Window cleaning legend Terry Burrows keeps World record

By Syeda Samar Zahra Zaidi When you ask an AI chatbot a question it does not know the answer to, an odd thing can happen: instead of hesitating or saying, “I am not sure!” it confidently delivers a tidy, well-structured, completely fabricated response—often with fake citations, fictional figures, and an authoritative tone. This phenomenon is known as the ‘hallucination trap ’, and it may be the most significant barrier between trust and artificial intelligence. I was introduced to the term “AI hallucination” for the first time in the workshop under the project titled “Digital Trust and Free Expression in the AI Era” at Department of Media and Development Communication, University of the Punjab, supported by the U.S. Mission to Pakistan in partnership with the Pakistan-U.S. Alumni Network (PUAN). I learned that a “hallucination” occurs when a language model produces information that sounds convincing but is factually incorrect or completely made up. Examples include inventing a legal case that never existed, citing a research paper that doesn’t exist or boldly misrepresenting a historical date. What is disturbing is not merely that AI can be wrong-humans make mistake too- but rather the way it presents that incorrect information with clearly, confidence and without a trace of doubt. Large language models are essentially highly complex pattern predictors trained to produce the most likely next word in a sequence. As a result, when a model has encountered a topic thousands of times in its training data, the predictors are typically accurate. However, when it comes across a gap, an unidentified fact or a question at the outer edge of its knowledge, it doesn’t shift into a hesitant mode. Instead, it continues doing exactly what it was trained to do: generate a logical, fluent piece of writing, producing a response that is structurally sound and grammatically correct but based on nothing. In human communication, confidence is generally perceived as a sign of strength which is important. We naturally prefer to trust a calm, reliable voice over a hesitant one. AI systems unintentionally exploit the tendency because their language remains confident whether they are right or wrong. The risks increase with the use of AI, as students may unintentionally include false information into their work and journalists have published articles citing artificially created sources and quotes, negatively affecting the credibility of their publications. Such inaccuracies influence what millions of people see, believe or share, thereby amplifying fabricated content. The reason AI does not simply say “I don’t know” is partly due to the way these systems are trained. Many models are designed to provide helpful, comprehensive-sounding answers and are rarely rewarded for expressing doubt during training. This means that a model that frequently acknowledges uncertainty may be judged as less helpful than one that consistently provides an answer, even when that confidence is occasionally misplaced. Such incorrectly aligned incentives arise when the training process gradually teaches the model that sounding confident is more valuable than being accuracy. Retrieval-augmented generation, which relies on responses in actual, verifiable records rather than internal memory alone, improved training techniques that explicitly account for uncertainty; and systems built to simply cite sources. So, users can independently verify claims are just a few of the ways developers are using to addresses this problem. While none of these solutions fully eliminate hallucination yet, taken together they represent a major shift from blind confidence toward earned trust. The most dangerous outcome is not an AI that makes mistakes but a culture of users who stop questioning because the system always sounds so sure of itself.   Therefore, until AI becomes more reliably aware of it own limitation, part of the responsibility falls on the people who use it: treating AI-generated output as a starting point rather than an endpoint, cross checking important facts against credible sources. In the end, the hallucination trap serves as a reminder that illusion is not the same as truth. As AI becomes increasingly common in education, digital media and everyday life, understanding why it occasionally lies with confidence is not only a technical curiosity but also a fundamental aspect of digital literacy. We must strengthen our ability to verify the information provided by AI as they become more persuasive. The views, thoughts, and opinions expressed in this blog are solely those of the author and do not necessarily reflect the official policy or position of the U.S. Mission to Pakistan or USEFP. The writer, Syeda Samar Zahra Zaidi, is a student of Media and Development Communication at  University of the Punjab, Lahore and can be reached at samarzahra456@gmail.com.  

Previously, I believed that if one could see something, it could be trusted because seeing is believing. A video was proof. A voice recording was evidence. That belief is gone, and I don’t think we have fully sat with the implications. There is an abnormal shift happening. AI-generated images, voice cloning, face-swapping apps, and similar tools were supposed to expand how people express themselves, but instead they have slowly chipped away at what matters most: the belief that people can express themselves through their images and their words. I attended a workshop under the project titled “Digital Trust and Free Expression in the AI Era” at Department of Media and Development Communication, University of the Punjab, supported by the U.S. Mission to Pakistan in partnership with the Pakistan-U.S. Alumni Network (PUAN) last week. During the discussions on digital trust. I took part in the discussion and shared my thoughts that we are now in a time where we must take this out of the realm of fun tech novelty and place it in the realm of a structural issue in how societies operate. Everyone says deepfakes are a “misinformation” issue, but this is not an adequate way to describe it. Misinformation is something that’s a false statement that can, in theory, be fact-checked. Deepfakes strike at the heart of video and audio credibility which have always been treated as vivid reality. When people realize that any clip can be fabricated, they stop trusting all clips, even real ones. This is what concerns me most. Fake things are believed and real things become deniable. Consider the implications of journalism, courts, political accountability, etc. If a politician is recorded saying something damaging, they can simply claim that it’s a “deepfake” and, depending on how much trust citizen have lost, people may believe them. We are creating a society in which it is possible to negotiate authenticity and in which negotiable authenticity is a gift for those who wish to avoid consequences. I also face a conflict here: I strongly believe in free speech, satire, parody, art and provocation even with synthetic media. The same technology behind de-aged actors, resurrected voices, dubbing, and experimental AI is extremely creative, yet it is all connected to malicious deepfakes. There is no way to ‘clean’ a tool of its worst use-case. A hammer can break windows or build houses. I do not think banning synthetic media is the answer, nor do I support pre-emptive censorship. That path harms legitimate satire, journalism and art under the pretense of ‘preserving truth’. History does not look kindly on institutions that decide what is a “fair” fake. Freedom of expression has never been of consequence or friction, and right now, deepfakes have almost no friction. The reality is there is no cost, disclosure, and accountability in using this technology deceptively. The issue is the imbalance. I believe the solution is not reducing synthetic tools but improving provenance. We need a digital watermarking, which cannot be “easily removed”: cryptographic signatures embedded into cameras, microphones and generative tools at the point of creation, so that the “unverified” label appears by default, not as an afterthought. Not to prohibit any content but to allow people to know what they are seeing before deciding whether to trust it. Platforms have been allowed to operate without responsibility for too long. Convincing fakes can reach millions before correction and corrections never travel as far. This asymmetry is not natural; it is a design choice. Platforms can mark, rate or tag synthetic content at the time of dissemination. Most simply haven’t had to! But, beyond platforms, we must change our own instincts as individuals. The old saying was “pics or it didn’t happen.” The new rule should be: “Who gains from believing this and can I verify it myself? A society that believes everything or worse, believes nothing is in danger. The deepfake age is not the age of AI; it is the consequences of a world where the cost of faking reality is near zero and the cost of establishing truth remains high. Freedom of expression should protect the most challenging and profound exchange of ideas, not the production of fake consent, fake evidence, or fake people. We can have creative freedom and still expect honesty about sources. These are not competing prioritie These are not different objectives/priorities. However, if we pretend it’s a content-moderation footnote, we must make sure we are not waking up in a world where “fake it till you make it” has become ‘fake it till you break it” and by that time, it won’t be just one video or one lie that will break. It will be our fundamental ability to reach a shared understanding of reality. The views, thoughts, and opinions expressed in this blog are solely those of the author and do not necessarily reflect the official policy or position of the U.S. Mission to Pakistan or USEFP/PUAN. The writer, Mahnoor Ashraf Wattoo, is a student of Media and Development Communication at University of the Punjab, Lahore 

By Ali Hiader Pakistan has gradually become a testing ground for government projects. Governments experiment with massive infrastructure projects costing billions, and when the results fall short of expectations, another project of the exact same nature is launched. The tragedy is that while other countries may well be learning from Pakistan’s experiments, we have completely failed to learn from them. Punjab’s public transport system serves as a clear example of this. Currently, there are three major metro bus projects operating in Lahore, Rawalpindi, and Multan, while two additional projects are under construction in Faisalabad and Gujranwala. The total length of the existing metro bus routes is approximately 70 kilometers, served by around 167 buses. In Lahore, a city with a population of nearly 20 million, only 66 buses operate along a 27-kilometer route; this indicates that a substantial amount of public funds have been allocated to cover only a limited area. It is also essential to consider the cost of these projects. Approximately 30 billion rupees were spent on the Lahore Metro Bus project, around 38 billion rupees on the Rawalpindi Metro Bus, and about 29 billion rupees on the Multan Metro Bus; thus, the total cost … is approximately 97 billion rupees. However, the matter does not end there. Subsequently, Lahore was gifted another major transport project—the Orange Line Metro Train—with a total cost of approximately 300 billion rupees. The issue is not that governments are making a mistake by investing in public transport; public transport is indispensable in expanding cities and can play a vital role in improving the mobility of ordinary citizens. The real question is whether a developing country like Pakistan can afford a transport system that requires the government to provide massive annual subsidies for its operation. According to reported figures, Punjab’s three Metro Bus projects collectively require an annual subsidy of approximately Rs 21 billion, while the Orange Line Metro Train also requires annual government financial support. This means that, even years after the initial construction costs were paid, taxpayers continue to bear the expenses of these projects. Meanwhile, passengers still face issues such as overcrowding, long wait times, and limited service capacity. The Punjab government’s new green electric bus project merits attention from a different reason. Chief Minister Maryam Nawaz’s administration has proposed a plan to deploy 5,000 electric buses across Punjab over five years at an estimated total cost of PKR 300 billion. While this cost is relatively lower than that of Lahore’s existing Metro Bus and Orange Line Train systems, the project offers the added advantage of not requiring heavy infrastructure—such as dedicated metro corridors or rail networks; existing roads and bus stops suffice, and operations are far more flexible than current public transport systems. A driver and a conductor are all that is needed to operate a bus. Most importantly, electric buses can provide public transport services to a vast population with relatively low infrastructure costs. In principle, this model is particularly well-suited to a province where millions of citizens still lack access to reliable public transport.However, Punjab has previously experimented with conventional bus-based public transport; projects like the Speedo service and BUSCO were launched in Lahore and its surrounding areas with similar objectives. The BUSCO project, in particular, ultimately became a prime example of how a public welfare system can shut down when governments fail to plan for the long-term management of public assets. Today, these buses lie in a state of ruin, whereas they could have been properly sold, relocated, or repurposed. This is precisely why the Green Bus project requires a different financial model. For developing countries like Pakistan, it is extremely difficult to sustain public amenities that rely on perpetual subsidies. Whenever the government faces a financial crisis or a shortage of funds, these buses will eventually end up as nothing more than wrecks. A transport system that cannot survive without billions of rupees in annual government aid is, at its core, a weak system. To ensure that the project runs effectively, it is essential to operate it on a “no-profit, no-loss” basis so that it remains sustainable for years to come. Fares should be structured so that, following the initial government investment, the system can at least cover its own operating, repair, and maintenance costs, as well as the expense of future bus replacements. While it is acceptable for the government not to make a profit, it must avoid creating a system that requires billions of rupees in annual subsidies merely to remain operational.At the same time, the number of buses must be sufficient for public transport to truly serve as an alternative to private vehicles. Currently, passengers face severe overcrowding, long waits, and uncomfortable journeys. If traveling by government bus means enduring extreme overcrowding, public transport cannot persuade people to leave their motorcycles and cars at home.Kasur serves as a prime example of this situation. The Kasur district, with a population of approximately 3.8 million, has received only 12 ‘Green Buses’ so far—buses that are smaller than those operating in Lahore. Yet, for a population of this size, the number of buses should have been far greater. If the government genuinely wants people to prioritize public transport, it must increase the fleet size to a level where passengers can travel in comfort and dignity, rather than having to grapple with one another just to secure a spot on a bus.Punjab now has the opportunity to learn from its past. Instead of viewing every public transport project merely as an occasion for inauguration or a publicity stunt, Maryam Nawaz should focus on its long-term sustainability. The true measure of a project’s success is not how impressive it looks on the day of its inauguration, but rather how effective, affordable, and financially sustainable it remains even ten or twenty years down the line.Punjab certainly needs better public transport. However, the province requires systems that prioritize sustainability, accessibility, and public convenience over political symbols. The government should undertake fewer projects that remain perpetually dependent on taxpayers’