When you talk about the future of artificial intelligence, it seems that we will all receive the same technology at the same time and under the same conditions. It does not happen that way. A tool that for someone assumes comfort can be a barrier to another person; what frees up time in one team can increase pressure on another.
The technological future is not a place we come to together. It is built with present decisions: who has access, who receives training, who can reject a tool and who must accept its results without understanding them.
Access is more than just connection.
Having a device does not guarantee that you can take advantage of a tool. It also takes time, confidence, knowledge and an environment to ask without feeling out of place. When these conditions are ignored, the digital divide is disguised as lack of interest or capacity.
Accessible technology allows us to understand what it does and offers alternatives when it doesn't fit. It does not require learning specialized language to solve a daily need. It does not penalize those who need more time, another language, or a conversation with a person.
In education, this difference is crucial. Two students can use the same platform and start from very different situations: one has accompaniment and quiet space; another shares device or works without support. Evaluating only the digital result can turn this context inequality into a label about their effort.
Benefits and costs are unevenly shared
An organization can save time with automatic responses while users put more effort into explaining cases that don’t fit. A system can improve a global figure and make the experience of a small group worse. If we only measure the average, that group disappears.
Asking who bears the cost changes the conversation about innovation. It forces you to look beyond the demonstration and follow the full path: who prepares the data, who checks the errors, who expects and who can claim.
It also matters the ability to choose. For one person, using an assistant can be voluntary; for another, artificial intelligence may be integrated into a process, selection or evaluation that cannot be avoided. The less the choice, the greater the transparency and human review.
Design futures in plural
A good test incorporates people with different experiences from the beginning. It is not only asking if the tool works, but whether it is understood, whether it allows correction and if it leaves someone out. This listening must be able to change the design, not become a decorative consultation.
Inclusion does not consist of adapting an already closed solution. It consists in recognizing from the beginning that there is no universal user. There are different bodies, languages, abilities, family economies and ways of learning.
The future will be shared only if people who live through its consequences can participate in the decisions that build it.
Talking about futures in plural does not reduce technological ambition. It makes it more concrete. It allows you to decide where a tool expands opportunities and where it needs limits, support or an alternative.
Artificial intelligence can help bring resources closer together and personalize support. That promise is only fulfilled if we do not confuse availability with access and speed with justice. Progress begins when we stop asking when the future will come and start deciding for whom we are building it.
This difference also affects communities and territories. A solution designed for a large city can depend on connections, services and habits that do not exist in a rural environment. Presenting it as universal forces people to adapt to a foreign reality and makes centralized design a measure of normality.
Languages are another example. A tool can deliver excellent performance in the dominant language and respond with less precision in other languages or local variants. If that difference is not communicated, the one who receives a worse service can interpret the failure as his own.
Preparing the future requires investing not only in tools, but in shared capacities: training for teachers and families, accompanied access spaces, understandable materials and equipment capable of evaluating suppliers. Without this social infrastructure, innovation benefits especially those who were already prepared.
We also need to preserve non-digital alternatives for essential needs, not as a rejection of change, but as a guarantee of access when a person does not have the necessary device, competence or confidence. A real alternative prevents modernization from becoming a condition for exercising a right.
Early participation improves decisions. Those who know a territory or a collective can anticipate problems that do not appear in a laboratory test. Incorporating their experience reduces subsequent corrections and creates more legitimate solutions.
The plural future requires accepting different rhythms. To move forward does not mean to force everyone to use the same tool on the same day. It means to ensure that each addition expands possibilities without closing the ones that other people still need.
Public and educational procurement can drive this approach by demanding accessibility, portability, clear explanations and independent evaluation. Choosing a provider is not just comparing functions; it is deciding what conditions people will have for years.
If these conditions are agreed before, technology can evolve without leaving anyone trapped. The future then ceases to be an imposition and becomes a construction that admits corrections.
It is also important to periodically assess who is using the tool and who has stopped doing so. Abandonment may reveal barriers that usage statistics do not explain. Asking for such absences allows correcting before a supposedly universal solution ends up caring only for those who best fit it.
