At first sight, the statement sounds reasonable. It reflects a deeply rooted way of thinking about technical objects. A knife can be used to prepare food or to injure someone. The object itself is not moral or immoral; what matters is the intention of the person using it. Applied to AI, the conclusion appears straightforward: the technology is not the problem. The challenge is simply to encourage good uses and prevent harmful ones.

But I increasingly think this analogy fails to describe what contemporary technological systems actually are.

The problem is not that technologies can produce both positive and negative effects depending on how they are used. Of course they can. What this analogy misses is that many modern technologies do far more than passively wait for individual users to decide what to do with them. They actively shape which behaviors become easier, more profitable, more legitimate, and eventually impossible to avoid.

From technological choice to technological inevitability

Seeing technology as neutral often appears alongside another common fallacy: AI is already here, so the only remaining question is how to make its uses ethical, responsible, or safe.

Both framings quietly remove something essential from the discussion: the possibility of questioning the technological trajectory itself. Public debates often move quickly from the question of whether certain technological trajectories should exist at all to the question of how they should be governed responsibly. The development and deployment of AI systems are treated as largely inevitable. The remaining discussion concerns risk mitigation, ethics frameworks, safety protocols, transparency standards, or responsible use guidelines.

Yet this framing already contains an important political assumption: that the trajectory itself is no longer really open to collective choice. The issue is no longer “Should we develop and deploy these systems in this way?” but rather “Now that these systems are here, how do we mitigate their effects?” In other words, governance becomes focused on adaptation rather than on the possibility of collectively shaping or limiting technological directions upstream.

Why technologies are not “just tools”

This matters because technologies are never deployed into a social vacuum. The idea that technology is neutral or that only "mitigation after the fact" is up for debate fails to describe how our technological systems actually emerge.

Let’s go back to our classic example: a knife can be used to cook or to kill. The object itself is supposedly neutral; morality depends entirely on the user. But large-scale modern technologies do not operate like knives.

A knife does not come with global financial incentives pushing toward specific uses. Entire industries are not organized around maximizing knife adoption. Governments do not redesign public infrastructures to integrate their use. Universities do not restructure research agendas and priorities to optimize them. Massive lobbying ecosystems are not built to normalize knife dependency across society.

Artificial intelligence, however, exists within exactly these kinds of dynamics.

Its development is shaped by enormous concentrations of capital, geopolitical competition, state strategies, industrial interests, military contracts, and research budgets. The uses that become dominant are not random. They emerge from systems of incentives and infrastructures that actively orient how the technology is designed, deployed, and integrated into everyday life.

This is why describing AI as “just a tool” is misleading. Not because individual responsibility disappears, but because individual intentions alone are insufficient to understand how technological systems evolve. A researcher may genuinely want to use AI to accelerate scientific discovery. A hospital may deploy AI systems hoping to improve diagnostics. A public administration may adopt automated systems to increase efficiency. None of these actors necessarily intend harm.

And yet the broader consequences may still include increased energy consumption, concentration of power among a handful of technology companies, new dependencies on opaque infrastructures, intensified surveillance capacities, or transformations of work that nobody fully controls. The issue is therefore not simply the morality of individual users. It is the structure of the socio-technical system itself.

The limits of the pharmakon

French philosopher Bernard Stiegler was among the thinkers who most powerfully challenged the idea of technological neutrality. Drawing on the Greek concept of the pharmakon, he argued that technologies are always simultaneously poison and remedy.[1] Writing, for example, can preserve knowledge but also weaken memory. Digital technologies can connect individuals while also producing attention capture and alienation. For Stiegler, technology could never be reduced to a neutral instrument because it always transformed human ways of living and thinking.

This remains an essential insight. But contemporary technological systems raise an additional problem that goes beyond the question of ambivalence alone.

The issue is not simply that technologies contain both danger and potential. The issue is that technological systems increasingly organize the conditions under which some uses become dominant, profitable, normalized, and scalable. They do not merely allow certain behaviors; they actively orient collective trajectories.

Social media platforms optimized for engagement are not neutral communication tools occasionally misused. Their architecture systematically incentivizes particular behaviors because attention extraction is built into their economic logic. Similarly, artificial intelligence systems developed within global competition dynamics, capitalist accumulation, and state surveillance infrastructures cannot be understood independently from those environments. The systems surrounding emerging technologies during their development and deployment make some outcomes vastly more probable than others.

Importantly, none of this requires bad intentions. Many actors involved in these transformations are acting rationally within the constraints and incentives they face. Institutions seek productivity gains. Governments seek strategic advantage. Companies seek markets. Researchers seek funding opportunities and publication capacity. Each individual decision can appear reasonable in isolation while still contributing to broader dynamics that nobody explicitly chose democratically.

Re-politicizing technological systems

This is precisely why the distinction between a tool and a socio-technical system matters politically. When technologies are framed as neutral tools, responsibility is individualized. Harm is explained primarily through misuse, unethical actors, or insufficient regulation. But once technologies are understood as systems structured by incentives, dependencies, institutional behavior, and social expectations, responsibility can no longer be reduced to individual morality alone.

The question then becomes much broader. It concerns how technological trajectories are collectively organized long before end users make any explicit choice. It concerns who finances technological development, which applications become economically viable, which forms of expertise shape deployment, and which visions of society become embedded into technical infrastructures.

This is also why current debates around “responsible AI” sometimes feel strangely insufficient. Very often, responsibility frameworks focus on reducing harms while leaving the overall direction of technological development largely unquestioned. The possibility that societies might collectively choose different trajectories, slower deployments, or even refusals in certain domains rarely enters the discussion in any meaningful way.

Yet this is fundamentally a political question, not merely a technical one.

What is ultimately at stake in the “AI is just a tool” argument is not merely a semantic disagreement. It is a political question about agency. If technologies are treated as neutral and inevitable, then collective deliberation becomes secondary. Society is reduced to adapting to trajectories already defined elsewhere.

But if technologies are understood as socio-technical systems shaped by institutions, incentives, infrastructures, and power relations, then technological development becomes something that can, and should, remain open to democratic questioning.

Not every technological trajectory needs to be accelerated simply because it becomes technically feasible. Not every application deserves deployment because it is profitable. And not every social transformation should be accepted as the unavoidable consequence of innovation.

The real challenge is therefore not only to use technologies responsibly once they exist. It is to recover our collective capacity to decide which technological systems we actually want to build in the first place.

[1] Bernard Stiegler develops the concept of technology as pharmakon across several works, notably What Makes Life Worth Living: On Pharmacology (Polity, 2013), drawing on the original Greek meaning of pharmakon as both remedy and poison.

But the more I work in innovation systems, the more I wonder whether the problem is not only the absence of responsibility frameworks, but also the kind of innovators these systems produce.

One of the most striking ideas developed by Xavier Pavie in L’innovation à l’épreuve de la philosophie[1] is precisely that innovation does not merely suffer from a lack of ethics. It suffers from a lack of freedom.

This may sound abstract at first but it’s not.

Because responsibility and freedom are inseparable. One can only be responsible for actions that one is genuinely free to choose. If decisions are entirely dictated by external constraints, then responsibility becomes blurred. At some point, we stop asking individuals to think morally or politically. We merely ask them to execute.

And if we look honestly at how innovation systems operate today, this is often what happens. Most innovators do not experience themselves as people making choices. They experience themselves as people navigating constraints. Funding priorities, quarterly objectives, industrial competition, technological race narratives, market expectations, investor pressure, publication incentives, institutional hierarchies: the space in which decisions can actually be made often feels extremely narrow.

Which choices are allowed

The result is not necessarily cynicism. Many researchers, engineers, founders or entrepreneurs genuinely want to contribute positively to society. But the system continuously reframes what counts as a legitimate choice. Certain questions become obvious (scalability, performance, market fit, growth potential) while others progressively disappear from the discussion entirely.

Who does the technology actually benefit? What forms of dependency does it create? What ways of living does it normalize? What environmental costs are rendered invisible because they happen elsewhere? And the biggest of them all: Should this technology exist at all?

These questions rarely disappear because individuals are malicious. They disappear because the system makes them difficult to sustain.

This is where Xavier Pavie’s argument becomes interesting. His point is not simply that innovation should become “more ethical”. It is that innovators themselves must change. In his words, innovation requires a form of “conversion”. Not a religious one but a cognitive and philosophical one.

Because if innovators perceive themselves as having no real choice, ethics becomes mostly decorative: a discourse added afterward to decisions already structured elsewhere. This is visible in the way many technological trajectories unfold. Once a field becomes economically strategic, the logic of acceleration tends to take over. Innovation becomes a race nobody claims to control anymore nor question.

A new innovator posture

And yet it is possible to break with this logic. Xavier Pavie gives the example of GE Healthcare and the development of a frugal electrocardiogram device initially designed for low-resource settings. What matters in this example is not the product itself. It is the fact that the initiative reportedly emerged from R&D teams who considered the lack of access to healthcare technologies unacceptable. The project was not initially driven by market optimization. It started from a social issue. Interestingly, once developed, the device also proved relevant in markets that had not even been targeted initially. The economic value followed the relevance of the solution rather than preceding it.

This may sound anecdotal, but it points toward something larger: innovation systems often assume that social concerns are constraints imposed onto technological development from the outside. In reality, they can also be starting points for invention itself. Pavie develops this idea through what he calls innovation-care: an approach to innovation grounded not only in efficiency or user experience, but in attention to interdependencies. Not only customers, but workers, ecosystems, vulnerable populations, non-human life, future generations. In that sense, it goes much further than the now-classic rhetoric of “human-centered design”, which often remains narrowly focused on users, often with purchasing power.

The difficulty, however, is that such an approach requires time, attention, doubt, and collective deliberation: precisely the things contemporary innovation systems tend to compress.

This is why the philosophical dimension matters. For decades, thinkers such as Hannah Arendt or Hans Jonas have warned that technological societies progressively separate action from responsibility. The scale and complexity of modern technical systems make consequences difficult to perceive, distribute responsibility across long chains of actors, and create situations where individuals participate in dynamics they no longer control.

Innovation is therefore not only a technical issue. It is a question about the kind of human beings institutions encourage us to become. Do innovation systems produce individuals capable of judgment, restraint, and responsibility? Or do they primarily produce actors trained to optimize within systems whose goals are never questioned?

The problem is not simply that innovation lacks ethics committees, impact metrics, or better governance frameworks. The deeper issue is that many innovators are structurally discouraged from exercising political and moral judgment in the first place.

The question is therefore not only how to regulate innovation differently. It is whether we are still capable of producing innovators who believe they are allowed to choose differently at all.

[1] Xavier Pavie, L’innovation à l’épreuve de la philosophie, Presses Universitaires de France, 2018.

But working on these tools progressively revealed an issue that long predates AI. Because the more research funding systems formalize what “good impact” should look like, the more actors adapt their language, behaviors, and proposal strategies to evaluation frameworks themselves. At some point, one starts wondering whether projects are still being designed for societal relevance, or increasingly for evaluative readability.

When impact becomes a compliance exercise

Most Horizon Europe proposals now require highly sophisticated impact sections: pathways to impact, KPIs, stakeholder engagement plans, dissemination strategies, societal outcomes, environmental benefits and increasingly detailed exploitation plans.

At the same time, calls are extremely competitive, timelines are short, and consortia are often assembled late in the process. Under these conditions, impact quickly becomes something teams must produce rather than something they genuinely have time to collectively design.

The result is a paradox. Systems designed to encourage socially meaningful research can end up generating increasingly standardized narratives about impact instead.

Certain formulations become recurrent. Certain stakeholder categories appear everywhere. Certain concepts become unavoidable. Researchers learn the codes of evaluation. Consultants, templates, and AI tools progressively reinforce these patterns further.

Little by little, projects risk converging toward what is evaluatively legible rather than what is necessarily scientifically or socially transformative. This creates a strange situation where everybody spends enormous amounts of energy producing the appearance of an impactful project while often lacking the conditions for deep collective reflection in the first place.

And everyone knows it.

Researchers feel overwhelmed by administrative layers. Project managers struggle to maintain coherence across large consortia. Evaluators read increasingly homogeneous proposals. Civil society partners are sometimes integrated too late for genuine co-creation. And societal impact itself risks becoming reduced to a formal section disconnected from scientific practice. In other words, ticking the “impact” box does not produce impact.

Research systems shape behavior

Researchers learn what kinds of projects, narratives, partnerships, and impact claims are considered credible or fundable. Institutions organize themselves around success rates and evaluation expectations. Support structures, consultants, templates, and now AI tools progressively stabilize certain ways of presenting research as legitimate.

None of this necessarily happens consciously or cynically. In fact, many researchers genuinely care about societal impact. But over time, repeated exposure to the same evaluation logics tends to normalize particular forms of discourse and project construction.

This is how entire scientific cultures progressively become influenced by metrics and funding structures. The issue is therefore not only administrative overload. It is the risk that research itself becomes increasingly formatted around what can be efficiently evaluated.

The irony is that many actors involved in European funding are fully aware of this tension. Over the past years, I have worked extensively with researchers, project managers, startup founders, innovation support organizations, and public institutions around these questions. Most people do not reject the idea of societal impact at all. On the contrary, many are deeply motivated by it.

What they reject is the feeling that impact is disconnected from actual research practices and reduced to a performative exercise. This matters because meaningful societal impact often requires exactly the things current funding structures struggle to accommodate: time, uncertainty, interdisciplinary dialogue, trust-building, conflict, experimentation, and long-term relationships with external stakeholders.

Real co-creation rarely fits neatly into accelerated proposal timelines.

AI will not solve this problem

The recent rise of generative AI tools in proposal writing makes these questions even more interesting. AI can genuinely help structure information, reduce administrative burden, support consistency across documents, and assist overwhelmed teams. But it also raises an important question: what happens when research proposals themselves become increasingly standardized through optimization tools?

If every consortium uses similar AI systems trained on previously successful proposals, there is a real risk of homogenization. Scientific originality becomes harder to distinguish. Certain ways of framing impact become dominant simply because they are statistically recognizable as successful.

This is why I have become convinced that the question is not whether AI should or should not be used in research funding. The real question is what role we want these tools to play. Should they help researchers think more clearly? (This is the design choice we made at SoScience.) Or should they progressively replace the difficult intellectual and political work of collectively defining what matters?

There is a major difference between tools that support reflection and tools that produce formatted discourse on behalf of users. The first can strengthen impactful research practices. The second risks accelerating the bureaucratization already underway.

Beyond compliance

The broader issue ultimately goes beyond AI, proposal writing, or European funding mechanisms. It concerns the growing tendency of innovation systems to prioritize compliance, sometimes at the expense of excellence. Of course, evaluation frameworks are necessary. Public funding requires accountability, coordination, and some degree of comparability between projects. Large-scale research programs cannot operate without procedures, indicators, and forms of assessment.

The problem begins when these mechanisms stop functioning as supports for meaningful research and start becoming ends in themselves. When societal impact is treated primarily as a section to complete under severe time constraints, it risks turning into a performative exercise disconnected from the actual organization of scientific work.

In that situation, everyone becomes frustrated. Researchers experience impact requirements as bureaucratic pressure rather than intellectual engagement. Evaluators read increasingly standardized narratives. Civil society participation becomes compressed into symbolic consultation exercises. And the very concept of impact slowly loses substance because too many actors learn to reproduce its vocabulary without being given the conditions to genuinely operationalize it.

The challenge is therefore not simply to ask researchers to “integrate impact.” It is to create systems where meaningful societal engagement becomes structurally possible, intellectually legitimate, and institutionally supported. Otherwise, impact risks becoming what many researchers already quietly perceive it to be: a mandatory narrative layer added onto projects whose trajectories were largely defined elsewhere.

[5] Pauline Gandré, « Les sciences : un nouveau champ d'investigation pour les gender studies », Idées économiques et sociales, 2012/1 (N° 167), p. 52-58.

Their impact however, is considered evident as long as they mention these challenges: it is rarely properly and seriously evaluated during the early stages of technological development, precisely when many of the most important choices are still open.

And this matters because impact is not something that appears afterward, once a technology reaches the market. Environmental considerations shape research directions themselves. They influence which technical options are pursued, which materials are selected, which applications are prioritized, which business models become possible, and ultimately which industrial trajectories emerge.

Through work conducted by SoScience with Bpifrance in 2024, several recurring challenges appeared across the Deep Tech ecosystem regarding environmental impact assessment and strategic decision-making.

The problem with impact measurement

One of the first difficulties is surprisingly basic: most Deep Tech startups cannot measure their environmental impact yet.

Unlike more mature companies, they are often still developing prototypes, validating scientific feasibility, or exploring industrial applications. There is no large-scale deployment, no stabilized production chain, and therefore very little measurable data.

This creates an uncomfortable situation. Startups are increasingly asked to demonstrate their environmental impact very early on, while the technologies themselves are still evolving.

As a result, many founders fear accusations of greenwashing if they communicate projected impact rather than measured outcomes.

But this confusion is problematic. Impact evaluation is not limited to measurement. It also includes projection.

At early stages, the relevant question is often not “What impact does this technology already have?” but rather “What kinds of impacts could this technological trajectory generate under different deployment scenarios?”

This is precisely why approaches such as impact pathways become important. They allow startups to explore how technological choices, industrial partnerships, market selection, or deployment conditions may shape future environmental outcomes long before large-scale commercialization.

A green technology is not automatically a positive technology

Another recurring issue is what is commonly called the rebound effect. A technology designed to reduce environmental harm can end up increasing overall consumption or extending unsustainable industrial systems.

One example discussed during the work conducted with Bpifrance was Sublime Energie, a startup developing liquefied biogas solutions. The environmental relevance of such technologies depends heavily on deployment conditions. The company must actively ensure that its solution replaces existing fossil fuel practices rather than simply adding another energy source on top of existing infrastructures. If biogas merely complements diesel-based transport systems instead of substituting them, the overall environmental benefit becomes far less clear.

This illustrates something essential: environmental impact is not only a property of technologies. It also depends on strategic choices regarding deployment, substitution, scale, and integration into existing systems.

But rebound effects are only part of the problem. Several startups interviewed during the study also highlighted what could be called “closet effects”: situations where solving one environmental issue generates new forms of ecological pressure elsewhere.

Highly efficient energy technologies, for instance, may depend on rare materials, complex extraction chains, or industrial processes carrying significant environmental costs of their own. This is one of the reasons why “green” narratives around technology can become misleading. A Deep Tech startup working on sustainability is not automatically producing environmentally positive outcomes across its full lifecycle.

The challenge is therefore not simply to optimize one indicator, such as carbon emissions, but to adopt a broader systemic perspective.

Technologies do not have one single future

One of the most fascinating aspects of Deep Tech is that many startups do not develop products. They develop technological platforms.

The same sensor technology, AI system, material innovation, or biological process may eventually serve radically different sectors.

And those sectors do not carry the same societal consequences.

A sensing technology may help monitor pollution and biodiversity degradation — or support new fossil fuel extraction systems. A robotics platform may improve accessibility for vulnerable populations — or reinforce labor automation dynamics under highly extractive business models.

In other words, impact is not only embedded in technology itself. It also emerges from strategic choices:

• which markets are prioritized,
• which investors are involved,
• which industrial partnerships are pursued,
• and which applications are considered legitimate.

This means that Deep Tech founders are not simply developing technologies. They are actively shaping future socio-technical trajectories, often long before these questions become publicly visible.

An ecosystem still poorly equipped

Despite growing awareness around sustainability, many Deep Tech startups still struggle to find adequate support structures regarding environmental impact.

Investors and public institutions increasingly ask startups to demonstrate positive impact. Yet few provide concrete operational tools allowing founders to actually explore, anticipate, and structure these dimensions during early R&D phases. Most support mechanisms remain centered on awareness and communication, while startups need practical frameworks integrated directly into technological and strategic decision-making.

The difficulty is not that methodologies do not exist. Over the past decade, numerous frameworks have been developed, particularly within European research and innovation policies, to help organizations anticipate societal and environmental impacts, explore alternative trajectories, and integrate impact considerations earlier in R&D processes.

The real issue is that these approaches remain poorly embedded within Deep Tech support ecosystems themselves.

Incubators, investors, technology transfer offices, and innovation support structures remain insufficiently equipped to guide founders through the concrete methodological implications of these questions. As a result, environmental impact is often treated as a reporting exercise rather than as a dimension capable of shaping technological choices upstream. This is especially important because Deep Tech startups occupy a very particular position: they contribute to shaping the industrial infrastructures of the coming decades.

This is where the challenge now lies: not only in training founders, but in transforming the innovation ecosystems surrounding them so they are capable of supporting different technological trajectories from the very beginning.

This article draws on work conducted by SoScience with Bpifrance on the environmental impact challenges faced by Deep Tech startups. The associated report and roundtable discussion (in French) are available in the Resources section.

This evolution explains the growing importance of concepts such as “impact pathways” within research and innovation systems.

At first glance, an impact pathway may appear to be just another bureaucratic tool. In practice, however, it reflects a much deeper transformation in the way research itself is imagined.

Because impact pathways do not simply ask researchers to measure outcomes after the fact. They ask them to project possible trajectories upstream.

From measuring impact to imagining trajectories

One of the recurring misunderstandings around societal impact is the assumption that impact evaluation is only about measurement. Of course, measurement matters. Once technologies, policies, or innovations are implemented, it is important to assess their effects and understand whether intended outcomes were achieved.

But in research and innovation, many important decisions happen long before measurable impacts exist.

Deep Tech startups, emerging scientific fields, early-stage technologies, or exploratory research projects often operate precisely in situations where direct impact data is unavailable. At these stages, what matters is not measurement in the strict sense, but projection: trying to anticipate plausible futures, possible applications, affected stakeholders, risks, dependencies, or unintended consequences.

This is where impact pathways become interesting.

In their best form, they are not simply reporting frameworks. They force innovation actors to make explicit the assumptions embedded within technological development itself. Thinking through an impact pathway means asking what conditions would actually be necessary for a technology to produce meaningful societal value, which actors would realistically benefit from it, what forms of infrastructure or governance it depends upon, and what kinds of environmental or social externalities could emerge along the way. It also means recognizing that technological choices are never isolated technical decisions, but always imply trade-offs, exclusions, and particular visions of what desirable futures should look like.

These questions are not peripheral to innovation. They actively shape technological trajectories.

Researchers were not trained for this

The difficulty is that most researchers were never trained to think this way.

Scientific training is designed to produce disciplinary expertise, methodological rigor, and technical specialization. Researchers learn how to formulate hypotheses, produce evidence, validate results, and contribute to scientific debates. They are not necessarily trained to think about stakeholder ecosystems, societal adoption, governance structures, public controversies, or systemic environmental effects.

And yet this is increasingly what innovation systems ask them to do.

Researchers are now expected to anticipate the social life of technologies before these technologies even exist. In practice, this often creates confusion and resistance. Many scientists perceive impact requirements as external administrative constraints disconnected from “real science.” Others feel uncomfortable making projections they cannot fully guarantee. Some worry that societal impact language remains vague, unstable, or politically instrumentalized.

To some extent, these concerns are legitimate. The language of impact can absolutely become performative, superficial, or bureaucratic when reduced to funding templates and standardized indicators.

But dismissing impact pathways entirely would also miss something important. Because these tools emerged in response to a real structural problem: innovation systems have historically been extremely poor at collectively discussing the broader trajectories of science and technology early enough for choices to still be possible.

The politics hidden inside technological development

One of the most interesting consequences of impact pathways is that they make visible something research systems often preferred to keep implicit: technological development is never neutral. Every innovation trajectory contains assumptions about what problems matter, which futures are desirable, who gets included, what forms of life are prioritized, and which risks are considered acceptable.

For decades, these questions were often treated as external to scientific work itself. Researchers produced knowledge. Society and politics would deal with consequences afterward.

This separation no longer holds. Climate change, biodiversity collapse, AI governance, resource depletion or public health crises, all illustrate the same issue: by the time technologies become fully operational, many structural choices have already solidified. Many decisions that will be critical for society are made during the research and development phase.

Impact pathways are therefore not simply reporting tools. They are attempts, even if imperfect ones, to reopen discussion around technological directionality before trajectories become locked in. In that sense, they subtly change the role of researchers themselves. Scientists are no longer only asked to produce knowledge. They are increasingly expected to reflect on the systems their work contributes to shaping.

A new literacy for research and innovation

Over the past decade, I have worked extensively with researchers, startups, public institutions, and innovation ecosystems on these questions. One thing has become increasingly clear to me: many actors are not opposed to societal impact at all. In fact, they often care deeply about the societal relevance of their work. The difficulty is rather that most of them have never been equipped with practical ways to collectively think through impact trajectories, stakeholder inclusion, unintended consequences, or the societal conditions necessary for technologies to generate meaningful outcomes.

This is why methodologies such as impact pathways matter when they are approached seriously. Not because they magically guarantee positive impact, but because they create structured spaces to reflect on the consequences, dependencies, assumptions, and strategic orientations embedded within innovation processes.

The danger is therefore not that research actors are increasingly asked to think about societal impact. The danger is that these practices become reduced to administrative exercises optimized for evaluation rather than used as genuine tools for reflection, deliberation, and collective orientation.

The future of responsible research and innovation will likely depend on which of these two directions ultimately prevails.

What if a “good” valorization is not what we think?

Valorization: an economic vision of the world

When we work in the world of research valorization, we have in mind first and foremost the economic question. Nothing more normal! After all, that is what valorization is all about.

Valorization of research is described as “the transformation of fundamental knowledge into new commercial products and services” [1] or the means of “making the results, knowledge and skills of research usable or marketable”[2]. At the European level the vision is similar.

In 2016, I was mobilized by the European Commission to assess European research policies in terms of Responsible Research and Innovation (RRI). I focused on the involvement of various stakeholders in funded research and innovation projects. Among my findings:

• the transformation of research into innovations (therefore valorization) concerned more than 90% of “academy – business” relations
• in 80% of cases, the “academy – non-market civil society” relationship never went beyond the debate. As if the transition to innovation was limited to the merchant circuit.

At all levels, the valorization process focus on economic results, real hardships. And the partners envisaged are reduced to companies.

Valorization of research : which indicators ?

Of course, it is not because the main attention is focused on the economic question that social and environmental issues are absent. On the contrary, there is not a strategic document that does not highlight them. However, the reflection on their subject is anecdotal. The content of the valorization work does not change in their light.

Let's do a little exercise:

• choose a research institute or university
• take 5 minutes to find the number of patents he co-owns
• take all the time necessary to find the number of beneficiaries of its valuation

It was not an easy exercise!

If it is an impossible mission, it is because the social and environmental issues are mentioned without being seriously integrated into the objectives to be achieved. The best way to be convinced of this is to ask yourself: what objectives are quantified? What is ultimately measured?

I was recently preparing a training course for the CEA on the integration of societal issues in the response to European calls for projects. By going to the websites of 3 major French research institutes (CEA, CNRS and INRAE), I found exactly the same indicators for measuring the success of valorization:

• number of patents
• number of start-ups created
• number of joint laboratories with companies

No mention of the major social and environmental challenges appears in the quantified results. This may seem anecdotal. Yet what is not measured will not be tracked or improved. Some will say that measuring social and environmental impacts is too complex! You might as well stick to monetary information, which is easy to count and compare.

“Can we deduce the societal impact of a project from its economic impact?” This question was asked to me more than once by development officers during my trainings.

Let's take a concrete example that shows that this approach is limited: a waste-to-energy project on a farm.

Let's assume that the technical device makes it possible to settle directly on the farm. Technological constraints will influence the economic dimensioning of the solution. The economic formula that will work for a given technology will define the type of beneficiaries having access to it. For example, it may take a farmer with a minimum number of hectares for the solution to be profitable. In a classic vision of valorization, you can find a direct link between the economic question and the societal question. By calculating my market (economic dimension), I can access my number of beneficiaries (social dimension). And even why not, to the tonnage of waste recycled into energy (environmental dimension). However, this approach is limited. Social impact stems from economic thinking, not the other way around. Economic success is not necessarily synonymous with societal success.

Research valorization: envisioning a different approach

Let’s imagine a translation process that places societal issues at the heart of its approach. In our example, we cannot stop at simply scaling the solution. The results of this scaling exercise—in terms of who stands to benefit—will prompt us to ask new questions: Who has access to it? Is it fair? Are certain populations particularly excluded from the solution? But also: is the energy generation capacity ambitious enough given the climate challenges? Depending on the answers to these questions, the work of the support staff and the researchers consists of revising the plan. How can we modify the technology to make it accessible? What economic mechanisms can we envision? What distribution model would be fairer? Should we modify the technology, and if so, how?

It is a shift in mindset. We cannot deduce social impact from an economic calculation, for the simple reason that calculation alone is not enough. We need a proactive approach to ensure that its commercialization is not purely economic. And to ensure this, we must use new measurement criteria!

To put climate and social issues back at the heart of research impact, we need to update the criteria and indicators that are being tracked.

Measure what matters

You are probably wondering: do you really have to go through this? The path seems strewn with pitfalls, and very few organizations know how to measure societal impact.

Wondering if it's worth it? The innovation director of a major European research institute even told me: “We don’t measure it but we do it, because it is the most important impact for us!” Other variants exist: “It happens naturally” “It is obvious” or “There is no need to force it because that is why the research exists”.

It may be my scientific background talking, but I find it hard to believe that we decide not to measure what matters most. Especially when the indicators exist!

Take the case of Nutriset.

This company was created with a very clear social mission: to end hunger in the world. All of its activities serve this mission. The promotion of research does not cut it! In partnership with the IRD, the company has revolutionized the humanitarian sector. In 1996, it launched the first ready-to-use nutrient paste (Ready-to-Use Therapeutic Food or RUTF) on the market. I can tell you about its patents, its intellectual property strategy (which it puts at the service of its social mission!), its research partnerships or its turnover. But the company uses another indicator of success that is too rarely found. Nutriset has 10 million beneficiaries. 10 million people saved from starvation thanks to the company R&D and products.

This approach to research impact is becoming more widespread. In the new European research funding programme, Horizon Europe, each project must explain in detail and relevance its “path to impact”. This impact is not only economic or scientific: it includes the social and environmental impact. Obviously, these impacts must be measurable.

To take the case of Nutriset, new indicators can be a real measure of success. 10 million people saved from starvation is still more inspiring and useful than 3 patents and 1 start-up created, isn't it?

Implementing this work of identifying beneficiaries is possible: this is what the social entrepreneurship sector has been doing for almost 30 years.

Another valorization is possible

Of course, to obtain good results on these new criteria, you will also have to modify your practices. When we are no longer looking for the same result, when performance no longer has the same face, jobs change accordingly.

This is what Assis de Souza shows in his article “A Conceptual Proposal for Responsible Innovation”. In this article, responsible innovation is compared to traditional innovation. The author shows that each building block of classic innovation finds its counterpart in responsible innovation. All the bricks are kept. In both models there is an ethical reflection, technological locks, economic impacts, consideration of stakeholders, etc. And depending on the model, these bricks do not cover the same thing, nor the same way of doing your job.

The deontological ethics of traditional innovation gives way, in responsible innovation, to an ethics of virtues and purposes. The context (country, sectors, competitors) moves from the backdrop to an active part of the innovation process. The management of economic risk becomes a management of resource and pollution risk, inclusiveness and acceptability. Beyond the quality and security of traditional innovation, the well-being is also covered by responsible innovation. Traditional innovation anticipates the next legal norms. Responsible innovation anticipates desirable futures.

At SoScience, we also insist on the fact that the results of responsible innovation are more likely to lead to “social” companies, that is to say companies whose primary objective is the social impact or environmental. Traditional innovation leads instead to traditional businesses.

Moving from traditional innovation to responsible innovation does not mean destroying the fundamental bricks of the innovation process. These are the same bricks that must gain new meaning and within which practices evolve.

Going from economic valorization to social and environmental valorization is exactly the same thing! At each stage of the process (R&D, prototyping, contracting, industrialization), new ways of thinking and new players enrich the process.

[1] Laperche, B. (2002). Le carré organique de la valorisation de la recherche: Le cas d'une jeune université dans un contexte de crise. Politiques et gestion de l'enseignement supérieur

[2] French National Council for the Evaluation of Higher Education

What is an industrial promise?

A promise is a commitment to a person or a group, formulated by a speech act, and whose performance must take place in the future. It is a complex object:

• it is both a statement (i.e. speech act, discourse) and a set of practices (from belief to the realization of the action),
• the promise most often binds two people, yet it is possible to promise oneself to oneself or to a group,
• finally it connects the present (the statement) and the future (the realization) putting the will to the test in time.

In the 17th and 18th centuries, Hobbes and Hume, among the first philosophers to think about the social contract, made the promise a central stone in their theory of human societies. The question they ask themselves is: “How can the state exist? How do human beings manage to collaborate in such large numbers, over great distances and time periods? The promise provides an answer to this question: whether it is to help each other harvest wheat (example used by Hume in A Treatise on Human Nature) or to build large infrastructures and launch rockets into space, it must be able to rely on the word of another (and many others) to successfully coordinate. Promising and keeping promises is necessary for the creation of human collectives.

While Hobbes and Hume were interested in the role of the promise in understanding how the existing society was constituted, our contemporary promises tell us of a future society. Here are two typical examples:

• Innovation promise – these promises accompany the announcement of the emergence of future technologies (regardless of their actual realization): 5G will allow us to navigate faster, space travel and terraforming will allow us to go live on Mars, etc.
• Ecological promise – this category corresponds to the "environmental pledges" for 2030 or 2050 which no large company, no more than the States, can do without today: amounts of carbon emission reduction are affirmed, the cessation or modification of certain activities, etc.

What are these promises worth ?

So what are these promises worth? It is important to note first of all that they go beyond the legal framework: companies are required to respect the law of course, but when a company talks about an action, or a product, that is going to perform in the future, the law can only serve as a mediator if these actions have been the subject of a contract in good and due form. This is not the case for innovation promises and ecological promises. These promises present the future as a horizon, that is to say a "tomorrow will be better", which makes them morally binding but not binding objects.

The question that every consumer would like an answer to is then: are they going to be kept?

Until the 20th century, it was always considered that time going by puts the promise in danger: this is the phenomenon of the weakness of the will. The will is maximum in the present moment, but gradually crumbles. Thus, I intend to do what I announce when I promise, but the passage of time is testing my will and I may change my mind. The one who promises changes his interests, his abilities, his desires and ultimately his ultimate decision to keep his word or not. With such a vision, it is legitimate to wonder how contemporary promises could be kept: innovative and ecological promises have a horizon of several decades, they are pronounced by institutional entities (companies, laboratories, public representatives) and given to an equally vague and changing entity: society. Such a time horizon should make these promises extremely uncertain and difficult to keep.

Why then do we see a proliferation of this form of discourse, a constant renewal of promises with such a distant time horizon? What is the interest of the actors concerned in making them, listening to them and believing them, if these objects are so fragile? The question that arises is not so much whether they will be held, but what purpose they serve.

How our promises shape the world, whether they are delivered or not

In the 20th century, Hannah Arendt made the promise one of the two pillars that allows man to act in her theory of human action, The Human Condition (1958). Arendt describes the "stabilizing power proper to the faculty of making promises"[1]. In a world where it is impossible to "predict the consequences of an act" just as it is impossible to "guarantee today" the continuity and stability of individuals, that is to say a particularly uncertain and unpredictable world that is impossible to control, it becomes very difficult to act. Everyone felt it during the health crisis: without knowing what awaits us in terms of confinement, displacement, etc. making decisions can be very complicated. However, the world in which we live, beyond the health crisis, is very uncertain, which could prevent us from acting (how do we know what are the right actions when nothing is controllable or predictable?) or lead us to the search for “security based on domination” of oneself and others (it would be possible to act only if everything is under control, thus leading to an authoritarian drift). For Arendt, the promise allows us to get out of these two extremes: it allows us to evolve in a world that remains uncertain, while preserving our freedom for humans. Indeed, the promises provide for men “islands of predictability” or “certainty” within an essentially unpredictable universe. It is because we are absolutely sure of our promises, of the fact that the word given will be kept, that we can act and sail almost on sight in the ocean of uncertainty that is the future.

Far from being fragile, we seem to rely on our promises as sure and strong pillars that let us know how we should act and what choices we should make in our lives. We act according to what we have been promised, considering that it is real: they promise to pick me up at the station, and I do not reserve a car to move myself; or else I hear this promise made to someone and then I know that on such a day at such a time the enunciator cannot be anywhere but at the station. I can use this information to act, and in general we do not deprive ourselves of it. This is because, to use the words of Arendt, those who are able to promise have the “capacity to dispose of the future as if it were the present”. When I believe in a promise, I act as if the world in which I found myself was the one where this promise is actually kept, and as long as nothing proves the contrary to me, it is indeed in reference to this world that guide all my decisions.

Those who are able to give their word are not only in control of the future, they also set the tone for our behavior today. Thus the promise, whether it is kept or not, has an effectiveness.

The counter-performance of the industrial world

Not everyone can activate this efficiency. Innovation promises and ecological promises cannot be made by just anyone. A major industrialist like Total can talk about its ability to reduce its CO2 emissions thanks to the development of carbon capture technologies. If a private individual holds this speech, it will not be credible since it is not able to develop these technologies. We place our trust in very specific players for this type of discourse: the vast majority of manufacturers. Historian François Jarrige reminds us that trust in innovative promises is less than 200 years old [2]. Two centuries ago they were greeted with great skepticism: faced with the ecological (deforestation), political (fear of the people) and social (pauperization and rising inequalities) crisis, the fact that technological innovation could provide answers to problems experienced was not obvious. However, "Science opens up a new regime of promise" says François Jarrige. It tells us that we no longer turn to religion or the state, that "politics" in general has been "powerless" and that is why we expect science and technique to meet our current challenges. This is what he calls the “reactivation of scientific promise”.

Knowing who has the right to speak about the future comes down in part to knowing who has the right to author it. Those in whom we place our trust by listening to their promises are today those who have the most real power over the future. Of course the legitimacy of speech is rooted in rational reasons, but that is not enough. The book Confiance, Croyance, Credit dans les Mondes Industriels under the direction of Bernard Stiegler, shows us that the actors perceived as legitimate to promise, and that we will believe, are not necessarily the best able to keep their promises. but those in whom we trust and of whom we have a good perception [3]. Beyond technical capabilities, it is community buy-in that makes pledges effective. It is therefore not because Total can develop a carbon capture technology (which is by the way not certain) that we believe it can, but because we believe it that we decide to trust in the future that is proposed to us, and that we give power to the promise by acting today as if what was promised was assured. The promises allow the actors who state them, for themselves but also for all those who believe in them, to act today as if they could bend the future according to their wishes.

However, the promises of future technical solutions such as carbon dioxide capture technologies are a dangerous trap that push us to further increase our consumption and our CO2 emissions, warn us climate scientists [4]. While the promise will not necessarily be kept, it is devilishly effective: all the players who believe in it do not change their consumption habits, since a solution will come, that's for sure! By doing so, what was thought to be avoided (an increase in CO2 in the atmosphere) is precisely what happens [5]. Contemporary industrial promises are effective, but they are often counterproductive: they achieve the opposite of what they promise.

Faced with our environmental emergency, can we content ourselves with accepting the promises that are made?

Those who do not have the power to promise today, who do not participate in the promises of innovation and ecological promises, that is to say the citizens, are both in a situation of vulnerability, at the mercy of those who promise, and of political powerlessness since their inability to be the enunciators sheds light on their lack of power. How to take back this power?

We could outline a beginning of the way forward to reclaim the future: give citizens the means to make and keep these promises, whether they are innovation promises or ecological promises. The democratic debate must spill over into our discussions of the future and of technical choices.

[1] Hannah Arendt, Condition de l’homme moderne, 1958, Editions Pocket, p.310-314, as well as for all the quotations in quotation marks that follow.

[2] François Jarrige is a historian, lecturer in contemporary history at the University of Burgundy and author of "Technocritics: From the refusal of machines to the contestation of technosciences". The quotes are taken from the Atécopol conference of 20/03/2019 available online: https://www.youtube.com/watch?v=0tUBJXHsjLs

[3] Confiance, Croyance, Credit dans les Mondes Industriels under the direction of Bernard Stiegler. See in particular Chapter Five, Trust and Political Economies written by Laurence Fontaine, historian.

[4] Climate scientists: concept of net zero is a dangerous trap, James Dyke, Robert Watson, Wolfgang Knorr, April 22, 2021, The conversation, https://theconversation.com/climate-scientists-concept-of-net-zero-is-a-dangerous-trap-157368

[5] On the ineffectiveness of the commitments made by large companies, see in particular Overselling Sustainability Reporting, Kenneth P. Pucker, May-June 2021, Harvard Business Review, https://hbr.org/2021/05/overselling-sustainability-reporting

This is why these collaborations remain the exception, not the norm.

A system designed around other actors

Scientific research and technological innovation do not take place in a neutral environment. They are embedded in institutional, financial, and political systems that define who can participate, under what conditions, and for what purposes.

Public research is often presented as open and oriented toward the common good. Political initiatives reinforce this narrative. When Emmanuel Macron launched #MakeOurPlanetGreatAgain in 2017, the message was clear: science must mobilize to address global challenges. In the end, the initiative only funded 42 researchers at best. That is not much for one of the most important issues of our century. Beyond the signal, the structure remains largely unchanged.

Even when new spaces emerge, such as the Innovation Campus for the Planet at IRD, the Societal Impact community launched by CEA, the D-Lab at MIT, or Tech4Impact at EPFL, they remain peripheral to the core functioning of research systems.

They create openings. But they do not redefine the rules.

Those who are able to engage with research are still those who already master its codes: large corporations, well-funded startups, and institutional actors. Others struggle to even enter the conversation.

Social enterprises, NGOs, and non-profit organizations rarely fit these models. The research world has little to no understanding of the concerns of civil society actors on the field. “Localized research on an issue specific to a territory or carried out by civil society, is of little interest to classic funders of research because the knowledge generated is less universally applicable” reminded us Lionel Larqué, Director General of Alliance Science-Société (ALLISS) during the Summer University of the Impact France Movement in 2020.

This is not an explicit exclusion. It is a system effect.

What gets built, and what does not

These structural conditions shape the direction of technological development. When research agendas are defined within specific institutional and economic constraints, the range of possible innovations narrows. Some questions are explored extensively. Others remain at the margins.

As Christophe Roturier highlights, Delegate for Science in Society at INRAE*, the evaluation of research does not encourage researchers to develop research collaborations in partnership with civil society. In France, the multi-annual planning which guides research budgets until 2030 does not mention (and therefore does not encourage) collaborations with citizens and civil society (NGOs, social entrepreneurs).

Technologies aligned with industrial demand are more likely to be developed. The result is not necessarily “bad” technology. It is a partial one.

And that partiality matters. Because what is not developed is as important as what is.

The limits of downstream action

In this context, social entrepreneurship tends to intervene where it can. Most initiatives operate downstream: using existing technologies, adapting them, building services around them, or attempting to mitigate their negative effects.

Some initiatives attempt to move upstream. For instance, programs such as The Future Of, developed by SoScience and recognized by the United Nations as a Good Practice for the Sustainable Development Goals, aim to connect researchers and social entrepreneurs to co-develop research projects rooted in social and environmental needs.

These experiments show that other configurations are possible. But they also reveal their limits. They require navigating institutional constraints, accessing funding mechanisms that are not designed for them, and building legitimacy in environments where impact is not the primary evaluation criterion.

Support and investment strategies do not include the social and environmental dimension in the criteria grid used for rating and funding research projects. Social impact is perceived as a positive externality of research projects, but it is not what sets the creation of research consortia in motion.

Support staff inside knowledge transfer offices and innovation department at universities are very well versed on economic return on investment. They mainly have to answer to economic policies, and do not know how to integrate the societal impact on society. The economic KPIs exist: economic growth, job creation, spin-off creation. The social KPIs are non-existent, or not well-known, and therefore the societal impact is not thought through.

Funding rules are not adapted to social entrepreneurs who want to invest in science and the barriers for scientists to collaborate with them are very high.

Beyond collaboration: a question of power

Calls for more collaboration between researchers and social entrepreneurs are increasing. They are necessary, but they do not address the core issue.

The problem is not only access. It is power.

Who defines research priorities? Who decides which problems deserve attention? Who controls the resources that make certain developments possible?

As long as these questions remain unchanged, new forms of collaboration risk remaining peripheral, integrating social actors sporadically without allowing them to shape the system.

Engaging with science and technology therefore requires more than exceptional partnerships. It requires the ability to influence agendas, to participate in decision-making processes, and to redefine what counts as success.

Tackling social and environmental challenges through research will not happen just through good will and public announcements. The research and innovation system has to favor collaboration with society, including social entrepreneurs, and integrate social and environmental criteria to the evaluation process.

The question is often framed as follows: how can social entrepreneurship engage more with science and technology?

A more accurate formulation would be:

What would it take for the research and innovation system to be shaped by a broader set of actors and values?

This is not a question of inclusion. It is a question of transformation. And it is, ultimately, a political question. Pressing challenges such as climate change forces us to rethink our innovation model. The social role of universities is being challenged and new ways to interact with society are needed. I therefore envision that answering this question will soon become inevitable.

Between maintaining distance and militant activism, we believe there is a third path for the academic world to explore: co-creation.

Feminist Studies: Knowledge Rooted in Society

Beginning in the late 1970s and throughout the 1980s, feminist studies demonstrated that the issue of sexism in science has not only a social dimension but also an epistemological one. Thus, sexism is not only present in laboratories, where there are fewer female researchers than male ones. Moreover, it influences research findings and conclusions [1].

By highlighting the sexist dimension [2] of much scientific knowledge (in the natural sciences as well as the social sciences), feminist studies have shown that it is impossible to ignore that the producers of scientific knowledge are situated social actors. Researchers belong to a particular context that necessarily influences their way of understanding reality.

“Behind the most mundane, least committed, and most neutral description lies a perspective on reality that is situated and rooted” [3].

In other words: absolute neutrality likely does not exist in the form we sometimes like to imagine it.

Neutrality as an Institutional Myth

Some stakeholders wish to continue defending the absolute objectivity of research findings. To this end, they cite errors, isolated cases, or instances of “bad science.”

“According to a general definition of sound scientific practice, this gender bias should not occur; it is a deviation; [...] if science (with a capital S) were to operate according to its very nature, it would not exhibit such aberrations” [4].

In short, while there are indeed sexist studies, this is not a reason to question research practices. It is individuals who are at fault for misapplying the scientific method.

On the contrary, the work of researchers in the humanities and social sciences shows us that it is “socially constructed differences that lead to inequalities and discrimination—hidden behind a façade of neutrality” [5].

This does not mean that facts become relative. It does mean, however, that the production of knowledge is never entirely separate from the social, cultural, economic, and political structures within which it is embedded.

This is where the question of co-creation becomes interesting.

If we must accept that every actor is inevitably situated, then it becomes essential to allow a wide range of actors involved in the same issue to participate in shaping research findings. Co-creation thus emerges as a new research practice that helps reduce certain biases stemming from individual blind spots.

Can scientists be engaged?

A researcher is a citizen. However, their profession also comes with a code of ethics: it is not the researcher’s opinions, ideas, or beliefs that should dictate the results of their research.

Nevertheless, as sociologist Eric Fassin points out, much research is conducted not out of “a kind of disinterested intellectual curiosity” but because there is behind it “a desire to change the world.”

Moreover, this tension extends far beyond the humanities and social sciences alone. Do researchers working on the next generation of batteries not have a vision of a different world in mind? Are researchers in ecotoxicology or ecology any more disinterested and impervious to societal issues?

The idea that only certain disciplines are “political” while others produce perfectly neutral knowledge is hard to sustain. The era we are living in is, moreover, no small factor in the redefinition of this profession.

“Science and research must increasingly learn to step out of their ivory tower, which means becoming involved in the social environment to which they belong ,” stated Joseph Taradellas, professor of ecotoxicology at the Institute of Environmental Engineering at the Swiss Federal Institute of Technology in Lausanne (EPFL), during the Citizen Science symposium held on April 6, 2018, on the topic of responsibility in scientific research.

Co-creation as a research practice rooted in societal challenges

At SoScience, we advocate for and promote science that is engaged, science that sees itself and positions itself as a stakeholder in society, attentive to societal concerns and capable of taking action.

This stance does not involve turning researchers into activists. It involves recognizing that research directions, the problems deemed legitimate, the uses of developed technologies, and the resulting impacts are already deeply shaped by collective and political choices.

This research must be able to guide and leverage research outcomes to address societal concerns. A number of researchers already adopt this stance, and the research institutes with which we collaborate are also beginning to integrate it at the policy and structural levels.

To achieve research outcomes that are grounded in our current societal challenges and social concerns, we offer methods that enable co-creation with civil society (citizens, NGOs, social entrepreneurs, and scientists).

Specifically, since 2016, our The Future Of open innovation programs have facilitated multi-stakeholder research collaborations with civil society on issues related to social and environmental challenges.

Whether it concerns water in urban environments, soil health, epidemic models at borders, or energy for mobility: all these topics benefit scientifically when approached with diverse stakeholders.

The question, then, is perhaps not whether researchers should be engaged. Rather, the question is: who is authorized to participate in defining scientific problems, technological futures, and the knowledge considered legitimate?

[1] Dominique Pestre (2006), Introduction aux Science Studies, Chapitre 5 Femmes, genre et science : objectivité et parti pris.

[2] « By ‘sexist,’ on the one hand, the fact that this body of knowledge reproduces the most common prejudices regarding relations between men and women, making them the backbone of its discourse and legitimizing them. On the other hand, it means naturalizing the difference and/or inequality between women and men. ».

[3] Florence Piron (2019), Et si la recherche scientifique ne pouvait pas être neutre ?, edited by Laurence Brière, Mélissa Lieutenant-Gosselin and Florence Piron, chapitre 9 « L’amoralité du positivisme institutionnel », pp. 135-168. Québec : Éditions science et bien commun.

[4] Dominique Pestre (2006), Introduction aux Science Studies, op. cit.

[5] Pauline Gandré, « Les sciences : un nouveau champ d'investigation pour les gender studies », Idées économiques et sociales, 2012/1 (N° 167), p. 52-58.

We can see this, and it has been widely denounced, in certain institutions. This is much less the case, however, when we move away from visible institutions (police, justice, employment) to focus on fields perceived as more neutral, particularly science and technology. These fields continue to enjoy a special status: they are seen as objective, neutral, and external to social relations. Here, another idea persists: that technical objects are, by nature, beyond suspicion.

It is precisely this idea that must be questioned.

Scientific Neutrality: A Convenient Fiction

Science enjoys a special status. It is associated with objectivity, rigor, and a form of truth that transcends individual interests and biases. This perception is not merely a matter of trust in scientific methods; it also has a political effect: it tends to place scientific and technical outputs beyond the reach of ordinary criticism. It also allows us to avoid asking certain questions: if technical objects are neutral, then the inequalities they produce can only stem from their uses.

In this context, inequalities are most often thought of in terms of access: who becomes a scientist, who does not, who funds research, who benefits from it. These questions are essential, but they overlook a more subtle and troubling aspect: the way in which technical objects themselves can embody and reproduce social hierarchies.

In other words, it is not just a matter of knowing who does science, but what science and technology do.

When Bias Becomes Infrastructure

Research on facial recognition systems offers a particularly illuminating example. In 2019, a study conducted by the National Institute of Standards and Technology showed that false positive rates were significantly higher for African American and Native American populations than for white populations. The gap is not marginal: it can range from one to ten, or even higher depending on the systems tested.

This is not a technical detail. When these tools are integrated into police or judicial systems, they result in mistaken identifications, unjustified stops, and the arrest of innocent people. And depending on your skin color, the consequences are not abstract.

In response to these findings, one explanation frequently arises: these systems are still imperfect and will improve over time. This interpretation has the advantage of preserving the idea of a fundamentally neutral technology that is merely temporarily flawed. All that is needed, it suggests, is to improve the models, expand the data, and invest further. But this interpretation misses the point.

The biases observed are not merely the result of technical errors. These biases have been known, documented, and criticized for years. Their persistence speaks to something else: they are not corrected because they are not treated as a priority. What is at stake is not just a technical limitation, but an implicit hierarchy of what is deemed worthy of being resolved and what is not.

What these systems reveal is not simply a difficulty in recognizing certain faces. It is a hierarchy, embedded in design practices, of audiences deemed important and those deemed less so.

What the History of Technology Teaches Us

This phenomenon is not new. Long before artificial intelligence, other technologies produced similar effects. For decades, early photographic film—widely used in the mid-20th century—was calibrated to accurately reproduce light skin tones. Black faces appeared underexposed, details were lost, and contrast was poor.

The problem was well known. Yet it was corrected only belatedly, and for reasons that speak volumes about the underlying logic at work. It was not primarily criticism from the affected communities that led to adjustments, but complaints in the 1970s from manufacturers—particularly in the chocolate and furniture industries—whose products were poorly rendered in advertising photos.

This is not an interpretation. It is a fact. Kodak’s former director of research acknowledged it himself: “Black skin was never considered a serious problem at the time.” In other words: the company only resolved the issue when it affected economic interests deemed significant.

But another question sheds even more light on this issue. Why didn’t the communities directly affected bring this issue to the forefront sooner, even though the 1970s were marked by major civil rights struggles?

The answer, documented by Lorna Roth, is brutally simple: “The general public believed that these things were based on science and therefore could not be changed.” [1]

Thus, scientific neutrality does not merely serve to describe the world: it creates a barrier. It makes certain injustices harder to challenge, because they appear technical, objective, almost natural. It is not merely that certain problems go unaddressed. It is that, in some cases, they become unthinkable as problems for those who are their victims.

A consequence of these two trends: what is considered a technical flaw depends less on its scale than on the importance accorded to the people or objects it affects. Thus, technical trajectories are not simply guided by objective constraints, but by value systems.

Shifting Perspectives

Taken together, these examples reveal a pattern. Technologies do not merely reflect the social world; they extend certain of its underlying logics. Through their modes of operation, they reflect priorities, blind spots, and hierarchies. They are not external to power relations: they are one of their expressions.

This does not mean that they are intentionally designed to discriminate. The question of intent, while it may have moral relevance, is insufficient for understanding the observed phenomena. What is at stake involves broader configurations in which technical, economic, and organizational choices converge.

If we accept this, then part of the debate must be reframed. It is no longer merely a matter of identifying problematic uses or malicious actors, but of examining the conditions under which certain technological trajectories become possible, legitimate, and sustainable.

To continue viewing science and technology as neutral spaces is to deprive ourselves of part of the analysis. This leads us to seek the causes of inequalities where they are most visible, while leaving intact the structures that contribute to producing them elsewhere.

Recognizing that everyday objects can embody forms of privilege does not mean attributing intent to them. It means taking seriously the fact that they are the product of social contexts, and that, as such, they can reproduce those contexts. Otherwise, the risk is clear: forgetting that the fight for justice also involves scientific and technical objects, from artificial intelligence algorithms to the molecules we bring to market.

[1] Lorna Roth, scholar at Concordia University: Roth, Lorna. April, 2009. “Looking at Shirley, the Ultimate Norm: Colour Balance, Image Technologies, and Cognitive Equity,” in Canadian Journal of Communication. Vol 34, No. 1, 2009: 111 – 136.

Alors pourquoi n’écoute-t-on pas les scientifiques climatologues quand on écoute les médecins ? Pourquoi acceptons-nous de tout changer du jour au lendemain alors que, pour des conséquences pourtant annoncées comme bien plus graves, nous refusons des modifications bien moindres ?

Trop tard

Aurélien Barrau souligne avec raison le double constat : dans les deux cas il s’agit bien 1) d’alertes scientifiques largement documentées et 2) d’un choix politique. Dans le cas du Covid-19, les épidémiologistes nous ont prévenus et le confinement est imposé par le gouvernement comme réponse politique coercitive touchant à la fois l’économie et les libertés individuelles. Dans le cas du réchauffement climatique, les climatologues alertent depuis des décennies, pourtant les mesures prises restent extrêmement faibles.

La différence fondamentale est que dans le cas du Covid-19, la crise est là. Ou plutôt : elle est exactement sur nous, avec ses conséquences directes, visibles, immédiates. Parce qu’il ne faut pas s’y tromper : les scientifiques alertent depuis plus de 10 ans sur l’arrivée d’une telle pandémie. Comme pour le réchauffement climatique, ils n’ont simplement pas été écoutés. L’oreille attentive n’a même pas été tendue en décembre, janvier ou février quand la crise touchait déjà la Chine puis l’Italie. Il a fallu attendre mars pour que la décision politique coercitive soit prise.

En temps normal, l’action politique reste principalement structurée autour d’un objectif de stabilité économique et de croissance. Toute donnée scientifique n’est réellement prise en compte que lorsqu’elle peut être intégrée à cet agenda. Nos systèmes de recherche eux-mêmes sont largement évalués à travers des indicateurs économiques : nombre d’entreprises créées, brevets déposés, capitaux levés, compétitivité. Lorsque les données scientifiques entrent en contradiction avec ces priorités — climat contre croissance, pandémie contre économie — alors l’action politique est retardée autant que possible, jusqu’au moment où la crise devient impossible à ignorer.

Pour le Covid-19 comme pour le climat, cela signifie souvent la même chose : agir trop tard.

L’irréconciliable trio science, société, politique

« Comment ? Nous nous gargarisons de la qualité de notre recherche, de nos technologies, mais nous n’écoutons pas nos scientifiques ? »

Pour faire simple, je vais m’intéresser à deux rôles communément endossés par l’institution scientifique : l’alerte et la création de technologies ou de solutions.

Quand ils sont lanceurs d’alerte, les scientifiques ne sont plus les acteurs dociles et efficaces de la politique de recherche. L’alerte devient alors un lieu de convergence entre science et société : les manifestations pour le climat en sont un exemple, tout comme Greta Thunberg demandant aux responsables politiques d’écouter les scientifiques. Mais cette configuration laisse le politique dans une position particulière. Car face aux enjeux climatiques, comme face aux coronavirus, sonner l’alerte ne suffit pas. Il faut aussi produire des réponses concrètes, des solutions efficaces : par exemple, si on reprend le cas du coronavirus, des médicaments [1].

Les scientifiques ne peuvent évidemment être qu’une partie de la solution. Ils peuvent indiquer, proposer, documenter, expérimenter. Mais ils ne peuvent pas décider seuls, car toute décision aux implications collectives reste fondamentalement politique. L’invention de solutions scientifiques et leur mise en application est une véritable union des politiques publiques et de la science. Le problème est que notre système de recherche et de valorisation n’est pas réellement organisé pour répondre rapidement à des crises systémiques de long terme [2] et laisse la société civile en dehors des décisions prises. Il sait produire de l’excellence scientifique et de la compétitivité technologique ; il peine beaucoup plus à orienter durablement ses capacités vers des besoins collectifs définis démocratiquement.

Notre système de recherche peut donner l’alerte, mais ne mène pas directement à l’action. Il peut développer des solutions, mais souvent pour les mauvais problèmes. Nous avons une abondance d’objets connectés permettant à des individus en parfaite santé de mesurer leurs constantes biologiques en temps réel, mais aucun traitement immédiatement disponible contre les coronavirus malgré les alertes répétées depuis le SARS en 2003.

Ce décalage n’est pas le signe d’un système absurde ou irrationnel. Il révèle surtout que ses priorités restent celles du siècle précédent : compétition économique, innovation marchande, croissance, souveraineté industrielle. Non pas que ces objectifs soient illégitimes en soi, mais ils ne suffisent plus à répondre aux enjeux contemporains.

La question n’est donc pas simplement celle de “l’écoute de la science”. Les responsables politiques savent parfaitement écouter les scientifiques lorsque cela sert leurs priorités stratégiques, économiques ou industrielles.

La question est plus profonde : quels problèmes notre système de recherche est-il réellement conçu pour résoudre ? Et qui participe à définir ces priorités ?

Réconcilier l’alerte et l’action

Réconcilier science, politique et société, c’est aussi réconcilier le temps de l’alerte avec celui de l’action.

Les sciences citoyennes, l’innovation ouverte, de nouvelles formes de co-construction entre chercheurs, citoyens, associations, institutions publiques et acteurs économiques sont autant de pratiques encore émergentes mais essentielles à explorer. Non parce qu’elles seraient miraculeuses, mais parce qu’elles ouvrent la possibilité d’un autre rapport entre production des savoirs et décisions collectives.

Nous restons aujourd’hui au tout début de ces transformations. Ces pratiques, mêmes naissantes, sont nécessaires aux politiques publiques de recherche à venir [3] et doivent préfigurer une nouvelle norme. Il devient de plus en plus difficile d’imaginer que les crises à venir pourront être affrontées avec des systèmes de recherche conçus principalement pour produire de la compétitivité économique.

Le problème n’est pas uniquement que les scientifiques ne soient pas écoutés. C’est que nos institutions savent encore mal transformer l’alerte scientifique en capacité collective d’action.

[1] Bruno Canard, chercheur au CNRS s’indigne du fait qu’un médicament pourrait déjà être fonctionnel si les recherches sur la famille des coronavirus avaient été financées de façon continue depuis la crise du SARS, également causé par un coronavirus, en 2003

[2] La mise en place d’appel à projet « flash » pour financer en urgence des recherches sur le Covid-19 est révélateur. A quand un appel à projet flash pour découvrir en 3 mois comment restaurer la biodiversité ?

[3] Notamment la future loi de programmation pour la recherche dont la discussion parlementaire a été retardée suite au Covid-19

The limits are not theoretical

Much of today’s technological development is built on an implicit assumption: that resource constraints can be managed, optimized, or eventually overcome. Efficiency gains, new materials, better recycling processes: the narrative of innovation is one of continuous adjustment.

This is where the idea of a circular economy fits in. In theory, materials are reused, waste is minimized, and growth can continue without increasing pressure on resources.

In practice, things are less stable.

Many contemporary technologies rely on complex combinations of materials — rare metals, dispersed components, intricate alloys — that are extremely difficult to recover once used. Recycling is partial, energy-intensive, and often results in a loss of functionality. Some materials are simply dispersed beyond recovery.

These are not marginal issues. They set physical limits. A technology that depends on non-recoverable resources cannot scale indefinitely, regardless of how efficient it becomes.

From this perspective, the question is not whether we can optimize our current systems. It is whether the direction itself is viable.

What are we actually innovating for?

Beyond resources, low-tech introduces a more uncomfortable question.Not how we produce. But why.

The current landscape of innovation offers no shortage of examples: connected devices designed to perform trivial tasks, products adding layers of functionality without clear utility, systems optimized for convenience at the cost of complexity.

A connected juice machine to press pre-packaged fruit. Socks infused with silver nanoparticles to control odor. A connected bikini measuring sun exposure. Automated pet feeders controlled remotely.

These examples are often dismissed as anecdotal. They are not. They reflect a broader orientation: innovation driven by technical possibility and market opportunity, rather than by necessity or collective benefit.

In this context, low-tech is not primarily about simplifying objects. It is about questioning the legitimacy of the needs they respond to. Do we need this product? What does it actually improve? What does it cost, in resources, in complexity, in dependency?

These questions are rarely central in innovation processes. Low-tech makes them unavoidable.

Shifting the terms of innovation

This is why low-tech cannot be reduced to a set of technical solutions. It is a shift in perspective.

It challenges the idea that innovation is inherently desirable, that more technology necessarily leads to better outcomes, that efficiency and productivity are sufficient criteria to guide technological choices.

It also complicates the opposition between low-tech and high-tech.

In practice, the question is not to reject advanced technologies altogether. Some of them create undeniable value: in healthcare, for instance, where imaging technologies or quite complexe medical devices can radically improve outcomes.

The issue is not the level of technology. It is where, how, and for what purpose it is deployed.

In many cases, simpler systems prove more robust, more accessible, and more adaptable. In urban logistics, for example, solutions like electrically assisted cargo trailers can replace short-distance deliveries otherwise carried out by vans, reducing congestion, emissions, and operational costs at the same time. In construction, modular wooden structures designed to be disassembled and reassembled challenge the permanence and waste associated with conventional building methods. In agriculture, approaches based on natural processes, such as using plants to extract excess metals from soils, offer alternatives to energy-intensive remediation techniques.

These examples do not form a unified model, a one-fits-all. They point to different ways of organizing production, use, and value.

What low-tech reveals

Seen from this perspective, low-tech is less a category of technologies than a way of interrogating innovation itself.

It forces a shift from means to ends.

Why do we innovate? Which problems are considered worth solving? Which constraints are taken seriously, and which are ignored?

In a context defined by resource limits and ecological pressures, these questions are no longer theoretical. They determine which systems can persist, and which will eventually fail.

Low-tech does not provide a ready-made alternative. It does something more demanding. It reintroduces the idea that technological choices are not inevitable, and that they can, and should, be debated.

In France, the deployment of autonomous vehicles is identified as a national priority[1]. Across the world, major technology companies and automotive actors are investing heavily, racing to bring these systems to market.

At this stage, the question is often framed in relatively simple terms: how do we make autonomous vehicles safe, reliable, and acceptable? This framing is part of the problem.

What is really being optimized

Autonomous vehicles are not a single innovation. They are the convergence of multiple systems: artificial intelligence, sensors, infrastructure, data flows, regulatory frameworks, business models.

And within this system, choices are constantly being made.

What level of safety is “acceptable”? What trade-offs between cost and redundancy? What data is collected, and for what purpose? What infrastructures are required, and who pays for them? Who gets access to these services, and under what conditions?

These questions have very concrete implications.

A system optimized for dense urban areas does not address rural mobility. A model based on high upfront costs and complex infrastructures limits accessibility. Replacing drivers with automated systems raises immediate questions about employment and transitions. Collecting large volumes of mobility data creates dependencies and risks that are not neutral.

Yet these elements are often treated as secondary constraints, rather than as central design choices.

When I was invited to speak about responsible innovation at Alphabet's X Development (ex- Google X Labs) - deeply involved in autonomous vehicle development — what was striking was not the absence of concern for these issues. It was how they were positioned.

The discussion is not whether autonomous vehicles should be responsible. It is which dimensions of responsibility can be integrated without slowing down development, increasing costs beyond acceptable thresholds, or creating too much uncertainty for deployment.

Responsibility is part of the equation, but it does not define it.

What “responsible” actually changes

Much of the discourse around autonomous vehicles assumes that innovation can be balanced: that economic, social, and environmental benefits can be aligned if the technology is well designed.

In practice, these dimensions do not carry the same weight.

Economic performance (growth, competitiveness, return on investment) remains central. It determines where investments go, which use cases are prioritized, and which actors can participate.

Other dimensions (accessibility, inclusion, environmental impact) are addressed, but under conditions. Rural accessibility, for instance, depends on infrastructure investments that are rarely prioritized. Environmental gains are often tied to assumptions about shared mobility that are not guaranteed. Inclusion depends on pricing models that may or may not emerge.

These are not implementation details. They are structural trade-offs. By the time questions of responsibility are formally addressed — through regulation, ethical guidelines, or design adjustments — many of these trade-offs have already been made.

At that stage, responsibility becomes a matter of correction: improving safety margins, protecting data, adjusting services.

But the trajectory is already set. This is the limitation of the idea that technologies can simply be made “responsible”. It assumes that responsibility can be added after the fact.

In reality, it depends on who is involved early on, which problems are prioritized, and which constraints are considered non-negotiable.

What this example shows

Autonomous vehicles make visible a broader dynamic.

Technological systems are not neutral. They are shaped by industrial strategies, political priorities, available infrastructures, and economic models. These elements define not only what is possible, but what is pursued.

Once a direction is taken, it becomes increasingly difficult to change. Investments are locked in, infrastructures are built, ecosystems align.

The question is therefore not only how to make a given technology more responsible.

It is whether we are able to influence the conditions under which it is developed in the first place, before certain choices become irreversible.

This raises a more concrete and often overlooked issue: who is actually in a position to shape these technologies?

Who is around the table when early design choices are made? Which actors have access to the resources, infrastructures, and partnerships required to participate? Whose needs are considered legitimate, and whose are not even formulated?

Today, these spaces are largely occupied by industrial actors, large technology companies, and public institutions aligned with their priorities. Other actors (local communities, civil society organizations, social entrepreneurs) are rarely involved at the stages where key decisions are taken.

This is where the question of co-creation becomes central. Without early and meaningful involvement of a broader set of actors, the same patterns tend to reproduce themselves: technologies are developed for the contexts, users, and models that are already prioritized.

The question is often framed in simple terms: how do we make autonomous vehicles safe, reliable, and acceptable? But this question hides a more fundamental one.

Who is this “we”?

Is it the companies developing the technology? Public authorities regulating it? Engineers designing the systems? Or the people whose lives, jobs, and environments will be affected by it?

As long as this “we” remains undefined, the framing itself is misleading. It suggests a collective capacity to act, while decisions are in fact concentrated in the hands of a limited set of actors.

This is why the problem is not only technical. It is about who has the power to decide what responsible innovation means, and for whom.

[1] Les Echos, 31/01 2018.

The tech world is going through a massive hangover.

The platforms that once embodied the promise of a more open and collaborative economy have revealed a very different reality: attention extraction, large-scale capture of personal data, distortions of competition, and, increasingly, direct impacts on democratic life. What was supposed to empower users has, in many cases, concentrated power.

For those who believed that technology could contribute to more inclusive and open societies, the moment is one of disillusionment. Not because the problems are new, but because they are now impossible to ignore.

This raises an uncomfortable question: was it avoidable? Were these platforms always bound to evolve in that direction, toward capital accumulation, market domination, and growing inequalities?

What makes this question particularly striking is that alternative models did exist. Some of the early experiments in collaborative and shared-use models did not come from venture-backed startups, but from the field of social entrepreneurship. Yet while Airbnb scaled globally, Couchsurfing did not.

The difference cannot be explained by product quality or timing alone. It reflects deeper asymmetries in access to capital, infrastructure, and strategic positioning.

If there is a lesson to be drawn from the past decade, it is not only that certain actors won. It is that others were not in a position to compete.

A blind spot in the making

Today, much of the debate still focuses on digital technologies. But the next transformations are already underway elsewhere: in fields such as robotics, synthetic biology, advanced materials or nanotechnologies. They are emerging from laboratories. The digital battle might be lost, however, the technological war is not. And the next battles are happening in laboratories.

This is where a structural blind spot appears.

Social entrepreneurship has grown significantly over the past decades. In France alone, the social and solidarity economy represents 14% of employment [1]. Yet its presence remains concentrated in sectors such as social services, education, culture, or finance. Technology-intensive domains, especially those rooted in advanced scientific research, remain largely outside its scope. In practice, the “tech for good” sector still looks at digital applications. Scientific research, as a space where future technologies are shaped, remains a very distant territory.

This absence matters.

There is a tendency to think of scientific and technological development as a linear process: discoveries are made, technologies emerge, and society adopts them. In this view, the role of social entrepreneurs is to use existing technology to achieve their social impact or to mitigate the effect of a bad technology.

This view is misleading. Technological trajectories are not autonomous. They are shaped by funding priorities, industrial strategies, regulatory environments, and institutional frameworks. Technology development largely depends on who is involved early on, which problems are considered worth solving, and which applications are pursued.

If certain actors are absent from these early stages, their ability to influence outcomes later on is limited. This is precisely what happened in the digital sector. And there is little reason to believe the pattern will be different in other technological domains. If social entrepreneurship aims to transform economic models and address systemic challenges, then leaving the production of technology largely untouched creates a significant limitation.

Where social entrepreneurship needs to be

Social entrepreneurship was not designed to adapt to existing systems. It was designed to change them.Yet when it comes to science and technology, it largely operates downstream: using existing tools, mitigating their effects, or reacting to their consequences.

Given the challenges posed by climate change and rising social inequalities, it is imperative that the places where the future is created, i.e. research laboratories, are more infused with the ways of social entrepreneurs.

If the time has come to rebuild our models of society and our economy according to new values in order to face societal challenges, as the actors of social entrepreneurship are striving to do, then scientific research and technological developments are a lever that must be seized upon.

If the places where future technologies are shaped remain outside its scope, then its ability to influence societal trajectories remains partial. Engaging with research is therefore not an option. It is a condition for relevance.

Again, let us be clear here: we are not talking about raising awareness, opposing some technological developments or educating scientists on specific social justice topics. We are talking about actually launching research partnerships to valorize research results, the same way companies and industrials are.

The question is therefore not whether social entrepreneurship should engage with science, but whether it can afford not to. If it does not, the risk is not simply to miss opportunities. It is to reproduce the same asymmetries that have already been observed: where a small number of actors define technological trajectories, and others are left to react to their consequences.

Early signs and open questions

This does not mean that no alternatives exist. Some organizations have started to build bridges between scientific research and social impact, experimenting with new forms of collaboration. I have seen these configurations emerge first-hand, sometimes through initiatives such as SoScience. Social enterprises that actually collaborate with research partners such as Leka, Alg&You, Nutriset, Sublime Energie and more, are becoming more and more common.

It is these partnerships between economic actors who wish to remain faithful to the values of social entrepreneurship and the world of advanced research that we need more of. The question of tomorrow's proteins for a sustainable food supply or robotics for social inclusion are only examples of the many scientific questions that we need to explore. Be it in nutrition, robotics, energy, agriculture, electronics : every field has its examples.

These initiatives remain limited in number, but they show that other configurations are possible. The question is whether they can scale.

And more importantly: whether the research and innovation system, as it is currently structured, allows them to.

This is the first part of a two-part series on social entrepreneurship and scientific research. The second article explores the structural barriers that limit these collaborations.

[1] according to a study conducted by ESS France (The French Chamber of Social and Solidarity Economy) and the National Observatory of the SSE.

Contemporary discourse on innovation is grounded in a set of promises that have remained remarkably consistent over time. Philippe Bihouix identifies three main ones: a promise of abundance, a promise of liberation from labor, and a promise of enhanced human capabilities.

These promises are not new. Traces of them can already be found in the 16th century in the writings of Francis Bacon, for whom technology was meant to correct the world’s flaws and restore control over it. The idea that innovation can solve the problems it encounters did not originate in the digital age.

What has changed, however, are the technical systems within which these promises are embedded. Far from being disruptive, new technologies are added to existing infrastructures, which they complicate and on which they depend. Energy networks, supply chains, transportation systems: every innovation builds upon pre-existing layers and contributes to reinforcing dependence on them.

In this context, technological “solutions” often appear less as transformative changes that will provide answers and more as extensions of the status quo. They allow us to maintain a certain way of life by shifting constraints rather than resolving them.

Innovating So That Nothing Changes

Another shift has taken place over the past few decades: Etienne Klein points out that the word “progress” has gradually disappeared from political discourse, replaced by “innovation.” This shift is not insignificant.

Progress implies a direction, a desirable transformation of the world. Innovation, on the other hand, has become an end in itself. It is supposed to generate growth, jobs, and competitiveness. Within this logic, the question of the values that innovation embodies is never raised.

There is no questioning of society through innovation. Innovation consists of doing things differently so that we can continue doing the same thing. Producing, consuming, and trading as before, without questioning the purposes of these activities.

From this perspective, innovation does not serve to transform the model, but to stabilize it. It allows us to absorb tensions, circumvent limitations, and delay any questioning of the status quo.

In other words: we innovate so that nothing changes.

Material Limits That Are Hard to Overcome

This gap between promises and reality comes up against a constraint that discussions of innovation tend to downplay: the finite nature of resources.

Contemporary technologies rely on materials whose extraction is costly, energy-intensive, and often destructive. Recycling processes, frequently presented as a solution, face well-documented physical limits. Material losses are inevitable; certain uses irreversibly disperse resources; complex alloys make recovery extremely difficult.

In this context, the notion of “green growth” appears as an attempt to reconcile objectives that are difficult to reconcile. Producing more while reducing environmental impact requires efficiency gains that, in practice, are often offset by increased volumes.

Examples abound. Technologies presented as sustainable (wind power, for example) rely on supply chains that are intensive in rare resources. Objects designed to optimize comfort or performance introduce a complexity that makes them more fragile, harder to repair, and more dependent on extensive technical systems.

How, for example, can we talk about “clean cars” when we’re focusing entirely on the power and size of vehicles, and when the paints, electronics, and interior materials used rely on metals and plastics that will never be reused? This paradox can be explained by the fact that green growth today is based solely on climate criteria, overlooking the finite nature of resources, the preservation of biodiversity, and the quality of soil and water...

What Our Technological Choices Reveal

In light of these observations, one question takes center stage: what criteria guide our innovation?

Today, the answer is relatively clear. Technological choices are largely shaped by goals of productivity, competitiveness, and growth. It is these criteria that determine what is developed, funded, and disseminated. Other dimensions—social, environmental, and political—do exist, but they often play a marginal role, functioning as adjustments or constraints rather than as guiding principles.

This imbalance has direct consequences. It leads to a preference for technologies that reinforce certain dynamics (acceleration, increasing complexity, dependence) at the expense of other possible forms of organization (simplicity, resilience, local autonomy).

The so-called “low-tech” proposals, championed notably by Philippe Bihouix, are part of this tension. They do not consist of rejecting technology, but of redefining its uses and priorities: favoring solutions that are simpler, more robust, more repairable, and less dependent on scarce resources.

But this approach immediately raises another question.

Political Tensions That Are Hard to Avoid

As Olivier Rey points out, technological choices are not merely technical or economic decisions. They are also tied to issues of power.

Major technological advances have historically been closely linked to the military, industrial, and political capabilities of states. In a globalized world, where power dynamics remain a defining factor, deviating from dominant trajectories is not without risk.

Can a country or society choose to prioritize technologies that are less resource-intensive, less focused on productivity, and more oriented toward well-being, without putting itself in a vulnerable position vis-à-vis other actors? Can we slow down, simplify, and relocate without losing our capacity for action or influence?

There are no simple answers to these questions, but they shift the debate. The initial question—what technological choices for a sustainable society—cannot be addressed solely in technical terms. It requires examining the ends these choices serve, the criteria that guide them, and the power dynamics within which they operate.

What today’s debates and deadlocks reveal is that technologies are not chosen in the abstract. They are extensions of models, priorities, and worldviews. Under these circumstances, the question is not merely which technologies to adopt, but what we are truly seeking to preserve or transform through them.

The ideas in this essay are drawn from presentations at the conference organized by the Association des Centraliens, “What Technological Choices for a Sustainable Society,” held on April 13, 2016.