Companies are fine-tuning tools like ChatGPT, Bing Chat, Gemini, and DALL-E 2 to cut costs in customer service, content creation, and more. However, most businesses remain cautious, testing these tools rather than deploying them at scale. Challenges include hallucinations (generating false information and presenting it confidently as accurate), biases, intellectual property violations, high energy consumption, and uncertain returns on investment.
Goldman Sachs’ April report, GenAI: Too Much Spend, Too Little Benefit?, questioned the $1 trillion investment in AI infrastructure without clear benefits. Similarly, Gartner’s July report predicts that 30% of GenAI projects will be discontinued by 2025 due to poor data quality and escalating costs, which range from $5 million to $20 million.
In India, the AI market is expected to grow by 25-35% annually over the next three-four years, according to Nasscom and EY’s 2024 AI Adoption Index. While 75% of Indian organizations have AI strategies at the proof-of-concept (PoC) stage, only 40% are ready to move to production.
GenAI adoption is slow, particularly in legacy sectors such as energy, utilities, and manufacturing, according to the report. While banking and financial services (BFSI), retail, and CPG (consumer packaged goods) are in the PoC stages, manufacturing and telecom have moved beyond PoCs to collaborate with industry disruptors.
The report insists that to transition from AI-ready to AI-first, large enterprises should focus on data standardization, strategic partnerships, and balanced AI use. Small- and medium-sized businesses (SMBs), on their part, should emphasize agile PoC processes, partnerships with tech SMEs, and strong leadership commitment to overcome barriers.
To be sure, advances in LLMs’ reasoning capabilities, such as those demonstrated by OpenAI’s o1 and o3 models and Gemini 2.0, do impress. And ‘Agentic AI’ models are capable of autonomous decision-making and action to achieve specific goals. But in asking whether these models are heralding artificial general intelligence (AGI), we must first agree on what it means to “reason” like a human, which is more in the domain of philosophy than technology. Further, real-world performance often diverges from LLM demonstrations. For instance, a September Stanford study found no evidence that LLMs can generate novel, expert-level research ideas, highlighting their current limitations.
Missing smartness in cities
On the face of it, India’s Smart Cities Mission, launched on 25 June 2015, is a grand success. Going by government figures, more than 90% of over 8,000 multi-sectoral projects worth ₹1.6 trillion, across 100 cities have been completed. Seventeen cities have achieved 100% project completion, while 34 cities have crossed 90%, and 24 have completed over 75%.
View Full Image
Key accomplishments include operational integrated command and control centres in all cities, installation of 83,000 CCTV cameras, monitoring of 9,900 km of water supply through supervisory control and data acquisition systems, and deployment of 5.2 million solar/LED (light-emitting diode) streetlights. The mission has also delivered 4,700 km of smart roads, 49,300 dwelling units, 1,300 public spaces, and 199 PPP projects worth ₹9,200 crore, says the government.
But if controlling air pollution, reducing chaotic traffic jams, building more public transport, and providing good quality water are part of the deliverables of Smart Cities, then India has done a woeful job. Globally, too, smart city projects have struggled with data privacy concerns, with a survey showing that 45% of urban residents are hesitant to share personal data for city optimization. The lack of standardized regulations further complicated collaboration between public and private sectors, leading to delays and increased costs.
“The global landscape of smart cities continues to change but at a slower pace than in recent years,” notes the International Institute for Management Development Smart City Index 2024, which was published in partnership with the Singapore University of Technology and Design. While Zurich, Oslo, Canberra, Geneva and Singapore were ranked the top 5 smart cities, in that order, not a single Indian city figured in the Top 100. New Delhi Municipal Council, which was included as a smart city in 2016, was ranked 106, followed by Mumbai (107), Bengaluru(109), and Hyderabad (111).
But there’s still hope. Data from the first smart cities mission is now being used to speed up processes and launch Smart City 2.0. Whatever data Smart City 1.0 is producing in 100 smart cities will come to the data exchange, called Urban India Data Exchange, which is fully funded by the ministry of housing and urban affairs. Further, in August, India approved 12 new industrial smart cities and other infrastructure projects. These cities are aimed at strengthening India’s domestic manufacturing ecosystem, attracting foreign investments, and boosting job creation.
Metaverse needs more than AI
Tuvalu, a Pacific nation threatened by rising sea levels due to climate change, has turned to the metaverse to preserve its culture, landscapes, and legal identity. Alongside building sea walls and reclaiming land, the government is creating a digital replica of the country. Often referred to as a ‘digital twin’, the metaverse is a 3D representation of real-world assets, offering immersive experiences through technologies such as augmented reality (AR), virtual reality (VR), and mixed reality (MR).
Despite the decades-old origins of these technologies, experts, including research and advisory firm Emergen Research, predict it could take 5-10 years for the metaverse to evolve into a cohesive, widely used platform. In the meantime, numerous private and open platforms, such as Sandbox and Decentraland, are emerging, with data portal Statista projecting metaverse revenue will reach $490 billion by 2030. The metaverse is already being explored in creative ways, such as Colombia hosting a court hearing within it.
View Full Image
Businesses worldwide, including Indian companies, are embracing the metaverse. Automakers Maruti Suzuki and MG Motors use it for 3D visualization, allowing clients to virtually customize vehicles. Companies such as Tanishq, Tata Tea, and MakeMyTrip have also entered this space.
Roland Busch, CEO of Siemens AG, emphasizes that the industrial metaverse relies on digital twins, software-defined automation, and AI. Meanwhile, tech giants Google and Apple have introduced concepts such as “ambient computing” and “spatial computing”, highlighting diverse interpretations of the metaverse.
Mark Zuckerberg remains optimistic about the metaverse’s potential despite Meta’s Reality Labs incurring a $4.4 billion operating loss in Q3 2023. Meta continues investing heavily in projects such as AI-powered glasses (Ray-Ban Meta), which integrate vision and language understanding. These glasses aim to simplify tasks like translating foreign languages or generating photo captions, potentially without needing a phone or app. Meta’s innovations also include holographic AR glasses, Orion, and its latest mixed-reality headset, Quest 3S.
However, challenges persist. VR and AR headsets remain expensive, bulky, and energy-intensive, with VR devices reportedly consuming three times the energy of standard gaming consoles. Interoperability between metaverse platforms is limited, and privacy and security concerns are significant. Critics also question the metaverse’s sustainability due to its energy demands. The rollout of 5G and advancements in edge computing is expected to address some issues by enabling faster data processing and supporting the metaverse’s computational requirements.
The metaverse represents immense potential but requires overcoming significant technical, economic, and sustainability hurdles to fulfil its vision.
Web3, NFTs take a backseat
In 2021, the term ‘Web3’ entered public consciousness, driving a surge in cryptocurrency and non-fungible token (NFT) activity. Venture capitalists invested over $30 billion in Web3 startups globally, while Indian blockchain startups raised $600 million, a 15-fold increase from 2020. Web3 is built on blockchain and related technologies such as decentralized finance (DeFi), decentralized autonomous organizations (DAOs), cryptocurrencies, and NFTs, offering users digital ownership.
Notable examples include Amitabh Bachchan’s NFT of Madhushala and digital artist Pak’s Merge, which sold for $91.8 million in 2021. Collections like the Bored Ape Yacht Club (BAYC) further cemented NFTs’ exclusivity, trading limited-edition digital art on Ethereum’s blockchain. But despite daily NFT trades worth millions, mainstream media attention has waned.
Cryptocurrencies, particularly Bitcoin, have faced significant volatility. Events like Donald Trump’s presidency and policy endorsements boosted Bitcoin, as did the approval of US-listed spot Bitcoin ETFs. Yet, Bitcoin’s value dropped 10% in late 2024, reflecting its susceptibility to regulatory decisions and global economics.
India’s approach to Web3 remains cautious. While the technology behind cryptocurrencies—blockchain—has potential, government policies have created hurdles. Gains from virtual digital assets (VDAs) like Bitcoin and Dogecoin are taxed at 30%, without the ability to offset losses or carry them forward. Transactions exceeding specified thresholds incur 1% tax deducted at source. Losses from NFTs cannot be adjusted against crypto gains, highlighting regulatory restrictions.
The Indian crypto space has also faced controversies, such as the WazirX-Binance ownership dispute, adding to investor uncertainty. Meanwhile, global incidents such as the Mt. Gox crypto exchange collapse further damaged cryptos’ reputation. Challenges like regulatory crackdowns, taxation, and price volatility cloud Web3’s prospects. However, its potential remains vast, with blockchain technology driving innovations in digital ownership and decentralized ecosystems.
Quantum leap years away
Governments, tech companies, and venture capitalists are investing billions in quantum computing, driven by the promise of revolutionizing industries and commercial and military advantages for global dominance. Quantum systems promise breakthroughs in cryptography, drug discovery, and logistics optimization with their incredible processing speeds. But they need to become stable before being deployed for everyday use. Hence, Google’s quantum computing chip, named Willow, made waves this month for its ability to reduce system errors despite adding qubits (quantum bits) and solving in under five minutes a computation that would take a supercomputer 10 septillion years (1 with 25 zeros), more than the age of the universe, to finish.
View Full Image
Two normal computer bits can exist in four possible states—00, 01, 10, or 11. But they can represent only one of these at any given time. A quantum computer allows two quantum bits (qubits) to represent the four states at the same time due to ‘superposition’ and ‘entanglement’ properties, akin to running four computers. A quantum computer’s power grows exponentially with more qubits but it also induces errors, making the system resemble your normal computer.
Stable quantum computers could accelerate AI model creation by performing parallel computations that classical systems cannot do. Google’s Willow claims to reduce errors when scaling. By doing so, it has indeed brought the world a step closer to running a more stable quantum computer. But the next challenge, as Google itself puts it, is “to demonstrate a first ‘useful, beyond-classical’ computation on today’s quantum chips that is relevant to a real-world application”.
With its National Quantum Mission, India plans to develop quantum computers with 50-100 qubits in about five years and accelerate it to 1,000 qubits and beyond in eight years. However, the focus will also be on developing quantum error correction to make quantum computers stable and functional for everyday use, and building quantum algorithms for practical applications. India currently has a six-qubit quantum computer built by the Tata Institute of Fundamental Research in partnership with the Defence Research and Development Organisation Young Scientists Laboratory for Quantum Technologies and Tata Consultancy Service.
IT services company Wipro believes enterprises will gear up for the Quantum Era with quantum computing as a service (QCaaS). The immediate priority for global enterprises will be to prepare for post-quantum cryptography, it cautions. The biopharma and chemical industries will be early adopters of quantum computing with applications in molecular modelling, predictive analytics, and drug design. Enterprises will also need to consider regulatory guidelines and ethical practices evolving around this cutting-edge technology. That said, despite the progress, stable quantum computers may see the light of day only by 2030.
