< PreviousINTERVIEW 20 EDGE | March 2025 Seven Years at Oracle: A Journey in Cloud and AI. Greg, a seasoned leader at Oracle, has been driving innovation in cloud technology for over seven years. With extensive experience across Oracle Cloud Infrastructure (OCI), he has played a key role in developing the company’s generative AI capabilities. “I’ve worked across most of OCI’s segments, including Infrastructure as a Service (IaaS) and Platform as a Service (PaaS),” Greg said. Greg Pavlik, Executive Vice-President AI and Data Management Services at Oracle Cloud Infrastructure, discusses how the tech giant’s AI strategy is helping businesses scale faster AI The Three Pillars of Oracle’s AI Strategy Oracle’s approach to AI is anchored in three key areas: infrastructure, platform services and enterprise applications. Advanced Infrastructure with Superclusters At the foundation of Oracle’s AI capabilities are superclusters designed to handle large-scale AI workloads and drive innovation. “These clusters leverage ultra-low latency networking and high-performance GPUs, scaling up to 130,000 GPUs for cutting-edge model training. Our expertise in cluster databases and network effi ciencies drives these innovations,” Greg pointed out. Powerful Platform Services for Developers Oracle’s OCI Generative AI Service provides API-driven access to curated AI models optimised for enterprise use cases. The platform supports tasks like automated job requisitions and developer productivity tools, such as Code Assist, which integrates with popular IDEs like IntelliJ. “Developers can seamlessly generate, document, or test code, boosting productivity without needing extensive setup,” Greg noted. Generative AI in Enterprise Applications Oracle has embedded AI capabilities into its Fusion Applications and NetSuite solutions, spanning 20+ industries, including banking, healthcare and HR. INSIDE ORACLE’S AI REVOLUTION 13,000 GPUs for model training and large- scale AI workloads Edge_Mar2025_20-21_Oracle Interview_13542577.indd 2024/02/2025 12:13INTERVIEW EDGE | March 2025 21 “For instance, in HR applications, managers can auto-generate performance reviews, saving hours of manual work. These AI features are pre-built, tested and ready to use, allowing businesses to start leveraging them instantly,” Greg emphasised. Addressing AI Adoption Challenges Transitioning AI projects from prototype to production can be daunting for enterprises, particularly when models aren’t well tailored for business or enterprise use. The Challenges: Scaling and Localisation “Enterprises often succeed in proofs of concept but struggle to scale due to inconsistent model performance or lack of domain-specifi c knowledge,” Greg observed. Additionally, language biases in early generative AI models posed challenges for non-English markets like the Middle East. Oracle’s strategy includes adapted models for enterprise use tested and refi ned through Oracle’s SaaS applications. Localisation: Enhanced model performance for languages like Arabic, ensuring relevance in diverse markets. Enterprise AI Trends: Effi ciency and Innovation While traditional machine learning remains vital, generative AI is redefi ning productivity across industries. Greg highlighted examples of customer support transformation, where Oracle’s generative AI and retrieval-augmented generation techniques improved case defl ection rates by 60–80%. “These gains signifi cantly reduce costs and response times, demonstrating the potential of AI to revolutionise human-intensive workfl ows,” he said. A Structured Approach to AI Agents Early adoption of open-ended AI agents often led to failures, Greg noted. To address this, Oracle developed opinionated agents tailored for specifi c business functions, ensuring consistent and reliable performance. AI in the ME: Opportunities and Challenges Greg acknowledged that while Middle Eastern enterprises are eager to adopt generative AI, some struggle to optimise its use. Avoiding the “Benchmarking Trap” “A common issue is endless benchmarking. Companies often chase the ‘best’ model instead of focusing on production workloads,” Greg observed. Oracle encourages businesses to commit to specifi c models and iterate for improved outcomes. Oracle’s AI strategy is deeply rooted in its enterprise-fi rst approach. /Greg Pavlik Executive VP AI and Data Management Services, Oracle Cloud Infrastructure Companies often chase the best model instead of focusing on production workloads “Our unmatched SaaS portfolio, covering supply chain, HR, ERP, and fi nancial applications, puts relentless pressure on generative AI models to improve,” Greg explained. This feedback loop ensures that Oracle’s AI solutions remain tailored to business needs, delivering value directly to customers. “No other cloud provider offers this depth of enterprise experience and application integration capabilities,” Greg added. Final Thoughts: A Bright Future for AI in the Region Greg expressed optimism about Oracle’s role in driving AI adoption across the Middle East. “The region is dynamic, and we’re excited to partner with enterprises here to accelerate their AI journeys. Oracle is committed to enabling meaningful business outcomes with cutting- edge AI innovations,” he concluded. Edge_Mar2025_20-21_Oracle Interview_13542577.indd 2124/02/2025 12:13SPOTLIGHT 22 EDGE | March 2025 What if the next big breakthrough in artifi cial intelligence wasn’t coming from Silicon Valley but from an emerging player in China? In the rapidly evolving landscape of AI, DeepSeek has emerged as a transformative force, challenging established norms and reshaping the industry’s dynamics. Founded by Chinese entrepreneur Liang Wenfeng, DeepSeek has introduced innovative AI models that rival leading Western counterparts in performance, effi ciency, and cost-eff ectiveness. This article delves into DeepSeek’s journey, its groundbreaking developments, and the broader implications for the global AI ecosystem. Developed in just two months with an investment of under $6 million, R1 demonstrated capabilities comparable to ChatGPT but at a fraction of the cost AI FOUNDING VISION Liang Wenfeng, born in 1985 in Guangdong, China, pursued electronics at Zhejiang University, where he developed a keen interest in computer vision and investment automation. In 2015, he founded High-Flyer Quant, a successful hedge fund leveraging quantitative strategies. By capitalising and building on this success, Liang established DeepSeek in 2023 aiming to democratise AI by making advanced models more accessible and aff ordable. BREAKTHROUGH WITH THE R1 MODEL In January 2025, DeepSeek unveiled its R1 reasoning model, a signifi cant advancement in AI technology. Developed in just two months with an investment of under $6 million R1 demonstrated capabilities comparable to leading models like OpenAI’s ChatGPT, but at a fraction of the development cost. This achievement underscored DeepSeek’s commitment to effi ciency and innovation. DeepSeek R1 is positioned as an open-source alternative to proprietary systems, making it highly attractive to researchers, startups and developers. Its open accessibility aligns with the growing demand for transparent AI systems that can be independently tested and improved. TRANSFORMING INTELLIGENCE Words by Anita Joseph 17% Drop in NVIDIA’s shares following the arrival of R1 Edge_Mar2025_22-23_Spotlight DeepSeek_13515546.indd 2224/02/2025 12:13SPOTLIGHT EDGE | March 2025 23 INNOVATIVE TRAINING APPROACHES DeepSeek’s success is attributed to its unique training methodologies. The company employs reinforcement learning techniques that minimise human intervention, allowing models to “think” more deeply and generate accurate responses. This approach not only reduces development costs but also enhances the model’s reasoning capabilities. GLOBAL IMPACT AND MARKET DISRUPTION The introduction of DeepSeek’s R1 model sent ripples through the global technology industry. Many major corporations experienced signifi cant market valuation shifts, with nearly $1 trillion in global market value aff ected. Nvidia, a leading chip manufacturer, saw its shares drop by 17% amid concerns that more effi cient AI models like R1 could reduce the demand for high-end chips. However, some experts argue that the increased accessibility of AI could drive overall demand, benefi ting companies like Nvidia in the long run. ETHICAL AND LEGAL CONSIDERATIONS DeepSeek’s rapid ascent has not been without controversy. The company’s methods, particularly its use of “distillation” to learn from existing AI models, have sparked debates about legality and ethics. Critics argue that DeepSeek may have utilised proprietary models from companies like OpenAI without explicit permission raising concerns about intellectual property rights and the ethical boundaries of AI development. DEMOCRATISING AI One of DeepSeek’s most signifi cant contributions is the complete democratisation of the AI technology. By off ering open-source models and signifi cantly reducing development costs, DeepSeek has lowered barriers to entry for smaller /Deep thinking and accurate responses DeepSeek’s rapid ascent and popularity have not been without controversy fi rms and researchers. This move challenges the dominance of tech giants and fosters a more inclusive AI ecosystem, encouraging innovation across diverse sectors. GLOBAL REACTIONS AND STRATEGIC RESPONSES DeepSeek’s emergence has prompted varied responses worldwide. In the United States, the development has been likened to a “Sputnik moment,” prompting a re-evaluation of the nation’s AI strategies to maintain technological leadership. Companies like Microsoft and Meta Platforms have acknowledged DeepSeek’s innovations and are exploring ways to integrate similar advancements into their systems. Meta’s CEO, Mark Zuckerberg, expressed confi dence in adapting DeepSeek’s innovations for Meta’s upcoming Llama 4 model, aiming to position it as a top-performing AI. CHALLENGES AND FUTURE OUTLOOK Despite its achievements, DeepSeek faces signifi cant challenges, including regulatory scrutiny over data practices and concerns about potential geopolitical implications due to its Chinese origins. However, the company’s focus on effi ciency, transparency and open-source development positions it well for future growth. As AI continues to evolve, DeepSeek’s approach may serve as a blueprint for sustainable and inclusive innovation. DeepSeek’s rise signifi es a pivotal shift in the AI landscape, highlighting the potential for great innovation beyond the traditional technology hubs. By prioritising effi ciency, accessibility and open- source collaboration, DeepSeek is defi nitely paving the way for a more democratised, accessible and highly dynamic AI future. Edge_Mar2025_22-23_Spotlight DeepSeek_13515546.indd 2324/02/2025 12:13INTERVIEW 24 EDGE | March 2025 Frans Hiemstra, Regional General Manager MEA, discusses how Uber is transforming the autonomous vehicle landscape in the UAE MOBILITY Can you share the key milestones Uber achieved in bringing autonomous vehicles to Abu Dhabi, and how this launch aligns with the region’s vision for smart transportation? At Uber, we’re building the future of transportation, which will be increasingly electric, shared, multimodal and autonomous. As the largest combined mobility and delivery platform in the world connecting users with transportation and delivery services through the Uber app, we are uniquely positioned to help bring this technology to our consumers. Rather than building autonomous vehicles ourselves, our goal is to work with the world’s leading autonomous technology developers to help deploy this technology at scale, learn as the industry advances and become the partner of choice for the industry. Over the past few years, we’ve hit signifi cant milestones in bringing autonomous vehicles onto the Uber platform. We’re currently working with 12+ autonomous technology companies – across mobility, delivery and freight. In December 2024, we launched our partnership with WeRide to bring autonomous vehicles on the Uber platform for the fi rst time outside of the US in Abu Dhabi. The launch aligns with the UAE’s Vision 2030, which prioritises sustainability, innovation, and advanced technology adoption. Through our partnership with WeRide, we’re working with Abu Dhabi to help realise its goal of becoming a global leader in intelligent transportation systems by integrating cutting-edge autonomous vehicles into its extensive mobility network. BUILDING THE FUTURE OF TRANSPORTATION Edge_Mar2025_24-25_Interview Uber_13515952.indd 2424/02/2025 12:14INTERVIEW EDGE | March 2025 25 What are the specifi c benefi ts Uber’s AVs off er to commuters in Abu Dhabi compared to traditional ride-hailing services? We’re excited about our partnership with WeRide in Abu Dhabi to bring autonomous vehicles closer to people. While it won’t happen suddenly overnight, we expect autonomous vehicles to be an increasingly important part of the transportation ecosystem and Uber’s business over time. The future of transportation is arriving now – Uber is making electric, shared, multimodal and autonomous transportation a reality – with 14 AV partners to date, and tens of thousands of autonomous trips every month. And just like with EVs, more people will try autonomous vehicles on the Uber app for the fi rst time than anywhere else. Uber customers in Abu Dhabi who are excited about riding in an autonomous vehicle can increase their chances of being matched with an AV by opting in via the Ride Preferences section of their Uber app under Settings. The future of transportation is arriving now and Uber is making it a reality How does Uber address safety and public trust concerns regarding the adoption of autonomous vehicles in a market like Abu Dhabi? Safety is a major consideration for all transportation types, and we take our role, promoting safety in mobility seriously. Our partners develop and build the technology necessary to power autonomous vehicles with safety in mind. In deploying AVs on the Uber platform, we review our partners’ approaches to safety and only partner with AV services that meet our own robust expectations for security, privacy and safety. To that end we have developed our safety guidelines for autonomous mobility partners on the Uber Platform, which incorporates the industry state-of-the-art best practices and standards on safety. In the launch period, each WeRide AV has a vehicle operator present to help provide a safe secure and reliable experience for riders and pedestrians, laying the groundwork for a fully driverless commercial service planned for later in 2025. We also work closely with our partner WeRide and abide by local regulatory guidelines to meet Abu Dhabi’s safety and operational standards. What role do local partnerships and government regulations play in the successful deployment of Uber’s AVs in the region? Strong local partnerships and regulatory support are essential to bring any new technology to a market. Our partnership with WeRide in Abu Dhabi endorsed by the Integrated Transport Center (Abu Dhabi Mobility) allows us to deploy autonomous technology that integrates seamlessly with Abu Dhabi’s dynamic and fast-growing transportation landscape. Tawasul Transport, a leading national transport company, is the fleet operator. Public-private partnerships play a critical role in establishing a sustainable mobility ecosystem with universal access to shared, electric autonomous and multimodal transport at the push of a button. Looking ahead, how does Uber envision scaling its AV operations in Abu Dhabi and across the Middle East, and what challenges do you foresee? Building on the success of our Abu Dhabi launch, we’re excited to expand AV services across the Middle East while adapting to the various local infrastructures and regulatory environments. Over time, we expect autonomous vehicles to be an increasingly important part of the transportation ecosystem and therefore Uber’s business. Not all routes will be able to be served by an autonomous vehicle, particularly over the next few years. Edge_Mar2025_24-25_Interview Uber_13515952.indd 2524/02/2025 12:14INNOVATION 26 EDGE | March 2025 Dubai is at the forefront of the global 3D printing revolution, positioning itself as a hub for innovation and technological advancement. The emirate has made signifi cant strides in adopting 3D printing across various sectors, from construction and healthcare to manufacturing and aerospace, aligning with its ambitious Dubai 3D Printing Strategy launched in 2016. This initiative aims to tap into technology for the service of humanity and promote the status of the UAE and Dubai as a leading hub of 3D printing technology by the year 2030. The three key sectors of this strategy are construction, medical products and consumer products. Dubai has already set world records with its 3D-printed structures. Always at the forefront of innovation, in 2019, it unveiled the world’s largest 3D-printed building, a two-story government facility spanning 640 square meters. The city also built the fi rst 3D-printed offi ce, proving the feasibility of large-scale, eco-friendly construction using robotic arms and advanced concrete mixtures. These innovations reduce construction time by up to 70% and cut labour costs by 50%, making sustainable, aff ordable housing more accessible and available. Dubai’s healthcare sector is leveraging 3D printing for customised prosthetics implants, and even organ models for complex surgeries. The Dubai Health Authority (DHA) is actively investing in medical 3D printing, aiming to produce personalised medical solutions that enhance patient care and reduce costs. Thanks to its eff orts, DHA now provides state-of-the-art healthcare and services to all who seek it. Dubai’s commitment to 3D printing is not just a technological leap but a strategic move toward a more sustainable, effi cient and innovative future, reinforcing its status as a global leader in emerging, innovative technologies. Against this backdrop, DA BUREAU the pioneering London-based architect’s studio, has debuted with the world’s largest 3D-printed restaurant interior design at MYATA Platinum Lounge, a hookah lounge with a restaurant and bar in C2 Licensed District, nestled in the heart of Dubai’s City Walk. The projec t ’s centrepiece comprises 27 unique “canyons” assembled from over 20,000 3D-printed components, produced using thirty- fi ve 3D printers. Each structure is crafted from recycled plastic and fi nished with textured paint, showcasing the potential of sustainable materials in luxury design. By using recycled plastic, the project contributes to the Arab building This Dubai lounge is breaking records with a mind-blowing 3D-printed interior CULTURE 3DPRINTING REVOLUTION Words by Anita Joseph Edge_Mar2025_26-27_Culture_13515722.indd 2624/02/2025 12:15INNOVATION EDGE | March 2025 27 sector’s goal of reducing carbon emissions by 8.8 million tons annually by 2030 through improved and enhanced technologies. “To materialise the idea, the studio built the interior on the image of a sand canyon. A canyon is a constantly changing geological formation shaped by water and wind. We recreated its fl owing amorphous forms running into each other in a modern interpretation. By employing 3D printing, DA BUREAU created whimsical fl uted structures that shaped an introverted, inward-looking space. Outside, these curving, enveloping rocky structures form a terrace isolated from the city. At the same time, indoors they shape the layout, dividing the interior into separate chamber zones,” said Boris Lvovsky, the CEO and co-founder of DA BUREAU. The canyon surrounding the entrance uses funnel principle — it narrows as if to succumb the visitor from the street into the lounge space. It is important to emphasise that visitors cannot see the place’s interior from the facade or the entrance hall. Entering the main lounge hall is reminiscent of a cave: through narrow, dark passages to a large open space. The entire eff ect is breathtaking. The project highlights Da Bureau’s commitment to innovative design solutions that are environmentally responsible and sustainable. “The design transforms recycled plastic into dramatic canyon-like formations, creating an immersive environment that transports visitors from Dubai’s bustling metropolis to a serene, cave-like sanctuary,” Lvovsky added. Edge_Mar2025_26-27_Culture_13515722.indd 2724/02/2025 12:15ANALYSIS 28 EDGE | March 2025 Google’s parent company Alphabet, has announced a staggering $75 billion capital expenditure plan for the year 2025 AI Alphabet is making its boldest bet yet on the future of AI. The Google parent company has announced a staggering $75 billion capital expenditure (CapEx) plan for 2025 setting the stage for a massive expansion in artifi cial intelligence, cloud infrastructure and next-gen hardware. This unprecedented investment signals that Alphabet is gearing up for an all-out AI arms THE BILLION- DOLLAR BOOST race, as it looks to outpace rivals like Microsoft, OpenAI, and Amazon. With the demand for AI- powered services and cloud computing surging, this spending spree could redefi ne the company’s competitive standing in the tech industry. AI: THE PRIMARY DRIVER OF CAPEX GROWTH Alphabet’s increasing CapEx spending is largely driven by its AI ambitions. Google has been aggressively pushing the development of AI-powered services, including Gemini (formerly Bard), enterprise AI solutions, and AI-integrated search enhancements. The company is in a fi erce race against Microsoft- backed OpenAI, which has gained signifi cant traction with ChatGPT. The increased capital spending will likely fund AI model training, data center expansions and custom AI chips (TPUs - Tensor Processing Units). Additionally, DeepMind, Alphabet’s AI research arm, is at the forefront of AGI (Artifi cial General Intelligence) research, and sustained investment is crucial to maintain its competitive edge against AI leaders like Meta, Anthropic and OpenAI. Edge_Mar2025_28-29_Analysis_13518643.indd 2824/02/2025 12:15ANALYSIS EDGE | March 2025 29 CLOUD COMPUTING & DATA CENTERS: STRENGTHENING INFRASTRUCTURE Another key area Alphabet is investing in is Google Cloud, which is competing with Amazon Web Services (AWS) and Microsoft Azure. The demand for cloud- based AI services, cybersecurity, and enterprise solutions is increasing, and Alphabet must scale up its data center capabilities to support this growth. The CapEx allocation will likely focus on building and upgrading data centers to accommodate AI workloads, expanding Google Cloud’s infrastructure for enterprise AI adoption, developing more energy- effi cient AI chips, reducing operational costs. Given the AI boom, cloud providers with the best infrastructure will have a signifi cant advantage, making this investment a strategic necessity. HARDWARE EXPANSION: PIXEL, FITBIT, WEARABLES AND AR/VR TECHNOLOGIES Alphabet is also expected to direct some of its CapEx toward hardware innovation, particularly in Pixel devices, Fitbit wearables, and augmented reality (AR)/virtual reality (VR) technologies. The rise of spatial computing (as seen with Apple Vision Pro) could push Google to enhance its AR/VR off erings, possibly reviving its AR hardware ambitions. /Refining Google’s search queries /Alphabet’s boldest bet yet The investment also refl ects Alphabet’s ambition to future-proof its business HOW THIS INVESTMENT WILL TRANSFORM THE FUTURE OF AI DEVELOPMENT Alphabet’s $75 billion CapEx push will have profound implications for the future of AI development, accessibility and commercialisation. Faster AI Advancements With increased investment in AI chips, data centers, and research, Alphabet can accelerate the pace of AI model development. This could lead to faster, more capable AI systems with improved reasoning, multimodal understanding and effi ciency. More Scalable AI Applications The expansion of Google Cloud means businesses and developers will have greater access to AI-driven tools and platforms. This could democratise AI capabilities, making advanced machine learning more accessible across industries like healthcare, fi nance and education. Potential Leap Toward AGI Alphabet’s DeepMind is one of the leading entities working on Artifi cial General Intelligence (AGI)—a form of AI that can perform a wide range of cognitive tasks at human-level intelligence. This funding could bring AGI research closer to reality shifting the landscape of automation and decision-making. Dominance in AI-Powered Search and Personalised Advertising As Google integrates AI into Search, this investment will help refi ne AI-powered search queries, personalised advertising, and automated content generation. Google’s dominance in search and ad revenue could be further solidifi ed with better AI models. Challenges to AI Ethics and Regulation With AI becoming more powerful, Alphabet will need to navigate regulatory scrutiny and concerns around bias, misinformation and ethical AI use. Governments worldwide are considering regulations, and Alphabet’s aggressive AI expansion may draw further regulatory attention. Strategic Implications for the Company Alphabet’s $75 billion CapEx plan represents a clear intent to dominate AI, cloud and hardware, reinforcing its long-term vision. The investment also refl ects its eff orts to future-proof its business amid rising competition, ensuring that it remains a leading force in AI innovation and digital infrastructure. If successful, this massive expenditure could reshape the Artifi cial Intelligence landscape, redefi ne human-computer interaction and power the next wave of technological breakthroughs and innovations. Edge_Mar2025_28-29_Analysis_13518643.indd 2924/02/2025 12:15Next >