The Super-Fast Follower | A Strategic Framework for AI and Semiconductor Sovereignty in the Global South

Executive Summary

The global race for AI and semiconductor supremacy has produced a frenzy of capital deployment, strategic anxiety, and policy reaction. Yet beneath the surface of this arms race lies a truth that few are willing to state plainly | the chips being deployed right now are not yet being used to train the best models and inference engines that we have. What we have right now is, sort of enough, sort of ground-breaking.

This whitepaper argues that countries and organisations in the Global South should not attempt to compete in the infrastructure arms race as it unfolds. Instead, they should adopt the Super-Fast Follower strategy | wait for infrastructure, but prepare for everything else. By doing so, they can become super-fast followers afterwards at maximum cost reduction, acquiring obsolete-but still transformative-infrastructure at 10 to 20% of the original cost. This is not a strategy of surrender. It is a strategy of disciplined patience.

Through the Totus Institute and partner organisations like ZedOptima, we have been advising certain groups-potentially at the margins and fringes of strategy and capability to implement this-but still representing a vast majority of organisations across the world. This report, released after one year of strategy definition and implementation under non-disclosure, sets out the framework in full.

1. Introduction | The Infrastructure Frenzy and the Sovereign Capability Gap

The current moment in artificial intelligence is defined by a paradox. Unprecedented sums of capital are being committed to semiconductor procurement, data centre construction, and compute capacity. Governments and enterprises are treating infrastructure acquisition as the singular metric of AI readiness. Yet the relationship between owning the hardware and deriving value from it remains poorly understood by most decision-makers.

Remember | the same chips that are being deployed right now are not being used yet to train the best models and inference engines that we have right now. And what we have right now is sort of enough, sort of ground-breaking. The frontier models that have shaken the developed world were trained on hardware that is, by the industry's own admission, already approaching obsolescence. This creates a structural insight that is the foundation of this report.

The sovereign capability gap between the developed world and the Global South is not primarily a hardware gap. It is a gap in preparation, strategy, talent, institutional readiness, and political will. Infrastructure is the most visible and most expensive dimension of that gap, but it is also the dimension most subject to rapid depreciation. This whitepaper reorders the priorities.

2. The Depreciation Window | Why What We Have Is Enough

So imagine what we have now-state-of-the-art and state-of-the-art for the next one or two years. Performance gains, performance capability, then extreme reduction of value, enterprise value, asset value, because they say that these same chips will be obsolete in a few years' time. But the truth is whatever has been built now and how they have been used, how these semiconductors have been used, how this infrastructure has been used, is sort of enough to have created massive frenzy now.

This is the central insight that is not being discussed right now. The semiconductor industry's own product roadmap guarantees that today's cutting-edge hardware will be available at a fraction of its cost within a predictable window. This is not speculation; it is the established pattern of every prior generation of computing hardware. What is unprecedented is the scale of capability that will depreciate.

Consider the implications-

• Current-generation chips are sufficient to train models that have already disrupted entire industries in the developed world.
• Next-generation chips will offer marginal performance gains over current hardware, not transformative ones, for the next one to two years.
• Depreciation timelines mean that infrastructure purchased today at full cost will be available at 10 to 20% of that cost within a predictable window.
• The capability ceiling of current hardware has not been reached. The models and inference engines built on existing infrastructure have not yet exhausted what that infrastructure can do.

If a few countries are patient, if a few organisations are patient, they will find that the depreciation window is not a period of waiting. It is a period of preparation.

3. The Super-Fast Follower Strategy

This is the one. This is THE report. This is the one thing that is not being discussed right now, and probably experts who know these things are waiting for this to happen and they are not disclosing this. We can proclaim that we have been advising countries and institutions to do the same and have been released from NDA after one year of strategy definition and implementation.

The Super-Fast Follower strategy is built on a single principle | if countries right now will wait for infrastructure but prepare for everything else, it is possible that they can become super-fast followers afterwards at maximum cost reduction.

Because of obsolete infrastructure, obsolete chips and semiconductors-and yes, probably we will still be 3 to 5 years behind in a few years' time. But if what we have right now is more or less more than enough to shake the world, at least the first world, the developed world, it makes no difference if we are late by 2 years, or 3 years or 5 years. If we are able to do this at a fraction of a cost, at a 10 to 20% cost reduction. The super-fast follower.

The strategy is not about catching up. It is about arriving. The difference between a nation that spends billions today to be 2 years behind and a nation that spends 10 to 20% of that amount to be 3 to 5 years behind is not a difference in capability. It is a difference in financial sustainability, strategic optionality, and institutional resilience. The super-fast follower does not compete in the race as it is currently framed. The super-fast follower redefines the terms of participation.

3.1 The Three Pillars of the Super-Fast Follower

Pillar One | Infrastructure Patience. Resist the political and institutional pressure to procure frontier semiconductors at peak pricing. Monitor the depreciation curve. Identify the inflection point at which sufficient capability becomes available at sustainable cost.

Pillar Two | Comprehensive Preparation. While waiting for infrastructure, prepare everything else | talent pipelines, data governance frameworks, regulatory readiness, institutional partnerships, model architectures, deployment strategies, and political consensus. When depreciated infrastructure becomes available, the nation that has prepared will deploy it in months, not years.

Pillar Three | Strategic Positioning. Use the preparation window to identify differentiation. Understand where the nation's unique factor conditions-geography, demographics, economic structure, diplomatic relationships-create advantages that the developed world's infrastructure-first approach cannot replicate.

4. Preparing for Depreciation, Not Collapse

Through the Totus Institute and partner organisations like ZedOptima, we have been advising certain groups. Potentially at the margins and fringes of strategy and capability to implement this, but still representing a vast majority of organisations across the world. The mandate is clear | not to prepare for collapse, not to prepare for doom, not to prepare for the bubble bursting, but to prepare for depreciation.

This distinction is critical. Much of the strategic discourse around AI in the Global South is framed in defensive terms | protecting against job displacement, guarding against cultural homogenisation, defending sovereignty against foreign model dominance. These are legitimate concerns, but they are reactive. The depreciation framework is offensive. It asks | what can you build, deploy, and achieve when the cost of capability collapses?

Preparing for depreciation means-

• Building the institutional muscle to rapidly absorb and deploy technology when it becomes affordable.
• Cultivating talent that is trained on current-generation tools and architectures, ready to scale when hardware arrives.
• Establishing data infrastructure and governance frameworks that are hardware-agnostic and ready to operate at scale.
• Developing deployment blueprints for priority sectors-healthcare, education, agriculture, public administration-that can be activated the moment compute becomes available.
• Creating procurement strategies that capitalise on the secondary market for semiconductors and pre-owned data centre capacity.

5. Factor Conditions | The Foundations of Sovereign Competency

Economics, finances, technology, talent, people, capacity, political will, political manoeuvering-what can the rest of us do? These are the factor conditions that determine whether a nation or institution can execute the Super-Fast Follower strategy. They are not optional. They are not sequential. They must be developed in parallel during the preparation window.

5.1 Economics and Finances

The financial discipline to resist premature infrastructure procurement is itself a competitive advantage. Nations that deploy capital at the depreciation inflection point will have more resources available for talent, data, and deployment than nations that exhaust their budgets at peak pricing. The economics of the super-fast follower are not about spending less; they are about spending differently, and spending later.

5.2 Technology and Talent

Technology capability is not the same as hardware ownership. A nation can build world-class AI research capability, model development expertise, and deployment engineering capacity without owning a single frontier chip. Open-weight models, cloud-based development environments, and collaborative research partnerships make this possible. The talent pipeline must be built now, so that when depreciated infrastructure arrives, the human capital exists to exploit it.

5.3 People and Capacity

Institutional capacity-the ability of government, academia, and industry to coordinate, plan, and execute-is the most underestimated factor condition. The Super-Fast Follower strategy requires institutions that can hold a multi-year preparation plan through political cycles, budget cycles, and leadership transitions. This is an institutional design challenge as much as a technology challenge.

5.4 Political Will and Political Manoeuvering

The hardest part of the Super-Fast Follower strategy is political. Leaders must resist the visible, symbolic act of procuring frontier hardware in favour of the invisible, substantive work of preparation. Political will must be sustained across electoral cycles. Diplomatic manoeuvering must secure access to depreciated infrastructure through bilateral relationships, technology transfer agreements, and multilateral frameworks before the window closes.

6. Two Paths | Scenario Planning for the Global South

There are two paths. If it is all doom and gloom, there is no point. But in case there is the likely event and likely scenario of us getting across and getting to the other side of this, what should be done?

Part of the strategy is, of course, to understand what you have and what you may have on the other side of this, and what factor conditions exist in your country, in your institutions-whether these are talent capabilities, political will, organisational interest, strategy wills, and so on.

And if we do find ourselves-whether the chips are down or whether the chips are up-what to do. So this is part of strategy and scenario planning | finding differentiation, meeting the demands of the world with the realities that might happen, and understanding the different scenarios that this might take.

6.1 Scenario A | The Depreciation Window Holds

In the most likely scenario, the semiconductor industry continues its established pattern | each generation of hardware depreciates as the next generation arrives. Current-generation chips, sufficient to train transformative models, become available at 10 to 20% of their original cost. Nations that have prepared comprehensively deploy this infrastructure rapidly and close the capability gap within 3 to 5 years. The super-fast follower strategy succeeds as designed.

6.2 Scenario B | Accelerated Obsolescence

In a more aggressive scenario, the pace of semiconductor advancement accelerates beyond current expectations, and the depreciation curve steepens. This scenario is more favourable to the super-fast follower, not less. Faster depreciation means lower acquisition costs and a shorter waiting period. The preparation window compresses, which rewards nations that began preparing early.

6.3 Scenario C | Doom and Gloom

Gloom and doom, then what for? But even doom and gloom, even global wars, even the worst that can happen-if we have forgotten about the Pandemic, if the pandemic is any reminder, we will get... well, at least the whole world, if not everyone, will get through it. One form or the other, whatever.

Even in the darkest scenarios-geopolitical conflict, supply chain collapse, economic contraction-the preparation work of the super-fast follower is not wasted. Talent pipelines, institutional capacity, data governance, and deployment blueprints are assets that retain value regardless of the macroeconomic environment. A nation that has prepared is more resilient in crisis than a nation that has merely procured hardware. The Super-Fast Follower strategy is robust across scenarios because it invests in capability rather than in depreciating assets.

7. The Advisory Framework | Totus Institute and ZedOptima

This is the strategy for the Global South. Through the Totus Institute and partner organisations like ZedOptima, we have been working with countries and institutions to define and implement this framework. The advisory process operates across four phases-

Phase One | Factor Condition Assessment. A comprehensive audit of the nation or institution's current position across economics, finance, technology, talent, people, capacity, political will, and political manoeuvering. This establishes the baseline and identifies the gaps that the preparation window must address.

Phase Two | Strategy Definition. Development of a multi-year preparation plan tailored to the client's factor conditions, geopolitical position, and strategic objectives. This plan defines what will be built, in what sequence, and through what partnerships, during the infrastructure waiting period.

Phase Three | Implementation Support. Hands-on assistance in executing the preparation plan-talent development programmes, institutional design, data governance frameworks, regulatory drafting, procurement strategy, and diplomatic engagement.

Phase Four | Deployment Activation. When the depreciation window opens, the advisory relationship shifts from preparation to deployment. Pre-built blueprints are activated, talent is deployed, and infrastructure is procured and integrated at a fraction of the cost paid by first movers.

8. Conclusion | To Leapfrog, You Must Get Close Enough to Leapfrog

This is the strategy for the Global South. To leapfrog, you must get close enough to leapfrog.

The Super-Fast Follower is not a strategy of delay. It is a strategy of sequence. It recognises that the most expensive moment to acquire AI infrastructure is the moment of peak frenzy, and that the same capability will be available at a fraction of the cost within a predictable window. It recognises that the work of preparation-talent, institutions, data, governance, strategy-is not diminished by waiting for hardware; it is enabled by it. And it recognises that the nations which prepare comprehensively while waiting patiently will not merely catch up. They will arrive.

The developed world is spending billions to be where it is today. The question for the Global South is not whether it can match that spending. The question is whether it can achieve the same capability at 10 to 20% of the cost, by being patient enough to wait, and disciplined enough to prepare.

If countries right now will wait for infrastructure but prepare for everything else, it is possible that they can become super-fast followers afterwards at maximum cost reduction. This is not a hope. It is a plan.

Infra et Ultra.