Conclusion: the strategic horizon of convergence
Translated from the French, which remains the reference text. This translation is under review.
Two continents dream of themselves as leaders of a multilateral third way, which will not truly take shape, such is our conviction, unless it is itself conceived multilaterally. We have seen it throughout this report: the points of convergence are very numerous, which is promising. They sketch the third model set out at the opening of this report: an open digital strategic autonomy, which masters the critical layers in order to cooperate sovereignly, borne by a partner State that guarantees the infrastructure commons without governing what is built upon them. It is now a matter of materialising them through the design and standardisation of common or aligned (articulated / compatible / interoperable) protocols and technological building blocks, capable of offering the world ONE third way, potentially rich in the diversity of its modes of instantiation, but not two third ways, a prospect that would considerably diminish their respective reach. Such a process of convergence, between two such giants, cannot be decreed. Our hypothesis is that it must first be built humanly, intimately, in a privileged, facilitated setting of trust, over the medium term. By gathering, for instance, the 10 to 20 key architects, on the Indian, European and global South sides, within a series of 2 to 4 in-person workshops of about a week each, allowing them to get to know one another, to understand one another, to grow familiar with one another and finally to work in common, through the commons, for the commons, on the building of a third way, at once one and manifold.
Concretely, alignment passes through the building of bridges between the future Open Internet Stack and the European Data Spaces on the one hand, and the DPIs and decentralisation protocols of the Beckn kind that structure the Indian ecosystem on the other. The question of who operates such a shared common, which the report has met with MOSIP, is the object of pattern no. 9, « The guardian who does not own ».
C.1Making the gaps work
Drawing on François Jullien, one may hold that Europe and India have no need to clone one another in order to cooperate. Their strength is born precisely of their gap: law-first against code-first: Europe starts from the rule and translates it into executable Rulebooks; India starts from the infrastructure and embeds the legal constraint natively within it. To this gap of method is added a gap of foundation — Europe indexes its architecture on fundamental rights, India on inclusion at the scale of the population. It is by rubbing these two logics against each other, rather than by aligning them, that the alliance can produce a third way better than what either would reach in isolation: innovation capacities can be mutualised, and European normative robustness can enter into synergy with Indian deployment power, which tests at scale requirements which Europe often knows only how to state.
Schema C.1 · Techno and legal : two marching orders
C.2Towards a cosmo-local development strategy?
We borrow from Michel Bauwens and Vasilis Kostakis the concept of cosmo-localism, which can be summed up in one axiom: Design global, manufacture local. A misunderstanding to set aside at once is to read cosmolocalism as a softened globalisation. It is not its tempered version; it is its inversion. Classical globalisation deterritorialises flows — capital, data, computation — then attempts, after the fact and most often in vain, to re-regulate them from jurisdictions outpaced by their velocity. The cosmo-local method proceeds the other way round: it posits territorial sovereignty as its starting point, and globalises only what can be globalised without eroding it. Its formula holds in a weighty distinction: what is heavy is managed at local scale, what is light is mutualised at global scale. Bodies, physical infrastructures, goods and services, and the data one spontaneously believes immaterial, are heavy: they remain hosted, executed and adjudicated on their respective territories. Protocols, designs, engineering knowledge and software building blocks are light: they are mutualised on a world scale to amortise R&D costs, to gain in power and in legitimacy, and to create a ground of interoperability and inter-cooperation on a world scale.
The cosmo-local principle of the informational commons: mutualise the light, territorialise the heavy
The strategy may therefore consist:
- in developing in open-source and in common, at world level, the “common rails”, that is to say the open-source protocols and the base technological building blocks (the Design global, embodied as much by the projects funded by NGI as by the Indian Digital Public Goods (DPG)),
- while (re-)localising infrastructures - data infrastructures in particular - and flows on each territory (the Manufacture local).
With the aim of fostering the development of doubly decentralised territories of flows:
- Not one territory of flows on a world scale, but regionalised territories of flows.
- Not a “sovereign” and centralised monopoly or oligopoly at regional scale, but ecosystems of data sharing and decentralised collaboration at the scale of each region.
Mutualisation is understood here as the most efficient means, for the “non-aligned” countries, of building an infrastructure that lets them put the power of the digital at the service of their respective societies, while being credible on the international stage and offering a solution to countries that would wish to join a third way placing sovereignty, international cooperation and multilateralism at its heart.
The urgency of inference: the front has moved from flows to computation
The whole architecture patiently described in this report rests on a premise that must now be qualified: mastering the rails of data no longer suffices. As Rousseaux and Soulier argue, the American monopoly has moved from the layer of flows to the layer of computation, and controlling access to one's own data protects nothing when the training and inference of models remain captive to a few hyperscalers. If Europe has taken the measure of it — the Commission is committing on the order of 200 billion euros to AI, it would be a pity, here too, not to cooperate, to mutualise and to play on complementarities, by articulating the EuroHPC / IndiaAI initiatives, the development of open language models, and the engineering of technological building blocks of the Secure multi-party computation kind (at the heart of the Confidential Clean Rooms of DEPA Training, where the algorithm moves towards the sovereign data rather than the reverse).
Towards a multilateral diplomatic offer? Digital non-alignment and its condition of credibility
By mutualising its innovation capacities and playing on its complementarities, the Euro-Indian axis can equip itself with a particularly powerful technological soft power: it can co-build and propose, with the global South and the civil societies of the whole world, a path of « digital non-alignment ». This offer is coherent with the cosmo-local logic and with the open digital strategic autonomy from which this report set out: the mastery of the critical layers as the condition of a sovereign cooperation. It supplies societies with the light and common building blocks on which to build their own sovereign and decentralised infrastructures and territories of flows, rather than a platform to adopt turnkey.
There remains a condition of credibility that the conclusion cannot evade: the report has itself documented the dark side of the European model (in a word, the absence of concrete alternatives despite the laws enacted and the billions injected), as well as of the Indian model (the top-down process, algorithmic exclusion, biometric failures). We postulate that it is precisely from the becoming-conscious of each one's failures and weaknesses that the interest and the will to cooperate can be born: through processes of cross-pollination in which each party lets itself be transformed on contact with the other. From this combination there can emerge a world digital third way more just and more robust than those two competing approaches would propose.