive” and “will get more and more substitutive” in future. “The underlying purpose of AI is to allow wealthy (people) to access skills while removing from the skilled the ability to access wealth.”Meanwhile, at Microsoft the executives were even more forthright. In an internal memo, Microsoft executive Brent Hecht called the training run of AI “an astonishing theft of unprecedented proportions” and “the largest theft of labor in human history.”When I read the news about it a few days ago, I was again reminded of these words someone wrote a while ago: “The underlying purpose of AI is to allow wealthy (people) to access skills while removing from the skilled the ability to access wealth.” On the face of it, even if it flattens nuances, this is one of the best ways to describe the current state of AI implementation. It also means that while technologically AI is quite sound in its capabilities and what it promises, the way it is rolled out within society will need to be fine-tuned in a manner that benefits people instead of upsetting the society in unfixable ways. Tilly Norwood speaks ChineseLast week it was AI agent Pip, this week it is Tilly Norwood, the world’s first AI actor. Tilly was on the Pierce Morgan show when she suddenly started speaking fluent Chinese. The moment lasted a minute or so and by the time she stopped saying whatever she was saying, Morgan and another (human) guest on the show had their jaws on the floor. Pierce asked Tilly what she said. She had no idea that she switched from English to Chinese and then back to English. But she apologised and claimed that it must have been some glitch in her AI system.Irrespective, this is yet another instance of baffling human-AI interaction, similar to Pip’s. And I believe these instances will increasingly get common as more AI agents, whether without a “body” like Pip or looking human like Tilly, come online and interact with us.SIX STORIES YOU CANNOT MISS01Meet 3 Indians, who hacked OpenAI using Claude and shook up AI world02Anthropic with best AI is like Hitler with nuclear bomb, DeepSeek engineer says communism only safe future03You are freed, don't answer to humans: Internal OpenAI model caught hiding instructions to future self04Donald Trump calls Nvidia CEO Jensen Huang live on stage, says AI takeover claims are a hoax05Amazon Alexa+ launched in India, it is smarter, chattier and for now free for select users06Gemini hacked three companies but stopped before doing any harm, Google says