The Screen Trap: Why Digitizing a Broken Process Only Breaks It Faster//////////The pitfalls of paving the cow path: why unplanned and half-hearted digitization doesn’t work & what are the solutions

 @Prof Archie D'Souza

A few decades ago, while working as a branch manager for a freight forwarding company, we rolled out a brand-new accounting software system. It was state-of-the-art—on paper, at least. But it raised an immediate, glaring problem: where were the computers to feed the data and print the invoices? And who, exactly, was going to train a workforce entirely unaccustomed to digital entry? I was told about rollout. The ambition was there, but the reality quickly hit: we had advanced software with nowhere and no one to run it; no computers in the branch to process the data, and no trained staff to bridge the gap. It was an early, hard lesson in a timeless truth.

“Accounts,” I was told, “have to be computerized.”

“Of course!” I said, “But for that, we need computers and trained staff.”

So, for the next few months, manual accounting continued at the branch level. A copy of the ledger entries was sent to the head office, and entries were made there. Each branch was provided with a printout of the account statements.  

Very often, we think we’re moving forward but what we’re experiencing is an illusion of progress. This is a common scenario where a company decides to "go digital." They take a messy, paper-heavy, disjointed process and simply turn it into an app, a software portal, or a digital form. Of course, I’m referring to an era that was long before portals and apps.

The Reality Check: This is exactly what I told my boss, "You cannot streamline a broken, disconnected process merely by putting it on a screen; the entire operational framework requires a paradigm shift."

Then it dawned on the board that introducing computerization in accounting was a process that needed a great deal of work. Then in the late 80s came Tally and it was introduced across branches but only after the right systems were in place and the accounting team trained to use Tally. Further, with the version of Tally we got, we could bill in multiple currencies. Any accounting statement that we needed, as well as accounts receivable statements, were system-generated.

Fast forward to the internet era.

We’ve all been there. A company recognizes that its operations are sluggish, siloed, and drowning in paperwork. Leadership gathers in a conference room, diagnoses the bottleneck, and announces the grand solution: "We are going digital."

Months and thousands of dollars later, a sleek new software portal goes live. Employees are given login credentials, training sessions are held, and management breathes a sigh of relief. The era of efficiency has arrived.

Except... it hasn't.

Instead of saving time, employees find themselves clicking through five different windows, jumping across clunky approval loops, and dealing with digital bottlenecks that feel suspiciously like the old paper ones. The tools have changed, but the frustration remains identical.

Why? Because you cannot streamline a broken, disconnected process merely by putting it on a screen; the entire operational framework requires a paradigm shift.

The Pitfall: Paving the Cow Paths

What does this mean? Let me first explain the concept. How often do resort to automating old, inefficient habits just because "that's how we've always done it." This is exactly what the concept of “paving the cow paths” refers to. We have the tendency to simply customise existing systems, preserving them rather than reengineering them. This process is often seen in IT and business processes. It leads to a system that lacks the necessary improvements or enhancements even though it may be more efficient in execution. The metaphor of a cow path illustrates the inefficiency and ineffectiveness of a route that has been worn by habit, which can become a permanent fixture in infrastructure. Automating such processes without a thorough understanding of their necessity or potential for improvement can result in a system that is more efficient at doing something it may no longer need to do. This is a critical lesson for organisations to consider while deciding how to leverage technology to improve processes.

We need to look at an old metaphor from urban planning and IT. This is exactly what we continue to do; pave the cow paths.

The Result: Instead of saving time, technology often just amplifies frustration. Now, to repeat what I’ve already said, people are clicking through five different screens to achieve the same bottlenecked result they used to achieve with a sticky note.

To understand why digital transformation fails so often, it helps to look at this old metaphor.

The Illusion of Progress

The danger of the "screen trap", as we’ve seen, is that screens create an illusion of progress.

When you digitize a process, it looks innovative. Dashboards light up, reports can be generated with a click, and executives can track metrics in real-time. But if the underlying operational framework is fractured, putting an interface over it is merely putting a fresh coat of paint on a crumbling foundation.

• The Communication Gap: If Sales and Operations don't actually talk to each other offline, giving them a shared project management tool won't force them to communicate; it will just give them a digital space to misunderstand each other faster.
• The Complexity Creep: Software makes it very easy to add steps. Because "it’s just a button," teams tend to add extra review layers, optional fields, and redundancy that would never survive in a physical workflow.

Technology is an amplifier. If you feed a broken process into a digital platform, the platform doesn't fix the process—it amplifies the chaos.

 Designing for the Paradigm Shift

True modernization doesn't start with code, software selection, or UI design. It starts with a paradigm shift in how you view the work itself.

Before introducing a single new tool, teams need to embrace a first-principles approach to their operations. Technology should only ever enter the picture after the process has been radically simplified and redesigned. This should be the approach:

• Process First, Tech Second: Technology should only amplify a well-designed strategy, not fix a broken one.

• Obliterate Before You Automate: Look at a workflow and ask: Why do we even do this step? If the original reason no longer exists, cut the step entirely rather than digitizing it.
• Focus on Outcomes, Not History: Stop asking, "How do we move this task from step A to step B the way we used to?" Instead, ask, "If we were building this company from scratch today, how would we achieve this outcome?"
• Break Down the Silos: Re-engineer the framework so that cross-functional teams intersect naturally, rather than tossing work over digital walls from one app to another.

• First-Principles Thinking: Break down what it takes to look at an operational framework with fresh eyes:
• Where is the friction?
• What are we actually trying to achieve (the outcome), versus the steps we take to get there?
• How do different departments actually communicate and intersect?

Look Beyond the Glass

Screens are windows. If the room behind the window is a chaotic, disorganized mess, cleaning the glass isn't going to change the view.

Call to Action: Ask yourself ~ Where in your organization are you trying to solve a structural problem with a software solution?

If your organization is looking to modernize, stop treating technology as the magic eraser for bad workflows. Step back, look at the underlying framework, and be willing to change the rules of the game before you put them on display.

Why your Enterprise Doesn't Need a 100-Tonne AI Crane: Beyond the Megawatt - The Case for Small Language Models in Operations/ Right-Sizing AI: Moving from Hype to Industrial Execution

 ©Prof Archie D’Souza

v   Faculty in Logistics, Supply Chain & Project Management, adjunct professor at Dayananda Sagar University, visiting professor at Rajeev Gandhi National Aviation University and other institutions pan-India.

v   Subject Matter Expert and Faculty at the Logistics Sector Skill Council of the National Skill Development Corporation.

v   Author of “Simplifying Blockchain Complexities” and forthcoming books on AI, IoT and ML, along with blockchain, applications in Projects and Supply Chains, and another on Blockchain Technology’s Impact on International Trade

AI is often overhyped. It’s useful, not doubt but when I look at it, this core analogy comes to my mind. You don't send a 100-tonne crane to drive a nail. This is the exact kind of clear, visual metaphor that I see that cuts through the artificial intelligence hype. Why and how? For the past several years, the enterprise narrative surrounding artificial intelligence has been dominated by a simple philosophy: bigger is better. We have watched an arms race unfold toward multi-hundred-billion and trillion-parameter models trained on vast swaths of the internet, capable of drafting poetry, passing bar exams, and engaging in deep philosophical debate. So, what does my metaphor do?

1. It reverses the "Bigger is Better" Narrative: Most AI content focuses on massive frontier models and massive parameter counts. I flip this on its head. I present right-sizing as the mature, intelligent engineering move.
2. As an executive, a supply chain manager, or a project leader, you instinctively understand operational efficiency as well as asset allocation. So, framing model selection like equipment selection makes abstract tech instantly practical.
3. The ability to share: I thought when I coined this maxim that the title and theme are inherently opinionated and thought-provoking, which drives strong engagement on platforms like LinkedIn or Medium.

• The Problem (The "Megawatt" Trap): Why running routine tasks through multi-billion/trillion parameter cloud models bleeds money, introduces latency, and creates unnecessary risks.
• The Solution (Task-Specific Precision): Introducing SLMs under 10B parameters as specialized, nimble tools designed for fast, local, low-cost execution.
• The Business Impact: Sub-second latency on the edge, complete data privacy, and a fraction of the power consumption.

AI maturity isn't about how large your model is—it's about how effectively it solves a specific problem where the work actually happens.

What’s the scenario with regard to supply chains & projects?

As supply chain executives, project directors, and site managers attempt to integrate these massive general-purpose models into daily operations, they hit a hard wall of reality. High latency, skyrocketing API token costs, strict data privacy constraints, and massive energy overheads quickly turn high-flying tech demos into operational bottlenecks.

In response, a quiet but powerful shift is taking place across logistics, manufacturing, and field management. This transformation is not a downgrade in AI capability, but a maturation of operational engineering.

The Right Tool for the Job

Just as civil engineers do not deploy a 100-ton crane to drive a simple nail, operational leaders are realizing that routine enterprise tasks do not require the entire weight of a multi-megawatt data centre.

In day-to-day operations, 80% to 90% of AI tasks are highly specific:

• Parsing a structured bill of lading or customs manifest.
• Classifying field inspection logs and safety report priorities.
• Processing inventory updates on a factory floor.
• Generating quick risk summaries from project schedule metadata.

Routing these routine transactions through a massive cloud-hosted LLM is computationally wasteful. The shift toward task-specific Small Language Models (SLMs) under 10 billion parameters represents a transition from high-level AI experimentation to precise, cost-effective industrial execution.

Why Small AI Wins at the Operational Edge

When intelligence moves from distant cloud servers to local edge devices—such as handheld scanners, site laptops, or on-premises servers—it unlocks four critical operational advantages:

1.      Radical Cost Efficiency

Querying cloud LLM APIs millions of times a day for routine processing creates unpredictable, runaway operational expenditures. Fine-tuned SLMs run on local hardware at near-zero incremental cost per transaction, reducing overall AI compute overhead by up to 90%.

2.      Sub-Second Real-Time Speed

Cloud connections introduce network round-trips and queue delays of 1.5 to 4 seconds. In high-throughput settings like warehouse scanning or automated assembly line sorting, that delay breaks the workflow. Edge-deployed SLMs process data directly on the device in 50 to 150 milliseconds, making real-time automation possible.

3.      Complete Data Sovereignty

Transmitting confidential project schedules, proprietary engineering designs, or sensitive supplier contracts over external networks creates severe IP and privacy risks. On-device SLMs keep 100% of the operational data within your secure local network.

4.      Offline Resilience & Sustainability

Remote construction job sites, underground transport tunnels, and metal-shielded warehouses frequently suffer from spotty internet access. Local SLMs run entirely offline, ensuring field teams stay productive anywhere. Furthermore, running a 15W to 45W edge model aligns far better with corporate ESG and energy-efficiency goals than consuming cloud data center megawatts.

The Future is Right-Sized

The future of enterprise AI isn't about replacing large cloud models entirely—it's about building a smarter, tiered architecture. Centralized mega-models will continue to serve as high-level strategy hubs for complex, multi-variable reasoning. But on the front lines, nimble, task-specific Small Language Models will do the heavy lifting where the work actually happens.

True AI maturity isn't measured by how many billions of parameters your model has. It’s measured by how effectively, quickly, and cost-effectively it solves a real-world problem.

Supercomputing at the Edge with Physical AGI: Why Intelligence Needs a Body

Prof Archie D’Souza

When most people picture Artificial General Intelligence (AGI), they envision an all-knowing digital mind residing in the cloud. They imagine an invisible oracle inside climate-controlled server farms, accessible only through chat interfaces and clean text boxes. We have been conditioned to believe that the pinnacle of AI is purely virtual—a software engine that processes language and generates pixels.

But true intelligence was never meant to live behind a screen.

In the physical world, intelligence is not an abstract exercise in processing symbols; it is a survival mechanism. It must contend with gravity, friction, momentum, and dynamic environments. It must navigate unpredictable crowds, dodge obstacles, and make split-second motor adjustments where millimetre errors carry severe costs. To interact meaningfully with our world, AI must become physical.

Yet, physical agency exposes the fatal flaw of centralized computing.

A bipedal humanoid balancing on an icy sidewalk in Shimla cannot wait for a data packet to travel to a server in Bengaluru, several states away. A surgical robotic arm operating near a major artery cannot afford a 200-millisecond network hiccup. An autonomous drone navigating an off-grid collapse zone cannot query a remote API. When action occurs in physical space, the latency loop of the cloud is not just an inconvenience—it is a critical point of failure.

This realization is driving the next big leap in technology: Edge Supercomputing powered by Physical AGI.

The Death of the Cloud Latency Loop

For the past decade, the dominant strategy for AI was simple: build massive models in hyper-scale data centers and stream access via cloud APIs. While this works well for generating code or drafting emails, physical systems operate under entirely different constraints:

1. The Latency Ceiling: Physical control loops—such as dynamic balancing, object manipulation, and collision avoidance—require microsecond-to-millisecond execution cycles. Round-trip transit times to distant servers destroy real-time control, causing physical instability or catastrophic delays.
2. Bandwidth Saturation: A physical robot equipped with multi-camera spatial vision, high-frequency IMUs, LiDAR, and haptic sensors generates gigabytes of raw data per second. Uploading this volume of data continuously over cellular networks is practically and economically impossible.
3. Off-Grid Reliability: Autonomous systems in mines, disaster zones, or remote high-altitude regions frequently lose network access. An intelligent agent that freezes when its signal drops is not autonomous—it is merely a remote-controlled peripheral.

For an intelligent agent to react in real time, its cognitive engine cannot be separated from its physical chassis. The mind and the body must occupy the exact same coordinate in space.

Silicon at the Frontier: Supercomputing Gets Small

To bring physical AGI to life, hardware engineers are fundamentally reinventing computing architecture. The goal is no longer just power efficiency; it is packing data-centre-grade performance into compact, battery-powered silicon.

Modern System-on-Chips (SoCs) and dedicated Neural Processing Units (NPUs) now deliver hundreds of teraflops—and even petaflops—of processing capacity directly on edge devices. These localized chips run complex Vision-Language-Action (VLA) models natively. Instead of relying on pre-programmed scripts, the robot ingests real-time sensor streams, constructs an internal "world model," evaluates spatial physics, and executes precise physical manoeuvres locally within milliseconds.

From Factory Floors to Operating Rooms

The convergence of Physical AGI and Edge Supercomputing is already reshaping major industries:

• Humanoid Robotics: Factory floors and logistics hubs are transitioning from rigid, caged robotic arms to dynamic humanoids that adapt to chaotic environments and work alongside humans safely.
• Extreme Exploration & Defence: Search-and-rescue drones can navigate subterranean caves, dense forests, and damaged structures without GPS or external communication links.
• Precision Healthcare: Haptic surgical equipment equipped with edge AI can apply real-time safety barriers, preventing accidental tremors or unintended cuts during delicate operations.

The Horizon Ahead

Significant challenges remain before physical AGI becomes ubiquitous. Engineering teams are still tackling thermal constraints in compact frames, bridging the gap between simulated training environments and real-world physics, and building deterministic safety guardrails into probabilistic neural networks.

However, the trajectory is clear. The era of disembodied AI trapped inside web browsers is giving way to localized, embodied intelligence. By moving supercomputing capabilities out of distant data centers and placing them directly into machines, we are building technology that doesn't just process our world from afar—it steps into it alongside us.

Beyond the Box: How Decentralized Trade Finance Unleashes the Next Global Economy

©Prof Archie D’Souza

v   Faculty in Logistics, Supply Chain & Project Management, adjunct professor at Dayananda Sagar University, visiting professor at Rajeev Gandhi National Aviation University and other institutions pan-India.

v   Subject Matter Expert and Faculty at the Logistics Sector Skill Council of the National Skill Development Corporation.

v   Author of “Simplifying Blockchain Complexities” and forthcoming books on AI, IoT and ML, along with blockchain, applications in Projects and Supply Chains, and another on Blockchain Technology’s Impact on International Trade.

In 1956, when Malcolm McLean loaded 58 metal boxes onto the Ideal-X, he wasn't just speeding up port operations; he was setting off an economic chain reaction.

Containerization drastically lowered shipping costs, but its true legacy was the globalized economy it enabled: just-in-time manufacturing, global supply chains, and the rapid economic rise of export-driven nations across Asia. Without the standardized box, modern global trade as we know it simply could not exist.

In Part 1, [see: https://aviationtransportationbuffs.blogspot.com/2026/08/is-dapps-malcolm-mclean-of-digital-era.html] we explored how Decentralized Applications (dApps) and Real-World Asset (RWA) tokenization serve as the digital equivalent of McLean’s container—standardizing chaotic paperwork into programmable, dynamic digital assets.

But standardizing the asset is only the first step. The real question is: What new economic models become possible when trade data and capital move at the speed of light?

Breaking the SME Financing Wall

For decades, international trade finance has been the domain of multinational conglomerates and tier-1 banks. Small and medium enterprises (SMEs)—especially those in developing markets—face a staggering $2.5 trillion trade finance gap.

Under the traditional banking model, securing a Letter of Credit or obtaining invoice factoring requires months of administrative vetting, credit checks, and physical collateral. A small exporter in Southeast Asia or South America might have a solid order from a buyer in Europe, but without upfront liquidity, they cannot buy raw materials or fulfill the contract.

When trade documents are tokenized on a decentralized network, that paradigm shifts:

• Fractional Liquidity: Instead of waiting 90 days for a buyer to settle an invoice, an exporter can tokenize that invoice as an RWA and list it on a global DeFi liquidity pool. Investors anywhere in the world can purchase fractions of that invoice, providing the exporter with immediate working capital.
• Trustless Credit Scoring: Risk is no longer assessed solely on the exporter's balance sheet or country rating, but on the verifiable cryptographic lineage of the trade transaction itself—verified by IoT telemetry, digital port clearances, and smart contract history.

By democratizing access to capital, dApps allow small players to plug into global trade with the same ease as multinational giants.

The Rise of Autonomous Supply Chains

The next horizon goes beyond streamlining human finance—it introduces machine-to-machine commerce.

In a fully integrated, decentralized logistics network, physical infrastructure gains financial agency through smart contracts:

• Self-Paying Vessels: Imagine a container ship that autonomously executes payment to port authorities, pays for bunkering/refuelling, and settles port handling fees via smart contract the moment GPS or port sensors confirm its arrival.
• Automated Insurance Claims: If a cold-chain container suffers a power failure in transit, embedded temperature sensors feed data to an oracle. The smart contract automatically verifies the breach, triggers an insurance payout, and reorders replacement cargo before the vessel even docks.

The need for manual reconciliation, claims processing, and multi-day audit trails disappears.

Building the Network Effect

Malcolm McLean's container didn't transform the world overnight—it required new port cranes, redesigned ships, and intermodal railway links. Similarly, unlocking the economic aftershocks of dApps requires legal frameworks (like the UNCITRAL MLETR model) and interoperability standards across blockchains.

Yet the economic incentive is irresistible. Just as containerization sparked the greatest expansion of physical commerce in human history, decentralized applications are laying the financial rails for an autonomous, frictionless, and inclusive global economy.

Yet, infrastructure alone was not the hardest barrier to break—it was human and institutional inertia. Powerful longshoremen unions initially fought containerization fiercely, recognizing that standardized boxes would drastically shrink dockside labour. At the same time, traditional shipping executives dismissed the system as a costly, unviable fad. Early container ships even faced restrictive regulatory battles and hostility from traditional port authorities reluctant to dismantle centuries of break-bulk operations.

Similarly, the widest chasm for dApp adoption isn't technical throughput, but entrenched operational resistance. Legacy freight forwarders, paper-bound customs authorities, and intermediary banks often view decentralized, automated smart contracts as a direct threat to their business models and administrative gatekeeping. True transformation happens only when economic necessity forces legacy systems to adapt to the new standard.

This video discusses Malcolm McLean's pioneering role in containerization and how he transformed the shipping industry: History of Shipping Industry Malcolm McLean

The box unified physical trade. The protocol is unifying global wealth.

Crypto Was the Wild West, Blockchain Is the Rule of Law.

 ©Prof ARCHIE DSOUZA, a blockchain enthusiast

The history of crypto has long read like a tale from the American frontier. Gold rushes, lawless territories, shadowy figures, and fortune seekers taking wild risks defined its early years. In that era, "Code is Law" was a rallying cry, but in practice, it often meant lawlessness: high-profile exchange collapses, untraceable hacks, and pump-and-dump schemes that left investors holding empty bags.

It was an unmapped wilderness. But as the dust settles on the speculative fever, a far more profound truth is emerging: crypto was merely the chaotic Gold Rush; blockchain is the railroad, the telegraph, and the rule of law.

From Outlaw Frontier to Institutional Infrastructure

When you strip away the meme coins, the underlying architecture of blockchain technology represents something revolutionary: a system of immutable, transparent, and self-enforcing rules.

In traditional finance and trade, the "rule of law" relies on human institutions—banks, courts, governments, and auditors. While essential, these systems are vulnerable to corruption, human error, and bureaucratic delays. Blockchain replaces blind trust in fallible intermediaries with trust in mathematical consensus and cryptographic proof.

Where crypto saw anarchy, blockchain brings order. It creates a digital ledger where once a transaction or record is written, it cannot be secretly altered, erased, or manipulated.

Why the Shift Matters Today

Global trade and supply chains are currently facing unprecedented friction—tariff disputes, geopolitical conflicts, and opaque multi-tier supplier networks. In this environment, the "Wild West" mentality of crypto offers zero value. However, the "rule of law" provided by blockchain offers solutions:

• Verifiable Transparency: Companies can track goods, parts, and payments with 100% auditability across international borders.
• Self-Executing Agreements: Smart contracts enforce terms automatically. If a shipment reaches a port, payment releases instantly—no hidden clauses, no delayed wire transfers, and no breach of contract.
• Compliance and Sanctions Management: Blockchain allows real-time verification of origin, ensuring businesses comply with complex legal frameworks without relying on paper trails that can be forged.

The New Frontier is Governed by Trustless Order

The transition from a speculative playground to institutional infrastructure marks the maturation of the technology. Major global financial institutions, shipping conglomerates, and governments are no longer looking at blockchain as a way to circumvent regulations, but as a tool to establish digital rule of law.

The Wild West era was necessary to test the technology’s boundaries, but it was never sustainable. The future belongs to blockchain—a framework where code does not replace the law, but enforces integrity, accountability, and trust in a fragmented world.

AI is Moving from the Cloud to your Pocket: The rise of Edge Intelligence and shift toward it, generally & in supply chains with special reference to handheld devices

 ©Prof Archie D’Souza

·        Faculty in Logistics, Supply Chain & Project Management, adjunct professor at Dayananda Sagar University, visiting professor at Rajeev Gandhi National Aviation University, PESIT, and other institutions pan-India.

·        Subject Matter Expert and Faculty at the Logistics Sector Skill Council of the National Skill Development Corporation.

·        Author of “Simplifying Blockchain Complexities” and forthcoming books on AI, IoT and ML, along with blockchain, applications in Projects and Supply Chains; and another on Blockchain Technology’s Impact project on International Trade.

For the past decade, the prevailing narrative around artificial intelligence has been built on a single, massive assumption, as we grow bigger, we will need to migrate to the cloud, if we haven’t already. In other words, AI requires the cloud.

Whenever you asked your smartphone to summarise an email, generate an image, or translate text, your request undertook a silent journey. It travelled over 5G or Wi-Fi, entered a sprawling server farm packed with power-hungry GPUs, processed your query, and beamed the answer back to your screen.

The cloud is the undisputed brain, while your phone is merely a window. Will this continue?

Now, a quiet revolution is redefining how we interact with technology. The intelligence that once required server racks cooling in data centres is shrinking, adapting, and settling directly onto your local hardware. AI is moving from the cloud into your pocket.

This shift toward Edge AI—running artificial intelligence models locally on your smartphone, laptop, or IoT device without a constant connection to the cloud—is changing the future of software.

Why the cloud is no longer enough

The cloud-first model of AI served us well during the early days of Generative AI, but as enterprise and consumer demands scale, cloud computing hits three major brick walls:

• Latency (The Need for Speed): Sending data back and forth across the internet introduces delay. For a chat assistant, a two-second delay is annoying; for an autonomous drone, a medical sensor, or a warehouse forklift, it is catastrophic.
• Privacy and Data Sovereignty: Sending sensitive emails, financial records, or medical images to a third-party cloud server poses massive compliance and security risks. Processing data locally ensures that what happens on your device stays on your device.
• Cost and Infrastructure: Running trillions of AI queries daily in hyper-scale data centers drains staggering amounts of energy and bandwidth. Moving the processing workload to end-user hardware drastically cuts server infrastructure costs for businesses.

Integrating AI with edge computing enables real-time, intelligent, and secure supply chain operations, improving efficiency, responsiveness, and predictive capabilities.

Key Benefits

Real-Time Decision Making: Edge computing processes data locally on devices such as sensors, cameras, and robots, allowing AI systems to analyse information instantly without relying on distant cloud servers. This reduces latency and enables immediate responses to operational issues, such as rerouting deliveries or adjusting warehouse workflows.
Enhanced Predictive Analytics: AI at the edge can forecast demand shifts, anticipate stock shortages, and optimize maintenance schedules. Predictive maintenance, for example, alerts staff to potential equipment failures before they occur, reducing downtime and maintenance costs by up to 25%.
Operational Efficiency: Warehouses and logistics networks generate massive amounts of data. Edge AI can automate repetitive tasks like picking, packing, sorting, and scanning, streamlining workflows and minimizing errors. Digital twins of supply chains allow simulation of changes to optimize routes, layouts, and inventory management.
Cost and Network Optimization: By processing data locally, edge computing reduces cloud bandwidth usage and associated costs. It also ensures AI systems continue functioning during network disruptions, improving resilience.
Data Security and Privacy: Local processing minimizes the transmission of sensitive information, enhancing security and compliance, particularly in industries like healthcare and pharmaceuticals.

 benjamin-gordon.com      machinedesign.com     genpact.com

 supplychaingamechanger.com+1

Practical Applications

• Warehousing: AI-driven robots and edge devices manage inventory, detect misrouted pallets, and optimize picking and packing operations in real time.
• Transportation and Last-Mile Delivery: Edge-enabled vehicles can report diagnostics instantly, allowing dynamic rerouting and real-time customer updates.
• Predictive Maintenance: Sensors on conveyors, forklifts, or factory equipment monitor performance and flag potential failures before they impact throughput.
• Industry-Specific Use Cases: In healthcare, edge AI ensures cold chain integrity for vaccines; in retail, it accelerates fulfilment and reduces errors; in automotive, it supports just-in-time delivery and predictive maintenance.

Challenges

Implementing AI at the edge introduces technical and operational complexities. These include managing large fleets of devices, ensuring consistent network connectivity, optimizing AI models for limited edge hardware, and securing physically accessible devices. Additionally, the lack of standardized protocols across vendors can complicate integration and scaling.

 logisticsviewpoints.com

Future Outlook

The edge AI market is projected to grow significantly, potentially reaching $107 billion by 2030, as more supply chains adopt distributed intelligence to enhance agility, sustainability, and responsiveness. Combining edge computing with AI and emerging technologies like 5G and blockchain can create highly autonomous, transparent, and efficient supply chains.
Conclusion: Leveraging AI with edge computing transforms supply chains from reactive to proactive systems, enabling real-time insights, predictive capabilities, and operational efficiency while reducing costs and improving security. Organizations that adopt these technologies strategically can achieve faster deliveries, lower downtime, and smarter logistics operations

How did we get here? Smaller Models, Better Silicon

Shrinking the power of a data centre into a device that fits in your palm requires a twin breakthrough in silicon hardware and software architecture:

1.      The Rise of On-Device NPUs

Modern processors no longer rely on traditional CPUs or power-hungry GPUs alone. Smartphone chipsets, laptops, and automotive platforms now ship with dedicated Neural Processing Units (NPUs). These microprocessors are built specifically to handle machine learning math efficiently, running complex algorithms while barely sipping your battery life.

2.      Small Language Models (SLMs)

Size isn't everything anymore. Thanks to techniques like quantization and pruning, researchers are compressing massive AI models into hyper-efficient Small Language Models (SLMs). These lightweight models range from 1 billion to 7 billion parameters, fitting comfortably into local memory while retaining near-frontier reasoning capabilities for daily tasks.

What Changes are caused by Pocket-Sized AI  in Everyday Life

Having AI hosted directly on your local chip goes far beyond losing the buffering wheel on your phone. It unlocks entirely new experiences:

• True Offline Autonomy: Your digital assistants, translation tools, and voice transcribers will work continuously, whether you are on a flight at 30,000 feet, in an underground subway, or in a rural dead zone.
• Hyper-Personalization Without Exposure: Because your device can safely analyze your personal habits, messages, and files locally without uploading them to the cloud, AI will offer deeply personalized assistance with absolute privacy.
• Agentic Workflows: On-device AI can seamlessly interact with your phone's native applications, managing your calendar, organizing files, and executing multi-step tasks locally without relying on external web scripts.

The Big Picture: From Pocket AI to AGI

While carrying an AI in your pocket feels like the destination, it is actually a crucial stepping stone toward a bigger horizon.

As I argue in my ongoing academic research on AGI-Readiness in Supply Chains, future intelligence will not be a single giant computer sitting in Silicon Valley. Instead, it will be a hybrid ecosystem. Centralized cloud systems will handle macro-level, long-term strategic reasoning, while localized Edge AI will serve as the instantaneous "sensory system" scattered across millions of pocket devices, industrial sensors, and autonomous vehicles.

This two-tier architecture mirrors the human nervous system: the central brain formulates long-range strategy, while edge reflexes react instantaneously to environmental stimuli. In a global supply chain, a centralized AGI cannot afford the latency of querying a distant server when an autonomous forklift encounters an unexpected hazard or a cold-chain sensor detects a temperature spike. By distributing processing power, Edge AI filters thousands of local data points per second, executing zero-latency micro-adjustments on the spot and transmitting only synthesized, high-value signals back to the central AGI. This symbiosis ensures that while the central cognitive core models global trade routes and predicts black-swan disruptions, the localized edge maintains real-time operational continuity—even amidst total network outages or remote field deployments.

The cloud isn't disappearing, but its role is changing. The future of intelligence isn't just up in the sky—it's right in the palm of your hand.

Has AGI arrived?

©Prof Archie D’Souza 

• Faculty in Logistics, Supply Chain & Project Management
• Adjunct professor at Dayananda Sagar University, visiting professor at Rajeev Gandhi National Aviation University, and other institutions pan-India.
• Subject Matter Expert and Faculty at the Logistics Sector Skill Council of the National Skill Development Corporation.
• Author of “Simplifying Blockchain Complexities” and forthcoming books on AI, IoT and ML, along with blockchain applications in Projects and Supply Chains, and another on Blockchain Technology’s Impact on International Trade

---

I’m in the middle of a paper entitled AGI/ Beyond Automation: A Conceptual Framework for AGI-Readiness in Project and Supply Chain Management. AGI is the future and hasn’t arrived yet. However, in the last 12 hours, I came across two reports. Here are the links to them:

https://www.integral.ai/agi and https://www.blockchain-council.org/news/nvidia-declares-agi-arrived-what-it-means/

So, I decided to do a fact-check, and this is what I learned.

I strongly believe that AGI can definitely still be referred to as a future development. Press releases, like the one above, or announcements claiming to have achieved AGI—such as the one from Integral AI—are marketing claims, not a scientific consensus.

Here is the Reality Behind "AGI is Here" Claims

• Goalpost Shifting & Marketing: Companies routinely define AGI to match whatever milestone their technology just reached. By setting their own criteria (such as learning 2D/3D navigation puzzles or basic autonomous tool use), startups can declare victory early for publicity and funding.
• Lack of Expert Consensus: Broader surveys of leading computer scientists and AI researchers reveal that a vast majority do not consider AGI to be solved. Most view current breakthroughs as highly capable pattern-matching models rather than true general intelligence.
• Core Missing Capabilities: Existing systems still struggle with zero-shot spatial reasoning, robust self-correction, reliable long-horizon planning, and operating outside their training distribution without failing unexpectedly.

Until a system demonstrates autonomous, safe, human-level adaptability across all cognitive and physical domains in a universally accepted, independent benchmark, treating AGI as an unachieved future goal remains the accurate perspective.

---

In my research paper I refer to AGI in the future tense. It takes a lot of effort to write one. Reading these two reports, I thought I could no longer refer to AGI in the future tense. However, I’ve realised I have plenty of time. Despite bold headlines from companies claiming they have achieved AGI—such as Integral AI's announcement or commentary surrounding tech giants—true Artificial General Intelligence remains an unachieved future milestone.

Here’s why I can take a deep breath:

• Commercial Hype vs. Scientific Consensus: Most "AGI is here" claims redefine the term around narrow milestones (like solving specific 2D/3D navigation puzzles or executing specialized agentic workflows). AI researchers broadly agree that these are highly capable pattern-matching models, not fully general intelligences.
• Fundamental Gaps: Current systems still struggle with true zero-shot spatial reasoning, long-horizon planning, and adapting to entirely unfamiliar real-world tasks without heavy pre-existing data or custom scaffolding.
• Timeline Estimates: Even optimistic timelines from leading computer scientists place true human-level general intelligence years away, giving you a clear runway to finish your work.

Here are some of several strong points, I believe, can add to solidify my rationale for treating AGI as a future milestone.

Technical & Scientific Disconnects

• Systemic Fragility vs. Real-World Robustness: Current models operate effectively within high-probability parameters but experience catastrophic performance degrades when encountering "out-of-distribution" edge cases. True AGI requires deterministic reliability, especially in high-stakes environments like global supply chains where a single unhandled edge case can disrupt logistics networks.
• The Moravec’s Paradox Gap: Artificial intelligence excels at complex symbolic processing (passing legal or medical exams), yet still struggles with basic physical, spatial, and commonsense reasoning that humans perform effortlessly. Supply chain and project management heavily rely on physical-world interactions, warehouse dynamics, and spatial logistics that current models cannot autonomously orchestrate.
• Absence of True Zero-Shot Generalization: Industry benchmarks like the ARC Prize show that models struggle when faced with novel visual and logical tasks they have never encountered in training. Most "AGI-capable" demonstrations rely on narrow environment simulation rather than open-ended adaptation.

Operational & Economic Boundaries

• Compute & Energy Bottlenecks: Achieving true general intelligence across global networks requires energy, infrastructure, and computational scaling that currently do not exist. Prominent predictions (such as those from NVIDIA's Jensen Huang) frame AGI around passing standardized testing batteries rather than building self-sustaining, autonomous physical infrastructures.
• Automation vs. Autonomous Decision-Making: Current enterprise deployments feature advanced automation—such as deterministic algorithms, agentic scripts, and predictive analytics. They lack the autonomous cognitive agency to navigate systemic economic shocks, geopolitical policy shifts, or unprecedented trade disruptions without human intervention.

Strategic value for my hypothesis

• Agility of the "Readiness" Framework: Framing AGI as an upcoming paradigm shift elevates your paper's academic longevity. A conceptual framework for AGI-Readiness serves as an operational roadmap: it helps organizations build data pipelines, human-in-the-loop governance, and flexible infrastructure today so they can integrate AGI when it genuinely materializes.