 New Delhi, (IANS): At least 89 per cent of Indian technology service enterprises have initiated generative AI (GenAI) proof-of-concept (POC) projects, with 33 per cent already in production, according to a report on Monday. Artificial Intelligence (AI) is fundamentally reshaping India's software services industry, driving innovation, efficiency, and new business models, said EY India in its report. AI investments also continue to surge as 78 per cent of organisations have dedicated budgets and strategies for GenAI implementation. Technology services sector is poised to witness a 43-45 per cent productivity boost across more than 500 roles by 2030 and 44 per cent of those surveyed recognize AI’s role in enhancing customer satisfaction while 28 per cent link it to revenue growth, the report noted. “Enterprises are moving beyond experimentation to putting AI into production at scale. The rapid transition from POC to enterprise-wide adoption reflects the industry’s confidence in AI’s potential,” said Abhinav Johri, Technology Consulting Partner, EY India. Scaling AI responsibly through robust infrastructure, governance, and talent development will be key to unlocking its full economic impact and strengthening India's competitive edge in the digital era, he added. The technology services sector is poised to witness a 43-45 per cent productivity boost across more than 500 roles by 2030. Roles in software development is expected to see approximately 60 per cent productivity improvement, while those in the BPO services and IT consulting could witness around 52 per cent and 47 per cent improvement, respectively. Collectively, these three areas are anticipated to account for 50 per cent to 60 per cent of overall productivity enhancement in technology services.Furthermore, organisations have prioritised specific areas for GenAI implementation -- around 56 per cent of companies focus on development, while 39 per cent emphasise operations, 28 per cent prioritise quality testing, whereas 22 per cent concentrate on designing and prototyping as well as integration and deployment each, the report mentioned. 89 pc of Indian tech service enterprises begin GenAI proof-of-concept projects | MorungExpress | morungexpress.com |
89 pc of Indian tech service enterprises begin GenAI proof-of-concept projects (2025-09-22T13:24:00+05:30)
Designing for impact: How product innovation is enabling scalable, intelligent infrastructure (2025-09-15T11:53:00+05:30)
 Posted by Harry Baldock, Contributed Article: By Pete England, Product Director, ITS  For decades, the business connectivity market has been defined by compromise. Organisations were forced to choose between two technologies: copper, cheap but unreliable, or leased lines, reliable but unaffordable for most businesses. Neither option truly reflected what businesses needed. Copper couldn’t deliver consistency. Leased lines often meant overpaying for bandwidth that was only essential during peak usage. Connectivity was a limiting factor, not an enabler. Today, that dynamic is changing. As the UK fibre rollout shifts into a “fibre utility” phase, the opportunity lies not just in building networks, but in designing products that match the way businesses and public services now operate. A market transformed by digital society Every aspect of modern work is underpinned by reliable connectivity. From cloud-based CCTV and IP-enabled cameras, to collaboration platforms like Teams and SharePoint, to guest Wi-Fi and cashless point-of-sale terminals – connectivity is no longer a utility, but the foundation of digital society. This creates new demands:Scalability, to grow as application use increases. Reliability, because downtime is no longer an option. Security, as every connection point becomes a potential risk. Traditional products can’t keep pace. Many businesses find themselves paying for capacity they only need at peak times, or trying to stretch inadequate services across mission-critical applications. As digital convergence accelerates, those compromises are no longer acceptable. That’s why product innovation has to move faster than traditional providers can adapt. At ITS, we’ve made it our mission to design services that evolve with applications, not against them. Smarter by design Connectivity should not be a hard-coded, one-size-fits-all commodity. It should be agile, consumable, and customisable. Our modular product framework gives partners and customers the ability to:Shape solutions for specific needs, sectors, and use cases. Flex bandwidth, service levels, and commercial models. Integrate seamlessly with partner platforms for quoting, ordering, and lifecycle management. This approach removes the perennial issue of overselling. Instead of locking customers into a leased line at 1Gbps when they only need 100Mbps, our products can be tuned and scaled remotely. No truck roll, no wasted spend – just a right-sized service that can grow as demand grows. We’ve also invested heavily in rapid provisioning through our partner portal. By cutting fibre lead times and enabling full integration with partner systems, we’re giving resellers the tools to compete harder in a market where speed of delivery is a true differentiator. This isn’t just infrastructure. It’s infrastructure designed for outcomes. Closing the gaps: A three-tiered portfolio Our product ladder has been deliberately shaped to address every segment of the business market, without forcing compromise.FibreOne – A reliable entry-level FTTP business broadband service. Perfect for organisations with basic connectivity needs, such as payment terminals or internet radio. It runs at consistent speeds and provides a clear upgrade path when requirements grow. FibreLight – The missing middle ground. Businesses that outgrow FTTP shouldn’t be forced into costly leased lines. FibreLight fills this gap, offering scalability, reliability, and performance at a price point that makes sense. It’s a commercial innovation that allows customers to unlock cloud-based applications without overspending. In many ways it’s the “EFM killer” – removing the awkward choice between too little and too much by creating a product that truly sits between. FibreBright – For those who demand absolute certainty. FibreBright is our premium leased line service, delivering dedicated, uncontended performance around the clock. It supports customers that require mission-critical resilience, with clear paths to higher speeds and optical services. What makes this portfolio powerful is not just the breadth of choice, but the designed-in upgrade paths. Businesses can start small, grow, and adapt without having to rip and replace, often as the competition needs to. Partners, meanwhile, can approach their customers with a suite of options that are both commercially compelling and technically future-proof. Looking ahead: From connectivity to smart infrastructure The next phase of our innovation journey is about making connectivity smarter, not just faster. That means embedding intelligence, security, and automation into the core of our services. It means designing for compliance and ESG goals, not as bolt-ons but as part of the network’s DNA. It also means recognising that, in a market where competitors still build siloed services, ITS has done something different. We’ve created a singular, software-configurable network that delivers economies of scale and allows for modularity at the service-spec level. In practice, this means:Care packages and service levels can be added like software modules. Speeds can be flexed remotely, without a site visit. Capacity can be reused rather than wasted. This is the connective tissue that enables partners and end customers to consume connectivity more like a cloud service: buy what you want, when you want it, and flex it as your needs change. Redefining the market Connectivity isn’t “done.” The market doesn’t need more commodity bandwidth layered with bolt-on services. It needs connectivity that is inherently better: faster, smarter, more flexible, and ultimately more affordable. By understanding the traditional pain points and addressing them with flexible, modular design, ITS is creating a portfolio that sets a new benchmark. It’s not about competing with incumbents on their terms. It’s about redefining the terms altogether – closing market gaps, making connectivity more equitable, and creating infrastructure that’s ready for the next decade of digital society. Want to learn more about the future of smart infrastructure? Stop by for a drink at the Dog & Duct pub at Connected Britain, sponsored by ITS g for impact: How product innovation is enabling scalable, intelligent infrastructure | Total Telecom. Designing for impact: How product innovation is enabling scalable, intelligent infrastructure |
High-tech plans to save polar ice will fail, new research finds (2025-09-11T14:06:00+05:30)
|
Steven Chown, Monash University: Our planet continues to warm because of greenhouse gas emissions from human activities. The polar regions are especially vulnerable to this warming. Sea ice extent is already declining in both the Arctic and Antarctic. The Greenland and Antarctic ice sheets are melting, and abrupt changes in both polar environments are underway.
These changes have significant implications for society through sea level rise, changes to ocean circulation and climate extremes. They also have substantial consequences for polar ecosystems, including polar bears and emperor penguins, which have become iconic symbols of the impacts of climate change. The most effective way to mitigate these changes, and lower the risk of widespread impacts, is reducing greenhouse gas emissions. Yet decarbonisation is slow, and current projections suggest temperature increases of roughly 3°C by 2100. Given the expected change, and the importance of the polar regions for planetary health, some scientists and engineers have proposed technological approaches, known as geoengineering, to soften the blow to the Arctic and Antarctic. In research published today in Frontiers in Science, my colleagues and I assessed five of the most developed geoengineering concepts being considered for the polar regions. We found none of them should be used in the coming decades. They are extremely unlikely to mitigate the effects of global warming in polar regions, and are likely to have serious adverse and unintended consequences. What is polar geoengineering?Geoengineering encompasses a wide range of ideas for deliberate large-scale attempts to modify Earth’s climate. The two broadest classes involve removing carbon dioxide from the atmosphere and increasing the amount of sunlight reflected back into space (known as “solar radiation modification”). For the polar regions, here are the five most developed concepts. Stratospheric aerosol injection is a solar radiation modification approach that involves introducing finer particles (such as sulphur dioxide or titanium dioxide) into the stratosphere to reflect sunlight back out to space. In this case, the focus is specifically on the polar regions. Sea curtains are flexible, buoyant structures anchored to the seafloor at 700 metres to 1,000m depth and rising 150m to 500m. The aim is to prevent warm ocean water from reaching and melting ice shelves (floating extensions of ice that slow the movement of ice from Greenland and Antarctica into the ocean) and the grounding lines of ice sheets (where the land, ice sheet and ocean meet). Basal water removal targets the ice streams found in the Antarctic and Greenland ice sheets. These streams are fast-moving rivers of ice that flow toward the coast, where they can enter the ocean and raise sea levels. Water at their base acts as a lubricant. This concept proposes to remove water from their base to increase friction and slow the flow. The concept is thought to be especially relevant to Antarctica, which has much less surface melting than Greenland, and therefore melt is more about the base of the ice sheet than its surface. Ocean fertilisation involves adding nutrients such as iron to polar oceans to promote the growth of phytoplankton. These tiny creatures absorb carbon dioxide from the atmosphere, which gets stored in the deep ocean when they die and sink. In our research, we assessed each of these concepts against six criteria. These included: scope of implementation; feasibility; financial costs; effectiveness; environmental risks; and governance challenges. This framework offers an objective way of assessing all such concepts for their merits. None of the proposed polar geoengineering concepts passed scrutiny as concepts that are workable over the coming decades. The criteria we used show each of the concepts faces multiple difficulties. For example, to cover 10% of the Arctic Ocean with pumps to deliver seawater to freeze within ten years, one million pumps per year would need to be deployed. The estimated costs of sea curtains (US$1 billion per kilometre) are underestimates of similar-scale projects in easier environments, such as the Thames Barrier near London, by six to 25 times. One project that planned to spread glass microbeads on ice has also been shut down citing environmental risks. And at their most recent meeting, the majority of Antarctic Treaty Consultative Parties made clear their view that geoengineering should not be conducted in the region. Polar geoengineering proposals raise false hopes for averting some disastrous consequences of climate change without rapidly cutting greenhouse gas emissions. They risk encouraging complacency about the urgency of achieving net zero emissions by 2050 or may be used by powerful actors as an excuse to justify continued emissions. The climate crisis is a crisis. Over the time available, efforts are best focused on decarbonisation. The benefits are rapidly realisable within the near term. Steven Chown, Director, Securing Antarctica's Environmental Future and Professor of Biological Sciences, Monash University This article is republished from The Conversation under a Creative Commons license. Read the original article. |
‘A Beautiful Idea’: This French Town Is Making its Cemetery a Source of Solar Energy (2025-08-11T11:31:00+05:30)
 Copyright Ville de Saint-JoachimBy Andy Corbley - Aug 11, 2025, A community on the Loire in France has come up with a solar-powered idea that will clear two hurdles in a single leap. The town of Saint-Joachim is located near the Brière marsh, a peat bog that becomes easily inundated with water. The local graveyard rests at 0 feet above sea level and standing water has become a major problem. The mayor eventually proposed covering the graveyard to divert the rainwater into catch tanks to water the grass around the cemetery and nearby sporting complex in the dry summer months. Solar panels—clear, see-through ones, were also proposed for the otherwise basic overhead covering, with each resident receiving a letter explaining the proposal and asking for their opinion on the idea. 97% of Saint-Joachim’s residents liked the concept and a power-sharing scheme was quickly determined whereby each resident would pay €5.00 for a share of the power generated there, which might be enough to supply around 1,000 people with enough solar power to save a couple hundred euro every year on their electric bills. The power company in charge of the project designed an algorithm that gathers data on energy usage and determines how much each home needs to meet a fixed percentile of their overall consumption. That way everyone, from a supermarket to the hairdresser to a single-family home, receives the same energy savings from the cemetery solar power. The municipality will fully finance the €3.35 million ($3.6 million), 1.2-acre installation, using a 7% property tax increase. “I think it is a beautiful idea,” Éric Broquaire, local resident and president of Brier’energie, told Euronews. “The purpose was to make it simple, to avoid someone saying ‘why don’t I have electricity for free’. Everybody, even companies, will have the same level.” ‘A Beautiful Idea’: This French Town Is Making its Cemetery a Source of Solar Energy |
Humane Innovates Wearable Device ‘AI Pin’ To Replace Smartphones (2025-06-21T12:25:00+05:30)
A wearable product named AI Pin is an innovative device that has been creating quite a buzz in the last 24 hours. AI Pin is essentially a small, wearable device, packed with advanced technology. The AI Pin, a groundbreaking wearable device, was launched by Humane, a startup based in San Francisco. This innovative company was founded by two former Apple employees. They have managed to secure backing from tech giants like Microsoft and OpenAI, the developer behind ChatGPT. The AI Pin, representing a significant leap in wearable technology, is Humane’s first product and showcases the potential to transform how we interact with technology in our daily lives. The AI Pin isn’t just any wearable tech. It stands out for its multifunctionality. It’s like having a personal assistant pinned to your clothing. With AI Pin, users can enjoy hands-free communication, access to information, and even interaction with other smart devices. It’s designed to understand and respond to voice commands, making it incredibly user-friendly. One of the biggest draws of AI Pin is its potential to make life easier. For starters, it can provide quick answers to questions, manage schedules, set reminders, and even help navigate places. Think about being in a new city and having a guide right with you, providing directions and local information without needing to pull out a phone. But can it replace smartphones? That’s a bit of a stretch at this point. While AI Pin offers convenience and ease of use, smartphones are still more versatile. They have larger screens, better multimedia capabilities, and a broader range of applications. However, AI Pin is a step towards a future where we might rely less on traditional gadgets like phones. As of now, AI Pin is a glimpse into the future of wearable technology. It’s not just about the convenience it offers but also about how it represents a shift in how we interact with technology. It’s a move towards more integration. With such innovations, it’s exciting to imagine what the future holds in the realm of technology and personal devices, Humane Innovates Wearable Device ‘AI Pin’ To Replace Smartphones
Visiting the Gym Today Could Trigger a Bright Idea Next Week, New Study Shows (2025-06-14T11:38:00+05:30)
 |
CES 2025: AI-powered tech to dominate world's biggest electronics trade show (2025-06-14T11:38:00+05:30)
 |
GenAI driving over 30 pc productivity gains for India's insurance industry (2025-06-14T11:37:00+05:30)
 New Delhi, (IANS) Artificial Intelligence (AI), especially Generative AI, is transforming the India's insurance industry this year, with productivity gains of over 30 per cent, a new report said on Friday. Insurers who are using AI in underwriting are seeing up to 36 per cent efficiency gains by using both structured and unstructured data more effectively, according to the report by Boston Consulting Group (BCG). In customer service, tools such as AI-powered knowledge assistants have led to over 30 per cent boosts in productivity, while also improving service quality. In claims processing, AI is helping resolve up to 70 per cent of simple claims in real-time, reducing costs by 30 to 50 per cent and offering a smoother experience for customers. Even in IT, AI is proving useful as smart automation tools are helping insurers cut their cloud migration timelines in half and save 30 per cent in costs, the report mentioned. Despite the growing potential of AI, the report found that many insurance companies are still stuck in pilot phases and have not fully scaled their AI projects. However, a few forward-looking insurers are using AI as a competitive advantage, particularly in underwriting, claims processing, customer service, and IT operations. Pallavi Malani, India Leader-Insurance Practice at BCG, said that GenAI is reshaping every part of the insurance business. She pointed out that while Indian insurers are experimenting with several AI-based proof of concepts, most of them have not yet scaled. “Particularly in India, we see that insurers are undertaking several proof of concepts but these use cases have not been scaled up,” Malani stated. To fully benefit from AI, companies must think beyond just technology and data, and include business impact, process changes, and employee readiness from day one. The report stressed that insurers who align their AI investments with business goals, and focus on high-impact areas, are pulling ahead of the competition. |
Why e-bikes can succeed where earlier bike-share schemes failed (2025-05-19T13:04:00+05:30)
Madison Bland, Griffith University; Abraham Leung, Griffith University, and Benjamin Kaufman, Griffith UniversityShared mobility devices such as bicycles and electric scooters have experienced significant growth across the globe and Australia is no exception. In cities with such offerings, users are able to get around in more convenient and flexible ways. The recent emergence of dockless shared e-scooters (i.e. Lime and Neuron) heralded a new-age of micromobility. In Brisbane, it signalled the end for the ten-year-old CityCycle bike-share scheme. Not long after announcing CityCycle’s demise in late 2020, Brisbane City Council proposed its replacement with shared dockless e-bikes and the topic started trending. The question is: why will the e-bike scheme succeed where its predecessors in Brisbane and other Australian cities failed? (See below for a summary of the evolution of shared mobility schemes in Australia.) Mobility is being offered more and more as a service. The uptake of share travel across Australian cities has undergone a transition from docked bikes to dockless e-mobility, aided largely by advances in technology and the proliferation of mobile devices. Sharing is being considered as an attractive alternative to owning a bike or car thanks to new ways to bundle mobility services into packages, in much the same way as we use entertainment streaming services instead of buying movies or records. What can we expect from e-bikes?E-bikes are pedal-assisted bicycles offering users electric motor assistance up to speeds of 25km/h. A shared bike scheme with self-locking and smartphone connectivity offers an extremely flexible riding experience. It isn’t yet clear how e-bikes will be deployed in Brisbane. What we do know is the scheme will be privately operated under a short-term tender. As with CityCycle, 2,000 bikes will be provided across Brisbane, similar to how e-scooters are managed. The e-bikes can improve on both e-scooters and CityCycle’s docked bikes in several ways. Trip flexibility: GPS tracking and smart lock technology remove the need to locate set docking stations. Users can start and end trips at places of their own choosing. This means they avoid the frustrations caused by docking stations reaching maximum capacity, especially in popular destinations such as the CBD. Wider appeal: unlike e-scooters and their younger target market, e-bikes can attract a wider demographic more familiar with riding bikes. They also offer greater load-carrying capacity and are permitted for use on roads whereas e-scooters are restricted to footpaths or bikeways in Brisbane. In New South Wales and Victoria, e-scooters are banned altogether - though changes could be on the way for Victoria. Assisted riding: electrically assisted bikes can make cycling easier and accessible for more people. For those who struggle to ride at the best of times, e-bikes can help overcome fitness issues, especially in Brisbane’s hot climate and hilly terrain. So, what punctured CityCycle?CityCycle was launched in 2010 under a 20-year single-operator contract. The scheme failed to achieve ambitious patronage targets and the goal of paying for itself. Despite usage growing until 2018, a shifting market has since resulted in significant declines. The reasons for the lack of use are clear:
The path to successAs Brisbane moves towards a dockless e-bike scheme, its ability to outperform its predecessor will ultimately rest with decision-makers delivering a safe and convenient rider experience. This involves several key considerations. Pricing and payment: the scheme will have to be competitive with current modes (particularly e-scooters), where registration and payment are integrated with existing systems. The rise of mobility as a service (MaaS) platforms can incorporate the service within shared mobility apps and bundle offers (packaging public transport and shared mobility services). Availability: the dockless model, while more flexible, will require operators to actively manage bike distribution and avoid cluttering. The blocking of access ways and even dumping of bikes have been sources of public opposition to other bike-share schemes. Though repositioning bikes (using service vehicles) will take up significant time and money, it is crucial in maintaining a balanced and orderly network that maximises bike availability. Initial launch: the scheme’s roll-out will be important, as positive perceptions are best achieved by people riding, rather than bikes sitting idle. Importantly, a winter launch should be avoided – as Melbourne found – when bike trips are at yearly lows. Cycle infrastructure: As with cycling in general, providing safe and connected bicycle networks is paramount for increasing participation rates. For Australian cities, the historic lack of funding for cycle infrastructure has limited ridership growth. Much work remains to be done, though Brisbane City Council has committed to trial improvements to its CBD on-road bike lanes. Ultimately, dockless shared e-bikes can deliver a more flexible mobility option as operators maximise user convenience and governments develop urban cycling infrastructure. * The regulatory environment for micromobility is rapidly evolving. There are many nuances across Australian jurisdictions, and users should check with their own state or territory for up-to-date road rules and regulations. Madison Bland, PhD Candidate, Cities Research Institute, Griffith University; Abraham Leung, Transport Academic Partnership (TAP) Postdoctoral Research Fellow, Cities Research Institute, Griffith University, and Benjamin Kaufman, PhD Candidate, Cities Research Institute, Griffith University This article is republished from The Conversation under a Creative Commons license. Read the original article. |
Explainer: Theory of evolution (2025-04-03T14:22:00+05:30)
|
The theory of evolution encompasses the well established scientific view that organic life on our planet has changed over long periods of time and continues to change by a process known as natural selection. Charles Darwin, the 19th century naturalist, is given credit for the theory, not because he was the first person to suggest evolution occurs, but because he proposed (in his seminal 1859 text, On the Origin of Species) a mechanism that explains the process of change. The theory of evolution contains two parts, both of which are unnecessarily contentious. The first is the word “theory”, which means something slightly different in everyday speech than it does in science. The second contentious word is “evolution”, because some people argue that there is insufficient evidence to support the idea that species change over time. Proponents of the latter view rely on our dual uses of the word theory to confuse the issues. Theory When most people use the word “theory”, they refer to a well argued bit of guesswork. Someone might propose a theory about why a football team lost, which could lead to a spirited debate with other football fans precisely because the theories under discussion derive from mere speculations and untested assumptions. When scientists use the word theory, they refer to a group of principles or laws developed over many years through rigorous hypothesis testing. Such theories are backed up with mathematical formulas and lines of evidence that when taken together explain a series of observations. Gravity is a fact because it summarises a number of observations that can be made by anyone. Despite the fact there are different theories explaining gravity (Newton’s law, Einstein’s theory of relativity and more recently, quantum field theory), no-one would say gravity does not exist. Similarly, biological evolution is supported by so much evidence from various scientific disciplines that it is a fact every bit as true as gravity. The theory of evolution, therefore, is not about whether or not evolution occurs, but how it occurs. In the same way that our understanding of gravity has changed over time, our understanding of the process of evolution has changed over time and continues to change. Natural selection There are two major components to the theory. The first is the mechanism of natural selection, which was proposed independently by Charles Darwin and Alfred Russell Wallace in 1858. In the same way artificial selection occurs when breeders choose seeds or studs that will improve their stock in the next generation, natural selection is the process of sorting living things according to how well adapted they are to their environment. In the case of artificial selection, humans choose which traits are desirable. In the case of natural selection, traits that increase the likelihood of survival and reproduction will become more common within a population or species over time. In the past, natural selection has been misrepresented by calling it the survival of the fittest. This statement oversimplifies the mechanism by making it sound like a tautology: the survival of those who survive. The truth is that individuals never survive. What survives is the process for making another individual, and this resides in genes found in populations. Natural selection has more to do with differential reproduction than survival, and what it selects are the genes that code for desirable traits or characteristics. The interaction of individuals with their environment provides a mechanism for sorting out which traits (not which individuals) will be passed on to the next generation. Nature of inheritance The second major component to the theory is the nature of inheritance, which follows the insights made by Gregor Mendel in 1865 and has advanced considerably since then due to our understanding of genes, DNA and the molecular processes of life. When natural selection was first formulated by Darwin, the nature of inheritance was not understood. Our current understanding of inheritance is very sophisticated and includes the precise mechanisms for passing genes on to the next generation, how genes are modified by mutation and how they are shared among sexual species. If we know enough about a gene and its various forms, it is possible to accurately predict the change in the frequency of those genes over time using mathematical formulae from population and evolutionary genetics theory. This alteration of gene frequencies is subtle and does not, at first glance, seem worthy of being called evolution. But it is precisely these small changes at the genetic level that lead to large changes in the organisms that carry them. The sorting of genes affects the fate of populations: populations drift apart and become species, and species diverge to create whole groups of plants or animals that dominate the landscape for millions of years. The intricate details of cellular processes are responsible for the glorious and majestic diversity of life on our planet. The theory of evolution includes large changes over vast periods of time and tiny changes made when one cell divides into two. These processes form a continuum that is the history of life on Earth. What element of science or technology would you like to see explained? Contact the section editor here.  Susan Lawler, Head of Department, Department of Environmental Management & Ecology, La Trobe University This article is republished from The Conversation under a Creative Commons license. Read the original article. |
AI doesn’t really ‘learn’ – and knowing why will help you use it more responsibly (2025-03-11T13:06:00+05:30)
|
What if we told you that artificial intelligence (AI) systems such as ChatGPT don’t actually learn? Many people we talk to are genuinely surprised to hear this. Even AI systems themselves will often tell you confidently that they are learning systems. Many reports and even academic papers say the same. But this is due to a misconception – or rather a loose understanding of what we mean by “learning” in AI. Yet, understanding more precisely how and when AI systems learn (and when they don’t) will make you a more productive and more responsible user of AI. AI does not learn – at least not like humans doMany misconceptions around AI stem from using words that have a certain meaning when applied to humans, such as learning. We know how humans learn, because we do it all the time. We have experiences; we do something that fails; we encounter something new; we read something surprising; and thus we remember, we update or change the way we do things. This is not how AI systems learn. There are two main differences. Firstly, AI systems do not learn from any specific experiences, which would allow them to understand things the way we humans do. Rather they “learn” by encoding patterns from vast amounts data – using mathematics alone. This happens during the training process, when they are built. Take large language models, such as GPT-4, the technology that powers ChatGPT. In a nutshell, it learns by encoding mathematical relationships between words (actually, tokens), with the aim to make predictions about what text goes with what other text. These relationships are extracted from vast amounts of data and encoded during a computationally intensive training phase. This form of “learning” is obviously very different to how humans learn. It has certain downsides in that AI often struggles with simple commonsense knowledge about the world that humans naturally learn by just living in the world. But AI training is also incredibly powerful, because large language models have “seen” text at a scale far beyond what any human can comprehend. That’s why these systems are so useful with language-based tasks, such as writing, summarising, coding, or conversing. The fact these systems don’t learn like us, but at a vast scale, makes them all-rounders in the kinds of things they do excel at. Once trained, the learning stops
Most AI systems that most people use, such as ChatGPT, also do not learn once they are built. You could say AI systems don’t learn at all – training is just how they’re built, it’s not how they work. The “P” in GPT literally stands for “pre-trained”. In technical terms, AI systems such as ChatGPT only engage in “training-time learning”, as part of their development, not in “run-time learning”. Systems that learn as they go do exist. But they are typically confined to a single task, for example your Netflix algorithm recommending what to watch. Once it’s done, it’s done, as the saying goes. Being “pre-trained” means large language models are always stuck in time. Any updates to their training data require highly costly retraining, or at least so-called fine-tuning for smaller adjustments. That means ChatGPT does not learn from your prompts on an ongoing basis. And out of the box, a large language model does not remember anything. It holds in its memory only whatever occurs in a single chat session. Close the window, or start a new session, and it’s a clean sheet every time. There are ways around this, such as storing information about the user, but they are achieved at the application level; the AI model itself does not learn and remains unchanged until retrained (more on that in a moment). What does this mean for users?
First, be aware of what you get from your AI assistant. Learning from text data means systems such as ChatGPT are language models, not knowledge models. While it is truly amazing how much knowledge gets encoded via the mathematical training process, these models are not always reliable when asked knowledge questions. Their real strength is working with language. And don’t be surprised when responses contain outdated information given they are frozen in time, or that ChatGPT does not remember any facts you tell it. The good news is AI developers have come up with some clever workarounds. For example, some versions of ChatGPT are now connected to the internet. To provide you with more timely information they might perform a web search and insert the result into your prompt before generating the response. Another workaround is that AI systems can now remember things about you to personalise their responses. But this is done with a trick. It is not that the large language model itself learns or updates itself in real time. The information about you is stored in a separate database and is inserted into the prompt each time in ways that remain invisible. But it still means that you can’t correct the model when it gets something wrong (or teach it a fact), which it would remember to correct its answers for other users. The model can be personalised to an extent, but it still does not learn on the fly. Users who understand how exactly AI learns – or doesn’t – will invest more in developing effective prompting strategies, and treat the AI as an assistant – one that always needs checking. Let the AI assist you. But make sure you do the learning, prompt by prompt. Kai Riemer, Professor of Information Technology and Organisation, University of Sydney and Sandra Peter, Director of Sydney Executive Plus, University of Sydney This article is republished from The Conversation under a Creative Commons license. Read the original article. |
Over 90 new products launched in first two days of Bharat Mobility Global Expo 2025 (2025-01-22T13:53:00+05:30)
 New Delhi, (IANS): The second day of Bharat Mobility Global Expo 2025 witnessed the launch of as many as 56 new products from 22 different brands at Bharat Mandapam in Pragati Maidan and 5 launches at Yashobhoomi, Dwarka in the national capital on Sunday, according to a Commerce and Industry Ministry statement. This takes the total number of launches to over 90 during the first two days at Bharat Mandapam, the statement said. Various advanced vehicles, cutting-edge mobility solutions, and the latest components were showcased ranging from super bikes, cars, to buses and even ambulances, the statement said. On the second day at Bharat Mandapam, VinFast Auto India announced the launch of its first two electric vehicles for the Indian market, the all-electric Premium SUVs, the VF 7 and VF 6. BMW India launched the all-new MINI Cooper S John Cooper Works Pack at an ex-showroom price of Rs 55,90,000 and BMW X3 at an ex-showroom price of Rs 75,80,000 – Rs 77,80,000. Additionally, BMW Motorrad India launched two new bikes in the Indian market, the new BMW S 1000 RR and the BMW R 1300 GS Adventure. The S 1000 RR super sport bike will be available in India at an introductory ex-showroom price starting from Rs 21,10,000. While the R 1300 GS Adventure (Base) introductory ex-showroom price will start from Rs 22,95,000. India's first solar electric car 'Eva' was also launched on Sunday by Vayve Mobility, at an introductory price of Rs 3.25 lakh (ex-showroom). The solar electric car will be available in three different options -- 9 kilowatt-hour (Whr), 12 kWhr and 18 k Whr with prices ranging between Rs 3.25 lakh (ex-showroom) and Rs 5.99 lakh (ex-showroom). JSW MG Motor India showcased 9 advanced models, launching MG Majestor. The showcased products include Hybrid Electric Vehicles (HEVs), Plug-in Hybrid Electric Vehicles (PHEVs), Battery Electric Vehicles (BEVs), and Internal Combustion Engine (ICE). Highlights included the IM5 sedan, IM6 electric SUV, MG HS PHEV, and MG7 Trophy Edition, showcasing cutting-edge technology and sustainable mobility. Eka Mobility showcased India's largest ever range of electric commercial vehicles, which includes over 11 distinct platforms spanning electric buses, trucks, and small commercial vehicles (SCVs). Launching the brand new EKA- Connect for the Indian market. Their bus portfolio included EKA COACH, EKA 12M, EKA 9M, EKA LF (Low Floor), and EKA 9M. Electric Trucks include EKA 55T and EKA 7T, while EKA 3.5T, 2.5T, 1.5T, EKA 3W CARGO, EKA 6S, and EKA 3S were unveiled under their SCV Range. BYD India launched BYD SEALION 7 Pure Performance eSUV at the Bharat Mobility Global Expo 2025. Additionally, BYD India also showcased the BYD SEALION 6, BYD Super Plug-in Hybrid EV with DM-i technology, and the stunning Yangwang U8. JBM Electric Vehicles launched 4 all-new electric buses on the second day of the Auto Expo 2025, ranging from luxury coach, and medical mobile unit to electric tarmac coach, among others. Highlights of the launch were Galaxy; electric luxury coach, Xpress; an intercity bus, e-MediLife; Low Floor Electric Medical Mobile Unit, and e-SkyLife; 9-meter electric tarmac coach. Montra Electric (TI Clean Mobility) launched ‘EVIATOR’ (e SCV) and Super Cargo (e 3-Wheeler) in the presence of the entire leadership team of the brand. Hyundai Motor Company launched 2 concept models of advanced electric three-wheeler and micro four-wheeler on the second day. Along with exploring contributions to the last-mile mobility market in India together with TVS Motor Company Ltd. (TVS Motor). Pinnacle Industries launched its next-generation ambulance range at the Bharat Mobility Global Expo 2025. The new range includes three cutting-edge ambulance models: AD-Gen Ultra, AD-Gen Ambulance, and Neonatal Ambulance. Numeros Motors on Sunday launched its multipurpose and reliable e-scooter 'Diplos Max' at an introductory ex-showroom price (Bangalore) of Rs 1,09,999, including the PM e-drive scheme. Alongside, the company unveiled another unique platform which will be India's first Bike-Scooter Crossover. SML Isuzu launched the Hiroi.EV while presenting 4 other products including AASAI MX, Premium Hiroi Bus, ATS-125 Multi stretcher ambulance, and the Samrat XT Plus Tipper. Cummins Group in India ("Cummins") announced the launch of its next-generation HELM™ (Higher Efficiency, Lower emissions, Multiple fuels) engine platforms, with the high-performance L10 engine, along with, an advanced Hydrogen Fuel Delivery System (FDS) with Type IV on-vehicle storage vessels and the innovative B6.7N natural gas engine. Godawari Electric Motors Pvt. Ltd. unveiled new additions to its product portfolio at the Bharat Mobility Global Show 2025. The company unveiled the Eblu Feo Z, Eblu Feo DX, and launched the Eblu Rozee ECO. The Eblu Rozee ECO is priced at INR 2,95,999/- (ex-showroom). To improve customer convenience, the company also announced the release of the EbluCare app. An innovative smartphone app called EbluCare has been designed to managing electric vehicles (EVs) easier. Sarla Aviation unveiled its first-ever product and India’s first flying taxi- Shunya on the second day. Motovolt Mobility unveiled a range of electric vehicles. The lineup includes the Hyper One, India’s first digital pedal motorbike, and the HUM NYC, a multi-utility speed pedelec tailored for demanding last-mile delivery needs. The versatile M7 and its sporty variant, M7 Rally, cater to both logistics and personal commuting, while the innovative CLIP, a portable e-bike conversion kit, enables seamless e-mobility for existing bicycles. Olectra Green Tech Limited unveiled a 12-meter Blade Battery Platform, a 9-meter City Bus, 12-meter Coach Bus, and Blade Battery Chassis. These products are underpinned by cutting-edge technology and inclusive design. Omega Seiki Pvt. Ltd. launched the M1KA 1.0 electric truck, priced at INR 6,99,000, at the Bharat Mobility 2025 exhibition. The company also unveiled the upcoming M1KA 3.0 model and presented the upgraded 2025 Stream City, a next-generation electric passenger vehicle.Bharat Mobility Global Expo 2025 is set to be a landmark event uniting the entire Indian automotive and mobility ecosystem under one roof. Scheduled from January 17th to 22nd, 2025, across three premier venues -- Bharat Mandapam and Yashobhoomi in Delhi and India Expo Mart in Greater Noida. The event was inaugurated by Prime Minister Narendra Modi at Bharat Mandapam. Over 90 new products launched in first two days of Bharat Mobility Global Expo 2025 | MorungExpress | morungexpress.com |