Análisis diarios sobre tecnología, inteligencia artificial y educación, elaborados con fuentes académicas verificadas y publicados en español, inglés y francés.
A critical narrative review (24 studies, 2021–2026) of school phone bans: the most consistent benefits appear in attention, classroom climate and reduced bullying; effects on achievement and mental health are small, heterogeneous or null, and depend on design, compliance, exemptions and the use of the time regained.
Leer artículo completoA critical narrative review (24 studies, 2021–2026) of high-dosage tutoring as a post-pandemic learning-recovery policy: meta-analyses and randomised trials show positive effects, but when evidence is aligned with programmes at scale they shrink to roughly half; what matters is the dosage actually delivered, small groups, tutor continuity and integration into the school day.
Leer artículo completoA critical narrative review (24 studies, 2021–2026) of teacher burnout and wellbeing through the job demands–resources model: burnout is associated with poorer classroom interactions, lower student motivation and intentions to leave; mindfulness and compassion help modestly, but sustainable change requires redesigning workload, leadership, collaboration and collective efficacy.
Leer artículo completoA critical narrative review (24 studies, 2021–2026) of retrieval practice and the testing effect in real classrooms: more tests do not equal learning; the benefit depends on the practice regime (spacing, interleaving, successive relearning, feedback), the type of knowledge, the learner profile and teacher and student adoption.
Leer artículo completoA critical narrative review (24 studies, 2021–2026) of dialogic reading and shared book reading in early childhood education: sharing the book is not enough; vocabulary and narrative gains depend on extra-textual talk, home–classroom continuity and intervention fidelity.
Leer artículo completoA critical narrative review (24 studies, 2022–2026) of formative feedback and feedback dialogue in higher education: delivering comments is not enough; uptake depends on student and teacher feedback literacy and on dialogic cycles that enable interpretation and action.
Leer artículo completoA critical narrative review (24 studies, 2024–2026) of cognitive offloading and dependence on generative AI in secondary and higher education: once the tool is withdrawn, unrestricted access can leave less competence than practice without AI, while guardrails, delayed support, and compulsory reflection attenuate that cost. Assessing what remains without the tool is the decisive test.
Leer artículo completoA critical PRISMA-lite review (2023–2026, 24 studies) of AI-generated text detectors in higher education: accuracy improves under controlled conditions, but editing, hybrid texts, possible bias against non-native writers, and base rates prevent treating a score as proof. Integrity is defended by validating authorship, not by detecting it.
Leer artículo completoA critical PRISMA-lite review (2021–2026, 24 studies) of parental technoference in early childhood: experimental evidence shows immediate infant distress and reduced adult attention when the phone interrupts interaction; developmental evidence is correlational and bidirectional. What protects is repairing contingency, not counting minutes.
Leer artículo completoA PRISMA-lite review (2021–2026, 24 studies) of the preschool-to-primary transition: school readiness is a relational capacity shared by children, families, and teachers at both levels, not an entry score; gatekeeping screenings, predictive dashboards, GenAI reports, and “readiness” apps do not replace pedagogical continuity.
Leer artículo completoThis article analyzes vulnerabilities in artificial intelligence systems, including jailbreaks, prompt injections, and hallucinations, and their impact on security and trust in technology.
Leer artículo completoA math-talk-rate detector, a GenAI number-talk script, an utterances-per-hour dashboard, a chatbot without an adult, or a fidelity checklist as ranking do not constitute mathematical conversation: the craft is dialogic mediation in play and routines across home, classroom, and professional learning.
Leer artículo completoA “yard safety score” app, a GenAI “play risk assessment” package, an incident dashboard, or a “digital parental permission” checklist without a pedagogical criterion of calculated risk do not constitute support for risky play; the craft is the teacher mediation that enables situated physical and social challenges in preschool.
Leer artículo completoA STEM loose-parts catalogue app, GenAI provocation prompts, a creative-exploration-minutes dashboard, or a creativity-score checklist without teacher mediation do not constitute support for play with open materials; the craft is the mediation that sustains situated material exploration in preschool.
Leer artículo completoA role-play script app, GenAI digital-theatre prompts, a symbolic-play-minutes dashboard, or a character checklist without teacher mediation do not constitute support for symbolic play: observation, plot expansion, voice inclusion, and props in service of shared fiction are required.
Leer artículo completoAn automatic-translation app, a monolingual drill, a target-language-minutes dashboard, or a filter that silences the home language without a translanguaging pedagogy do not constitute plurilingual support: full repertoire, observation, family ties, and digital tools in service of shared meaning are required.
Leer artículo completoA learning-analytics dashboard, a checklist of documented moments, a photo-count score, or a GenAI caption without narrative observation, teacher interpretation, family dialogue, and evidence of process do not constitute pedagogical documentation: the craft is the learning story, not the count.
Leer artículo completoA brain-training score, a working-memory app, a screen-time dashboard, or a cognitive-exercises pack without situated play and teacher scaffolding do not constitute support for executive function: social context, observation, transfer, and not replacing play with drill are required.
Leer artículo completoA GenAI comment bank, an LLM rubric score, a similarity dashboard, or a “write better” chatbot without teacher mediation do not constitute formative assessment: goals, criteria, dialogue, revision, and judgment attribution are required.
Leer artículo completoPISA 2025 recorded the OECD's lowest reading average. This state-of-the-art review examines the role of devices and generative AI, proposes the 4D model, and sets out a roadmap for Mexico and Latin America.
Leer artículo completoA chat that “teaches”, a confidence score, a generated plan, or an “autonomous” agent without an HITL protocol (acceptance criteria, traces, veto, attribution, escalation) do not constitute professional support: the craft is teacher judgment.
Leer artículo completoA cherry-picked demo, a leaderboard screenshot, a single-run score, or a video of “the agent solved X” without a fixed evaluation harness do not constitute agent evaluation: a verifiable criterion outside the model is required.
Leer artículo completoA GenAI of circle/morning meeting scripts / lesson packs / carpet-time banks, an engagement/attention/participation quality score or circle compliance, a circle-minutes / morning-meeting-fidelity dashboard, a robot or agent that leads the meeting in place of the adult, or a vision/NLP participation-quality ranking without situated judgment do not constitute support for circle time in early childhood: it is the adult’s dialogic relational craft with the group.
Leer artículo completoA larger model, a “better” prompt, or a tools demo without an observe–act–verify loop do not constitute a harness: typed tools, memory, stop policies, verifiers, budgets, logging, and isolation turn generative capacity into reliable work; gains on agent benchmarks are explained more by harness design than by scaling the LLM alone.
Leer artículo completoA GenAI of cooperative play scripts / peer collaboration lesson packs / shared-goal activity banks, a cooperation quality score or cooperative minutes/hour, a cooperative-minutes dashboard, a “cooperative peer” robot without contingent adult mediation, or a vision/NLP collaborative-effectiveness ranking without situated judgment do not constitute support for cooperative play in early childhood: it is peer co-construction and the adult’s relational craft.
Leer artículo completoA GenAI of guided play scripts / lesson packs / provocation banks, a guided play quality score or minutes of guided play/hour, a guided-minutes dashboard, a robot that “guides play” without a contingent adult, or a vision/NLP scaffolding ranking without situated judgment do not constitute support for guided play in early childhood: it is contingent co-construction and the adult’s relational craft.
Leer artículo completoA facial/voice/ML detector of EQ score or dysregulation alerts, a GenAI of emotion scripts / calm-down recipes, a dysregulated-minutes-per-hour dashboard, a chatbot/robot/wearable that “co-regulates” without an adult, or a teacher emotional sensitivity ranking without situated judgment do not constitute support for co-regulation or emotional self-regulation in early childhood: it is presence, affective attunement, and relational craft.
Leer artículo completoAn ASR/NLP/ML detector or SSTEW auto-score, a GenAI of SST scripts / open-question banks, an SST-minutes-per-hour dashboard, a chatbot/robot that “does SST” without an adult, or ChatGPT co-coding turned into a teacher ranking do not constitute support for sustained shared thinking in early childhood: it is prolonged dialogue, co-construction of meaning, and situated pedagogical interpretation.
Leer artículo completoA CV/multimodal ToM or perspective-taking-score detector, a GenAI of false-belief worksheets/scripts, a trials-per-hour dashboard, a closed VR/robot treatment, or a gaze/emotion ranking without mediation do not constitute support for theory of mind / perspective-taking in early childhood: it is adult scaffolding, mental-state language, and situated role play.
Leer artículo completoA CV/multimodal detector of prosocial acts or kindness index, a GenAI of sharing worksheets, a helps-per-hour dashboard, a closed VR treatment, an automated-praise robot, or a proximity ranking without mediation do not constitute support for prosocial behaviors (helping/sharing/consoling) in early childhood: it is the other’s real need, offer/help/comfort language, and children’s agency.
Leer artículo completoA CV/multimodal conflict detector or SEC score, a GenAI of resolution/sharing scripts, a conflicts-per-hour dashboard, a “say sorry” robot, or a proximity/synchrony ranking without mediation do not constitute support for peer-conflict negotiation / turn-taking in early childhood: it is co-presence, request/wait/offer language, and children’s agency to repair.
Leer artículo completoA CNN/CV that scores strokes/shapes/DAP or fine-motor/prewriting/drawing-age/VMI scores, a GenAI of copy-the-shape worksheets, a drawing-minutes/stroke-accuracy dashboard, a correctness app, or a developmental ranking without mediation do not constitute support for mark-making / early drawing-writing in early childhood: it is real materials, graphic agency, and adult co-presence that interprets meaning and process.
Leer artículo completoA chatbot/LLM agent that “does dialogic reading” with the child, a comprehension/vocabulary/dialogic quality score, a GenAI of CROWD/PEER scripts, a shared-reading-minutes dashboard, or an app that replaces adult–child shared reading do not constitute support for dialogic reading in early childhood: it is a shared book with contingent PEER/CROWD and adult co-presence.
Leer artículo completoA “pouring accuracy” / “STEM water score” sensor, a CV/water-table camera with “mess / off-task splashing” alerts, a GenAI of “water experiments” / sink-or-float lesson plans, a minutes-at-the-water-table dashboard, or an AR layer that guides pouring or predicts sink/float do not constitute support for water play in early childhood: it is exploration of the liquid with flow, capacity, sink-float and adult co-presence.
Leer artículo completoA “sandcastle accuracy” / “STEM sand score” / “molding accuracy” sensor, a CV/sandbox camera with “mess / off-task sand spilling” alerts, a GenAI of “sandcastle plans” / sand play lesson plans, a minutes-in-the-sandbox dashboard, or an AR layer that guides molding or predicts castle stability do not constitute support for sand play in early childhood: it is exploration of the granular with flow, compaction, moldability and adult co-presence.
Leer artículo completoA “playdough molding accuracy” / “fine-motor dough score” / “shape compliance” sensor, a CV/table camera with “mess / off-task dough spilling” alerts, a GenAI of “figure plans” / playdough lesson plans / STEM dough activities, a minutes-with-playdough dashboard, or an AR layer that guides kneading or predicts the correct shape do not constitute support for playdough / clay play in early childhood: it is exploration of the deformable with tactile properties and adult co-presence.
Leer artículo completoA wearable or “fall-risk score” sensor, a CV/playground camera with “unsafe play” alerts, a GenAI of “safe activities” / risk-free alternatives, a “risk exposure score” dashboard, or an AR layer that avoids height/speed do not constitute support for risky play in early childhood: it is child calibration with Sandseter categories, risk≠hazard and adult co-presence.
Leer artículo completoA chatbot that “suggests constructions” or “loose parts ideas,” a GenAI generator of provocations by prompt, a CV system that scores “creativity,” an AR layer that guides arrangement, or an “open-endedness score” dashboard do not constitute support for loose parts play in early childhood: it is situated practice with open-ended physical materials, child agency, extended time, peers and adult co-presence.
Leer artículo completoAn “eco-tips” chatbot or “green classroom” plans, an app scoring “green footprint,” computer vision of “sustainable acts,” or a prompt-based environmental-unit generator do not constitute early childhood environmental education: it is situated care of place, sensory nature exploration, everyday habits and adult mediation (ECEfS / nature pedagogy).
Leer artículo completoA chatbot that “teaches nature,” an AR identification app, an outdoor-minutes/green-score wearable or dashboard, a prompt-based outdoor-unit generator or a VR forest do not constitute support for outdoor play in early childhood: it is situated practice with body outdoors, peers, real materials and adult co-presence.
Leer artículo completoA role-playing chatbot, a pretend-play script app, an autonomous voice agent directing roleplay, computer vision scoring “symbolic-play quality,” or a prompt-based dramatic-play unit generator do not constitute support for symbolic play in early childhood: it is situated peer practice with open-ended objects, metacommunication and adult scaffolding (teleoperated StoryCarnival = peripheral contrast).
Leer artículo completoA chatbot that “teaches spatial,” a puzzles app scoring “spatial ability,” computer vision of construction “complexity,” or a prompt-based spatial-training unit generator do not constitute support for spatial reasoning in early childhood: it is situated block play, spatial language, gestures and human scaffolding.
Leer artículo completoA STEM/5E chatbot, a sensor or computer vision scoring curiosity or engagement, or an experiment-script generator from a prompt do not constitute early scientific inquiry: in early childhood it is situated practice of observation, experimentation with real materials and adult mediation at the science table.
Leer artículo completoThis article analyzes how AI-powered virtual assistants are transforming the workplace, enhancing productivity and redefining human interactions.
Leer artículo completoThis article analyzes the impact of artificial intelligence in the financial sector, focusing on fraud detection, credit scoring, and the implementation of banking assistants.
Leer artículo completoThis article analyzes the fine-tuning process of artificial intelligence models, discussing when it is beneficial to train your own model and when it is better to use pre-trained models.
Leer artículo completoAn eye-tracking system that marks “correct/incorrect” if the child follows the gaze, a social robot with RJA/IJA prompts, or CV that scores joint visual attention do not constitute pedagogical joint attention: at ages 0–6 it is situated triadic practice, serve-and-return and mediation of shared focus.
Leer artículo completoA CLASS video scorer, a “respond now” wearable, or a conversational-turn classifier does not constitute pedagogical–relational interaction quality: early process quality is sensitivity, reciprocity and scaffolding (serve-and-return; CLASS ES/CO/IS; OECD Starting Strong).
Leer artículo completoAn app that completes the child’s drawing, an image generator from a prompt, or a creativity scorer does not constitute artistic creativity: in early childhood it is process art, material–bodily exploration and teacher mediation, not an output or a score.
Leer artículo completoA pose-estimation system that scores a jump or run as “correct/incorrect,” a motor-milestone wearable, or an FMS-from-video app does not constitute motor development: in early childhood it is situated practice of locomotion, object control and stability in mediated active play.
Leer artículo completoA PEER/CROWD dialogic-reading chatbot, a prompt-based story generator, or an app that marks letters or comprehension as “correct/incorrect” do not constitute emergent literacy: in early childhood it is situated practice of shared reading, print awareness and adult mediation in the book corner.
Leer artículo completoA pitch-correction app that marks singing as “correct/incorrect,” a lullaby generator from a prompt, or an affective-computing system that classifies musical engagement do not constitute music education: in early childhood it is situated practice of shared singing, rhythmic-body play and adult mediation.
Leer artículo completoA prompt course, a webinar, a badge or AI-generated microlearning does not constitute teacher professional development in a kindergarten: effective TPD in early childhood is inquiry situated in practice with peers and accompaniment, not consumption of automated content.
Leer artículo completoA cognitive training app, an adaptive game that scores attention or a model that classifies impulsivity do not develop executive functions in a 3–6 kindergarten: EFs are inhibitory control, working memory and cognitive flexibility built through shared play, conversation, mediated waiting, movement and adult scaffolding on real everyday conflicts; not digital drill or algorithmic attention scoring.
Leer artículo completoAn app that drills counting, a chatbot that poses problems or a model that personalizes worksheets do not constitute numeracy in a kindergarten: early numeracy is embodied number sense, comparison, composition–decomposition and mathematical conversation, not digital drill.
Leer artículo completoA programmable robot kit, a block app or a model that suggests STEM challenges do not constitute a STEM experience in a kindergarten: early STEM is bodily, material inquiry shared with adults, not interface training.
Leer artículo completoA platform that personalizes, an accessibility dashboard or a model that adapts content does not constitute UDL: UDL is design of barriers in advance with the adult as designer of the environment, not an algorithmic adjustment on a classified child.
Leer artículo completoA progress dashboard, a continuous score or a model that diagnoses the child does not constitute formative assessment in a kindergarten: formative assessment in early childhood is the adult’s situated judgement in the process—feedback, activity adjustment, a view of the group—not a board that classifies.
Leer artículo completoA generated story, a chatbot that “converses,” or a speech-correction app does not constitute oral-language mediation: at ages 3–6, speech is situated dialogue, listening and expansion by the adult, not a model output.
Leer artículo completoA “feedback” button, an avatar that “consults” the child or a model that summarises their opinion does not constitute child participation: in early childhood education, participation is agency, situated listening and shared decision-making with responsible adults, not an input captured by an interface.
Leer artículo completoA model-generated lesson plan, a ready-made sequence or a prompt that produces circle time does not constitute pedagogical planning: in early childhood education, planning is group observation, situated decision and in-the-act adjustment, not a textual artefact generated a priori.
Leer artículo completoAn analysis of the risks associated with artificial intelligence systems, including jailbreaks, prompt injections, and hallucinations.
Leer artículo completoA notice chatbot or an automatic digest of the day does not constitute pedagogical family–school communication: that communication is situated dialogue among adults responsible for the child, not a generated feed.
Leer artículo completoAn emotions chatbot, an affect detector or an app that names feelings does not constitute SEL in kindergarten: SEL at ages 3–6 is situated interaction, co-regulation and play, not face classification or dialogue with a model.
Leer artículo completoA gamified app, a robot that directs turns, or a model that suggests playful activities does not constitute free play: play at ages 3–6 is the child’s initiative, open temporality and the absence of an imposed product, not an AI-optimised sequence.
Leer artículo completoA school-readiness score or a model that predicts preparedness for primary school does not constitute a pedagogical transition: the passage at ages 3–6 is a relational process of children, families and teachers, not an algorithmic threshold.
Leer artículo completoBuying a licence, naming an AI lead or circulating an internal memo does not constitute pedagogical leadership in a kindergarten: leadership is played out in protecting play time, educators’ judgement and children’s rights.
Leer artículo completoThis article analyzes the integration of text, image, and audio in artificial intelligence models, highlighting its potential in education and business.
Leer artículo completoAn AI module does not prepare the kindergarten teacher if it only teaches prompts: intention to use and self-reported AI-TPACK are not didactic judgement in initial teacher education.
Leer artículo completoThis article analyzes how artificial intelligence is transforming the labor market by automating certain tasks and enhancing others, as well as its impact on the future of employment in Latin America.
Leer artículo completoThis article explores the evolution of computer vision from image classification to its application in real-time detection, highlighting its impact on education and business.
Leer artículo completoAI can transcribe and tag the early childhood classroom; it does not document in the pedagogical sense without situated adult interpretation. The risk is not only extracting data: it is that what the adult no longer looks at ceases to count as documentation.
Leer artículo completoThere is not sufficient evidence to claim that AI reduces burnout or “returns time to the classroom” in early childhood and basic education: it can reorganize teachers’ work and, at times, intensify it.
Leer artículo completoWithout connectivity, energy and technical support, educational AI does not arrive late in rural kindergartens and basic schools: it arrives as inequality. Innovation that presupposes broadband, a device and the cloud is realized where a network already exists.
Leer artículo completoArtificial intelligence is not linguistically neutral. A system trained mainly in English can widen language inequality in early childhood and basic education: Spanish scores below English, Indigenous languages operate with high-error ASR, and AI literacy that only works in English is not literacy.
Leer artículo completoThe arrival of generative text models has shifted academic integrity toward a problem of evidence: if a system can produce the artifact the school grades, what counts as proof that a child learned? In basic education the problem is not hunting cheats with a detector, but not evaluating a text without a subject.
Leer artículo completoEducational artificial intelligence is often announced as classroom personalization. This article advances a restrictive thesis: in early childhood and basic education that personalization does not operate without cost, because it extracts voice, image, and behavior from girls and boys who cannot give informed, free, and reversible consent.
Leer artículo completoAI literacy in basic and early childhood education is not using ChatGPT in class: it is a curriculum of ideas (data, pattern, prediction, limit, agency) compatible with the age-13 threshold for independent generative conversation.
Leer artículo completoAI does not include by itself in early childhood: it can expand communicative access as assistive technology and, at the same time, medicalize, bias voices and languages, or displace the specialist and the support teacher.
Leer artículo completoUNESCO’s AI competency framework for teachers does not transfer to kindergarten as a tools workshop: it must be translated into play, observation, documentation, care, and inclusion, and anchored in empirical evidence without promising that AI improves development.
Leer artículo completoEducating children in AI is not giving them ChatGPT: it is parental mediation of coviewing, talk, limits, privacy, and literacy. The home is the first scene of mind attribution to machines; the evidence does not show that AI develops the child.
Leer artículo completoA doctoral synthesis of how play and oral language can be scaffolded when generative artificial intelligence enters early childhood education. Drawing on seventeen studies from 1995–2021, PopBots (n=80, ages 4–7), AI robotic toys in the 2020 Australian lockdown (n=5) and child-agent interaction (n=26, ages 3–10), it argues that algorithmic mediation is defensible only when subordinated to the adult, to relational pedagogy and to the age limit on independent conversations with generative platforms.
Leer artículo completoA deep analysis of the internal workings of large language models and their impact on education and business.
Leer artículo completoWe explore the impact of artificial intelligence in education through personalized tutors and adaptive assessment.
Leer artículo completoThis article analyzes the impact of AI chatbots in customer service, focusing on conversational design and effectiveness metrics.
Leer artículo completoThis article analyzes the structure and function of artificial intelligence agents, emphasizing their architecture, memory capacity, and levels of autonomy.
Leer artículo completoEfficient use of tokens in artificial intelligence can reduce costs and enhance the effectiveness of prompts.
Leer artículo completoAn analysis of the evolution of artificial intelligence interfaces, from simple prompts to autonomous agents that perform complex tasks in various environments.
Leer artículo completoThis article explores the impact of artificial intelligence on cybersecurity, analyzing its applications in automated defense and attacks.
Leer artículo completoThis article reviews the scientific state of the art of voice agents in telephony, with verified sources published between 2021 and 2024, and their future in communication.
Leer artículo completoThe incorporation of artificial intelligence (AI) into classrooms has opened a new family of educational inequalities that transcends the classic digital divide. This article presents a state-of-the-art review (2021-2026) of educational gaps in AI learning,…
Leer artículo completo