Deep learning is currently moving from policy into practice. At this\nstage, the important question is not only whether the concept is sound,\nbut also how it spreads and is adopted by implementers. This is where\nEverett M. Rogers\u2019 diffusion of innovations theory becomes relevant.\nRogers explains that an innovation is never accepted by everyone at the\nsame time. Adoption speeds differ because of perceptions of benefit,\ncompatibility, complexity, the possibility of trial, and the\nobservability of the innovation\u2019s results (Rogers, 2003).<\/p>\n
Rogers divides adopters into five groups. Innovators, around 2.5 per\ncent, are the group quickest to try new things. Early adopters, at 13.5\nper cent, follow more selectively and often serve as social references\nfor other groups. The early majority, 34 per cent, only move once they\nhave seen evidence that the innovation makes sense and works. The late\nmajority, also 34 per cent, tend to be more sceptical and wait until the\ninnovation becomes increasingly common practice. Meanwhile, laggards, at\nroughly 16 per cent, are the slowest group; some only change when\nenvironmental pressure is strong, and some may even cling to old\npractices (Rogers, 2003).<\/p>\n
These proportions are certainly not the result of measuring deep\nlearning implementers in Indonesia. They are more appropriately used as\nan analytical lens to understand that policy adoption speeds can never\nbe uniform. These categories are also not permanent identities or labels\nattached to individuals; they primarily describe a relative position\nwithin the adoption timeline of an innovation.<\/p>\n
A DIFFERENT APPROACH<\/p>\n
Deep learning deserves to be read as an innovation because it demands\na transformation in the way of learning: mindful, meaningful and joyful\nlearning; learning experiences that involve understanding, applying and\nreflecting; and support through pedagogical practice, learning\nenvironments, partnerships and digital utilisation (Kemendikdasmen,\n2025). With demands for change that large, it is impossible for all\nimplementers to move at the same speed.<\/p>\n
Rogers also reminds us that an innovation is adopted more quickly if\nit has relative advantage, meaning it is perceived as better than\nprevious practice; compatibility, meaning it fits the adopters\u2019 values,\nexperiences and needs; low complexity; trialability, meaning it can be\ntried on a limited basis; and observability, meaning its results can be\nseen. The term divisibility, often used to describe innovations that can\nbe applied bit by bit, is essentially close to the notion of\ntrialability. The easier it is to try, adapt and demonstrate the\nbenefits of deep learning, the faster its adoption will be.<\/p>\n
The implications for policy are highly strategic. Innovators should\nnot be burdened with excessive procedures. They need room to experiment,\nbuild prototypes and produce examples of practice. Early adopters need\naccess to evidence, networks and spaces for sharing, because they are\nthe prospective opinion leaders who can give deep learning social\nlegitimacy. The early majority cannot be won over with slogans alone;\nthey need tested models, clear guidelines, implementation examples,\nmentoring, and evidence that deep learning can improve the quality of\nlearning.<\/p>\n
The late majority requires yet another approach. This group usually\nonly moves once uncertainty has diminished. They need simple practices,\nlow implementation risk, close technical support, and evidence from\nenvironments they trust. As for the slowest group, they should not be\nimmediately branded as opponents of change. They may face resource\nconstraints, bad experiences with previous changes, or simply have not\nyet seen the relative advantage of deep learning. A policy that merely\nissues commands can actually amplify resistance.<\/p>\n
DIFFERENTIATED POLICY<\/p>\n
For this reason, the implementation of deep learning should not adopt\na \u2018one size fits all\u2019 philosophy. One size fits all may be\nadministratively practical, but it is weak as a theory of change. Groups\nwith different levels of readiness require different interventions. Some\nsimply need room to act, some need to be convinced with evidence, some\nrequire intensive mentoring, and some must first be helped to reduce the\ncomplexity and risk of change.<\/p>\n
The ministry can build a differentiated policy by mapping\nimplementers\u2019 readiness, rather than merely administrative compliance.\nSchools or regions that are already innovators and early adopters can\nserve as practice laboratories and demonstration centres. The early\nmajority can be strengthened through the replication of proven\npractices. The late majority needs more systematic mentoring. Meanwhile,\nthe slowest group needs a diagnosis of the underlying causes: whether\nthe problem lies in perceived benefit, compatibility, complexity,\nresources, or trust.<\/p>\n
Deep learning will fail as a transformation if its success is\nmeasured only by how quickly the term becomes known. The diffusion of\ninnovation demands more than socialisation; it demands changes in\nbelief, practice and experience. The ministry\u2019s challenge is not to make\nall implementers move in unison, but to help each group move forward\nfrom its starting position. If policy can read these variations in\nreadiness, deep learning will not remain a central policy, but will\ngenuinely diffuse into a culture of learning.<\/p>", "url": "https:\/\/jawawa.id\/newsitem\/deep-learning-which-groups-are-slow-reluctant-and-refusing-1789512620", "image": "" }, "sponsor": "Okusi Associates", "sponsor_url": "https:\/\/okusiassociates.com" }