Intelligence artificielle : quels sont les apports concrets de la régulation européenne ?

Source: The Conversation – France in French (3) – By Chloé Dornbierer, Docteure en droit et Professeure à Kedge Business School, Kedge Business School

L’intelligence artificielle s’impose désormais au cœur de décisions qui engagent directement la vie, la santé et les droits des citoyens européens. Dans des entreprises, des administrations, des hôpitaux ou sur des plateformes numériques, des algorithmes trient, évaluent, recommandent et parfois décident. Face à cette automatisation croissante, une question centrale se pose : la réponse de l’Union européenne pour encadrer l’intelligence artificielle est-elle à la hauteur des risques qu’elle prétend prévenir ?


Imaginons qu’une entreprise décide d’utiliser une intelligence artificielle (IA) pour automatiser son processus de recrutement. L’algorithme, conçu pour analyser des milliers de CV, est entraîné sur les données RH de l’entreprise des dix dernières années. Or, les profils recrutés durant cette période sont majoritairement masculins. L’algorithme déduit alors de ces données que les candidats masculins sont préférables et exclut de fait les profils féminins.

Ce scénario n’est pas fictif : il s’inspire directement du cas d’Amazon, dont l’outil de recrutement par IA a dû être abandonné après la révélation de ces biais eu égard aux risques juridiques qu’ils engendraient. Cet exemple illustre précisément les risques que le législateur entend prévenir : erreurs algorithmiques, atteintes aux droits fondamentaux, captation et fuite de données, dépendance technologique.

En réponse, l’Union européenne a bâti depuis plusieurs années un arsenal juridique ambitieux, allant du Règlement général sur la protection des données (RGPD) au Règlement sur l’intelligence artificielle (RIA). Ces textes doivent relever un défi de taille : concilier innovation et protection, réguler les acteurs internationaux sans les éloigner du marché européen, et garantir les droits fondamentaux sans censure – le tout face à des technologies transversales, opaques et en constante évolution. Cet ensemble de règles constitue-t-il une réponse satisfaisante ?

Pas de régulation de l’IA, sans régulation des données

La régulation de l’intelligence artificielle ne saurait être envisagée indépendamment d’une politique structurée de la donnée. La donnée constitue en effet le carburant indispensable au développement et au perfectionnement des systèmes d’intelligence artificielle. En cela, la quantité de données compte tout autant que leur qualité.

C’est pourquoi, le législateur européen a d’abord cherché à réguler la donnée avant de réguler l’IA : d’une part avec le RGPD dont la finalité est de protéger les données à caractère personnel (les informations concernant une personne physique identifiée ou identifiable, comme un numéro de téléphone ou une adresse postale).

D’autre part avec une série de textes portant sur la régulation des données non personnelles (les informations ne se rapportant à aucune personne identifiable, comme une statistique ou un nombre d’accidents sur une autoroute visée). Ceux-ci comprennent le DSA, le DMA, le DGA, le Data Act et visent notamment à limiter la domination des géants du numérique et organiser le partage et l’accès de certaines données entre plateformes, entreprises, administrations publiques et utilisateurs, condition indispensable au développement de l’IA.

Ces textes relatifs aux données ont été le préalable à l’élaboration de la première réglementation générale au monde sur l’intelligence artificielle : le RIA.

Quelles sont les obligations imposées aux entreprises par le RIA ?

Contrairement aux États-Unis, qui privilégient une approche dérégulée, et à la Chine, qui encadre principalement certains usages ciblés, l’Union européenne a fait le choix d’un cadre juridique général et contraignant, s’appliquant à l’ensemble des secteurs et des systèmes d’intelligence artificielle, afin de poser les bases d’un véritable droit commun de l’IA au sein du marché européen depuis son entrée en vigueur le 1er août 2024.

Son objectif est d’encadrer

« le développement, la mise sur le marché et l’utilisation de systèmes d’intelligence artificielle, qui peuvent poser des risques pour la santé, la sécurité ou les droits fondamentaux ».

Pour atteindre cet objectif, le RIA commence par définir précisément le champ des technologies qu’il entend encadrer. Il distingue : les systèmes d’intelligence artificielle, qui correspondent aux applications concrètes utilisées par les utilisateurs (ex : un outil de tri de candidatures, ChatGPT) et les modèles d’intelligence artificielle à usage général, qui constituent le socle technologique sur lequel reposent ces applications, comme GPT-4 pour ChatGPT.

Pour assurer un équilibre entre protection des personnes et innovation, le RIA classe les systèmes d’intelligence artificielle en quatre niveaux de risque :

  • le risque inacceptable (ex : système de notation sociale des citoyens, présentant un risque d’une atteinte grave aux droits fondamentaux) impliquant l’interdiction du système d’IA.

  • le risque élevé (ex : une IA utilisée pour évaluer des salariés, présentant un risque d’atteinte au droit du travail et à la vie privée) imposant des obligations comme celle de réaliser des audits ou celle d’entraîner le système sur des données qualitatives et exemptes de biais.

  • le risque limité (ex : assistant conversationnel d’un service client, présentant le risque d’une collecte de données sans consentement éclairé) obligeant notamment les opérateurs à informer les utilisateurs de l’usage de l’IA.

  • le risque minimal (ex : IA intégrée dans un jeu vidéo) dont ne découle aucune obligation spécifique.

Pour chaque niveau de risque, des obligations spécifiques et proportionnées sont donc édictées en fonction de leurs effets potentiels sur les droits fondamentaux, la sécurité et la santé des personnes. Ces obligations s’imposent à différents opérateurs selon le moment où ils interviennent dans la chaîne de valeur de l’IA : fournisseur, mandataire, importateur, distributeur…

Par exemple, un système d’IA utilisé pour l’analyse d’images médicales, qualifié de système à risque élevé, est soumis à des obligations spécifiques qui varient selon l’opérateur concerné : le fournisseur doit concevoir le système en conformité avec les exigences de sécurité et de gestion des risques comme s’assurer qu’il ne peut pas y avoir de faux positif ou de faux négatif dans le diagnostic mais également que le système ne présente pas de biais du fait d’une sous-représentation de certains groupes ethniques ou de genre, l’importateur et le distributeur doivent vérifier que cette conformité a été respectée, et l’hôpital qui déploie le système doit s’assurer qu’il est utilisé correctement et sous supervision humaine.

Concernant les modèles d’IA à usage général (ex : GPT-4, algorithme sur lequel repose ChatGPT), le législateur européen a édicté différentes règles selon qu’ils présentent ou non un risque systémique.

L’article 3 § 65 du RIA définissant le risque systémique comme :

« un risque spécifique aux capacités à fort impact des modèles d’IA à usage général, ayant une incidence significative sur le marché de l’Union en raison de leur portée ou d’effets négatifs réels ou raisonnablement prévisibles sur la santé publique, la sûreté, la sécurité publique, les droits fondamentaux ou la société dans son ensemble, pouvant être propagé à grande échelle tout au long de la chaîne de valeur ».

Ce risque est évalué selon la puissance de calcul du modèle. Les modèles d’IA à usage général qui présentent un risque systémique doivent notamment signaler les incidents graves (des biais, des faux positifs, une cyberattaque), prendre d’importantes mesures de cybersécurité ou encore tester le modèle et atténuer les risques systémiques identifiés (grâce à des audits, des mesures de protection contre les piratages ou encore des mesures de détection de biais). Plusieurs autorités ont été désignées pour contrôler la bonne application du règlement, c’est le cas notamment du comité européen de l’intelligence artificielle au niveau européen, et de la CNIL au niveau national.

Les limites du RIA

Le RIA a suscité des critiques des experts du droit et du monde professionnel à propos de sa mise en œuvre. Ainsi, 46 dirigeants d’entreprise ont dénoncé, dans une lettre ouverte destinée à la présidente de la Commission européenne, « des règlementations européennes peu claires, qui se chevauchent et deviennent de plus en plus complexes ». La technicité du RIA conjuguée à la complexité des algorithmes compliquent l’application du texte, en particulier pour les obligations de transparence – incompatibles avec l’opacité des algorithmes – et pour l’identification des responsabilités, les entreprises peinant à déterminer quelles obligations leur incombent selon leur position dans la chaîne de valeur et le niveau de risque du système.

En découle alors une autre problématique, le risque de freiner les acteurs français et européens, comme Orasio, H Company ou DeepL, qui tentent de concurrencer les géants américains de l’intelligence artificielle qui ont davantage de moyens (juridiques, technologiques, financiers) pour mettre en œuvre les obligations du RIA.

À cet égard, il convient de relever que le législateur européen a finalement renoncé, le 11 février 2025, à adopter une législation spécifique relative à la responsabilité civile des acteurs de l’intelligence artificielle, dont la finalité était d’harmoniser les règles de responsabilité civile pour les dommages causés par l’IA, afin de protéger les victimes et d’unifier le cadre juridique au sein de l’UE. Cet abandon est intervenu suite à une opposition de certains États membres, notamment la France, mais aussi des critiques formulées par le lobbying de grandes entreprises technologiques qui dénonçaient un risque de surrégulation et une insécurité juridique susceptible de freiner l’innovation. En l’absence d’un tel cadre spécifique, ce sont donc les règles de droit positif de responsabilité civile qui demeurent applicables aux dommages causés par les systèmes d’intelligence artificielle.

Les enjeux de la régulation

En combinant protection des données, encadrement des usages et obligations pesant sur les acteurs les plus puissants, l’Union européenne a posé les bases d’un véritable droit commun de l’IA unique au monde.

Cependant, cette réponse n’est satisfaisante qu’à une condition : que son application ne transforme pas la régulation en barrière à l’entrée pour les acteurs européens tels que Mistral AI, NEURA Robotics ou Dust. En effet, le risque est de renforcer la domination des géants américains et chinois plutôt que de la combattre. D’autant que cette domination ne se limite pas au plan économique : elle s’exerce également sur le plan politique. Un rapport de Corporate Europe Observatory et LobbyControl, deux ONG spécialisées dans la surveillance du lobbying, montre ainsi que les Big Tech auraient une influence significative sur la rédaction des Codes de bonnes pratiques prévus par le RIA notamment en parvenant à affaiblir certaines obligations.

La véritable réussite de la régulation européenne ne se mesurera donc pas seulement à la protection des citoyens, mais à sa capacité à faire émerger une intelligence artificielle européenne, à la fois éthique, compétitive et souveraine.

The Conversation

Chloé Dornbierer ne travaille pas, ne conseille pas, ne possède pas de parts, ne reçoit pas de fonds d’une organisation qui pourrait tirer profit de cet article, et n’a déclaré aucune autre affiliation que son organisme de recherche.

ref. Intelligence artificielle : quels sont les apports concrets de la régulation européenne ? – https://theconversation.com/intelligence-artificielle-quels-sont-les-apports-concrets-de-la-regulation-europeenne-274299

Trois récits de vie pour comprendre le mal-être dans la fonction publique

Source: The Conversation – France in French (3) – By Aude Deville, Professeur des Universités, Université Côte d’Azur

« J’aimais mon métier d’enseignante, mais là, je n’avais plus le sens. Mon médecin m’a dit, “C’est un burn out” », témoigne une enseignante interviewée. FrameStockFootage/Shutterstock

Face à l’austérité budgétaire, certains agents de la fonction publique font face à des injonctions contradictoires – moins de moyens, mais une meilleure qualité de service –, des responsabilités floues et un manque de reconnaissance qui affectent leur bien-être. Pour comprendre ces enjeux complexes, une étude donne la parole à trois agents de la fonction publique d’État.


Une directrice d’école en burn out faute d’aide. Un cadre hospitalier en charge des décisions, sans pouvoir réel. Un fonctionnaire dont les compétences restent ignorées. Ces situations ne sont pas des cas isolés. Elles suggèrent une transformation silencieuse du management dans la fonction publique.

La fonction publique présente des défis managériaux spécifiques, documentés depuis plusieurs décennies. Ces défis oscillent entre sens du devoir, convictions et valeurs partagées, et contraintes telles que l’austérité budgétaire.

À partir de récits de vie d’agents publics, notre étude révèle des situations où les responsabilités augmentent, alors que les moyens et les marges de décision diminuent. Cette évolution contribue à affaiblir l’engagement des fonctionnaires et contractuels. Notre approche narrative permet d’appréhender les enjeux contemporains de la gouvernance publique d’État à partir de l’expérience vécue de professionnels.

Injonctions contradictoires

L’équilibre entre contrôle de l’activité des agents de la fonction publique et liberté décisionnelle nécessaire à toute prise de décision reste un sujet de débat. Depuis plus de vingt-cinq ans, le chercheur Christopher Hood questionne la mise en œuvre du New Public Management, tandis que Geert Bouckaert invite à repenser la notion de performance et ses critères dans le secteur public. Dans ce contexte, les fonctionnaires se retrouvent face à des injonctions contradictoires : préserver une qualité de service envers les concitoyens avec des moyens limités et des obligations de résultat.

Le New Public Management trouve sa transposition en France à travers des programmes d’actions interministériels lancés en 2007 et 2012. La Révision générale des politiques publiques (RGPP) a été notamment consacrée à la réorganisation des structures de l’administration publique. La modernisation de l’action publique (MAP) à l’amélioration de la qualité des services publics ou la simplification et de réorganisation administrative.

Notre analyse met en lumière les formes concrètes du désenchantement des ressources humaines dans la fonction publique et la dégradation de la qualité du management liées à ces évolutions, notamment à travers deux cadres théoriques :

  • La justice organisationnelle qui analyse les appréciations des salariés du caractère juste du traitement dont ils font l’objet au sein de la structure qui les emploie ;

  • Les conflits de rôles qui naissent quand les périmètres d’activités ne sont pas suffisamment bien définis ou l’existence d’une répartition déséquilibrée de la charge de travail entre agents de la fonction publique.

Manque de reconnaissance (premier récit de vie)

Appel aux consultants extérieurs

« Ce qui me désole au sein du ministère, c’est l’indifférence manifeste envers les CV des fonctionnaires. Il est vrai que nous avons été embauchés selon certaines conditions de niveau et de compétences, mais après plus de quatre ans comme moi au sein des services, les agents ont eu l’occasion d’acquérir de nouvelles compétences. Plutôt que de faire appel à des consultants externes pour leurs compétences, on pourrait exploiter les compétences internes qui sont réellement abondantes et variées. »

Tâches routinières

« Avant de rejoindre la fonction publique, j’ai occupé plusieurs postes dans divers cabinets d’expertise comptable, où j’ai géré de manière autonome la comptabilité d’entreprises reconnues. J’ai choisi la fonction publique pour satisfaire mon père qui valorise la stabilité et la sécurité qu’elle offre. Cependant, mon rôle actuel à la direction générale se limite à des tâches routinières et subalternes. Ce manque de reconnaissance de mon niveau de compétence et d’expérience me désole. »

Qui est responsable ? (second récit de vie)

« Je suis en position de responsabilité, en tant que cadre depuis cinq années. J’ai la charge de la gestion des soignants au sein d’une organisation hospitalière multi-sites. Il y a des difficultés pour accompagner tous nos patients, accueillis sur un site et quelquefois opérés sur un autre site. Il faut alors trouver des places pour l’hospitalisation. »

« Nous expliquons que c’est au chef de service de déterminer qui sera le premier patient de la journée et l’heure à laquelle il sera hospitalisé. À notre niveau, nous sommes des représentants administratifs. Nous devons répondre à des sujets qui sont pour eux administratifs, mais qui en réalité sont organisationnels. C’est compliqué : qui est responsable ? »

Gérer seule une situation difficile (troisième récit de vie)

Problématique en dehors des clous

« Je suis directrice dans une petite école et institutrice dans une classe de CE2. J’ai vingt-cinq ans d’expérience et on peut dire que je fais partie des piliers de l’école. Jusqu’à cette situation, j’étais passionnée et ce métier a toujours été une véritable vocation. »




À lire aussi :
Quand le cynisme mine l’engagement dans la fonction publique…


« Depuis la petite section, nous accueillons un enfant avec des troubles importants – sans connaître l’origine de ses troubles, car sa maman n’a jamais voulu fournir de bilan médical à l’école. Il a pu bénéficier d’un accompagnant d’élève en situation de handicap durant 5 ans, mais plus aujourd’hui. C’est donc un peu moi qui joue ce rôle, mais en plus je dois m’occuper des autres élèves de ma classe. »

Pas de soutien de l’académie

« La situation était tenable tant qu’il bénéficiait de cette personne dédiée. Il a toujours eu tendance à frapper les autres et je pense qu’une bonne partie des enfants de l’école y ont eu droit. Cette année, les problématiques se sont amplifiées. Il y a eu de nombreuses plaintes des parents. On m’a même dit : “Mais que faites-vous madame, vous ne surveillez pas ?” “Convoquez la mère !” »

« Je me suis donc retournée vers l’académie et l’inspecteur, en demandant conseils et aides. Il m’a été répondu qu’il ne faudrait surtout pas céder aux plaintes quotidiennes des parents. Je n’ai pas reçu d’aide ou de moyens supplémentaires. Je pleurais tous les jours. J’aimais mon métier d’enseignante, mais là, je n’avais plus le sens. Mon médecin m’a dit, “C’est un burn out”. »

Hypocrisie organisationnelle

L’ensemble de ces récits permet de proposer plusieurs perspectives d’analyse à la lumière de l’hypocrisie organisationnelle – discours et actes contradictoires. Par exemple, le troisième invite à se demander si cette hypocrisie ne sert pas à protéger l’organisation, ici l’éducation nationale.

Une autre perspective est celle du leadership. Les trois récits suggèrent que :

  • Les agents publics sont contraints d’endosser des responsabilités pour gérer des situations complexes, sans recevoir de reconnaissance lorsqu’ils réussissent à trouver des solutions. L’écart entre les missions confiées et les moyens disponibles devient difficilement supportable ;

  • Ils ne disposent parfois pas du pouvoir décisionnel nécessaire pour résoudre les problèmes auxquels ils sont confrontés ;

  • Ils restent souvent prisonniers d’un statut ou d’une organisation hiérarchique rigide, qui freinent l’évolution de leurs missions, malgré l’acquisition d’expérience ou de nouvelles compétences. Cette situation génère à terme démotivation ou démission, qu’elle soit silencieuse ou effective.

Ces récits illustrent une forme de « management du risque » de la part des niveaux hiérarchiques supérieurs, qui consiste à laisser les collaborateurs affronter seuls certaines situations, tout en revendiquant la maîtrise de la décision et de l’action. Ces extraits mettent en lumière l’importance cruciale de la reconnaissance et de la valorisation du travail, tant au sein de la fonction publique qu’au service de l’usager ou du citoyen. En filigrane se dessine la nécessité d’un leadership éclairé et bienveillant, capable de soutenir et guider ses équipes.

The Conversation

Les auteurs ne travaillent pas, ne conseillent pas, ne possèdent pas de parts, ne reçoivent pas de fonds d’une organisation qui pourrait tirer profit de cet article, et n’ont déclaré aucune autre affiliation que leur organisme de recherche.

ref. Trois récits de vie pour comprendre le mal-être dans la fonction publique – https://theconversation.com/trois-recits-de-vie-pour-comprendre-le-mal-etre-dans-la-fonction-publique-271891

Gamblers don’t understand ‘free bets’ – and the costs can be huge

Source: The Conversation – UK – By Jamie Torrance, Lecturer and Researcher in Psychology, Swansea University

Gorodenkoff/Shutterstock

“Welcome bonus: get 150% up to £150 on your first deposit”. It’s the kind of offer that greets anyone who visits a British online betting site. What it doesn’t say is that if you decide to spend £50 on this offer, you’d need to stake an additional £750 of your own money before any winnings could be withdrawn.

Recent research by colleagues and I asked nearly 600 UK bettors to work out the true cost of exactly that kind of offer. Nearly everyone got it wrong, underestimating the real amount often by hundreds of pounds.

Financial inducements, “free” bets, deposit matches and welcome bonuses are a standard part of signing up with almost any UK operator. Their behavioural harms are well established. They encourage people to gamble more often, push bettors towards riskier wagers and are linked to chasing losses. The heaviest effects tend to fall on those already experiencing gambling harm.

But behaviour is only half the story. Harm can also stem from something more basic: not understanding what an offer actually requires of you in the first place. That’s where wagering requirements come in – the rules saying you have to bet the bonus amount a certain number of times over before any winnings attached to it can be withdrawn.

Until recently, these multipliers could be as high as 50 times the bonus. Since January 2026, the UK Gambling Commission has capped them at ten times and required operators to make their terms clearer. It’s a meaningful step. But it stops short of requiring operators to show consumers what that ten times multiplier actually means in pounds and pence, and that omission turns out to matter.

Our research

We ran an online experiment with 585 adults who had gambled in the past year. Each participant saw a realistic welcome bonus modelled on a real 2025 promotion, fully compliant with the 2026 rules. Half saw it in the standard industry format. The other half saw the same offer with one addition: a three-sentence example spelling out what the 10 times wagering requirement actually meant for a £50 deposit.

The correct answer was £750. The median estimate was £500. More than 90% of participants underestimated the true cost. Only around 5% got it right.

The £500 figure is telling. It is exactly what you would get if you applied the 10-times multiplier to the £50 deposit but ignored the 150% bonus on top. Most people understood part of the calculation but missed the compounding effect.

Matched bonuses combined with wagering multipliers are among the most common inducements in the UK. Together, they appear to obscure the true cost in a systematic way.

Crucially, this misunderstanding was not confined to any one group. People at low risk of gambling harm miscalculated at almost the same rate as those at high risk. The issue is not that some bettors are bad at maths. It is that the offer itself is structured to make the true cost hard to calculate.

When we added the worked example, attractiveness ratings dropped significantly. Once people could see what the offer required, they found it far less appealing.

Man using online sports betting services on phone and laptop
Over 90% of UK bettors misjudge ‘free bets’ gambling bonuses.
Kaspars Grinvalds/Shutterstock

What bettors told us

Participants’ responses revealed three consistent themes. Many described the offers as manipulative, using words such as “predatory” and “deceptive”. Others argued they were economically worthless, with one participant saying “99% of people will fail to benefit”. Many also called for stronger regulation.

Several made a comparison worth taking seriously: gambling inducements, they argued, should follow the same upfront disclosure rules as credit products. One 23-year-old said wagering requirements should be shown on the advert itself, “similar to how interest rates need to be shown clearly on sites offering loans”.

They may have a point. The Annual Percentage Rate was introduced in UK consumer credit precisely because people couldn’t compare loan products when costs were hidden behind different headline formats. Gambling inducements present an almost identical problem.




Read more:
Why people are watching livestreams of influencers gambling – and how it could be fuelling addiction


Capping wagering requirements at ten times is welcome. But it’s not the same as making costs visible. Even a reduced multiplier still requires a multi-step calculation, and an understanding of compounding that many people do not have.

A worked example, shown in the same print size as the headline offer, would take only a few lines. It would not ban anything or restrict choice. But our study suggests it would change how people evaluate these offers. Denmark already requires something similar. Australia, Spain, Belgium and Italy have gone further, banning inducements to new customers altogether.

Worked examples are not a complete solution. But as a low-cost addition to existing Gambling Commission rules, they could help consumers see these offers for what they are before deciding whether to take them up.

The Conversation

Jamie Torrance has received, in the last three years: (1) Open access publication funding from Gambling Research Exchange Ontario (GREO), (2) Conference travel and accommodation funding from the Academic Forum for the Study of Gambling (AFSG), (3) A minor exploratory research grant from the ASFG and GREO, (4) Seed Grant funding from the International Centre for Responsible Gambling (ICRG), (5) Studentship funding from the Economic and Social Research Council (ESRC), (6) Rapid evidence review (RER) funding from UK Research and Innovation (UKRI), (7) Policy Fellowship funding from UKRI, and (8) A Gambling Harms Research and Innovation Partnership (GHRIP) award from UKRI.

ref. Gamblers don’t understand ‘free bets’ – and the costs can be huge – https://theconversation.com/gamblers-dont-understand-free-bets-and-the-costs-can-be-huge-279873

Video game nostalgia is powerful – but it only works if developers understand the psychology behind it

Source: The Conversation – UK – By Alan Stewart, Senior Lecturer in Games Development, University of Portsmouth

Nostalgia is a powerful emotion for gamers. Gorodenkoff/Shutterstock

When thinking back to the gaming experiences of your youth, it’s easy to get misty-eyed. Depending on your generation, you might yearn for Dizzy on the Commodore 64, marathon GoldenEye sessions with friends or your first Minecraft world.

And games publishers know this. Recognising the nostalgic appeal of games, there has been a recent slew of re-releases of beloved classics. Tomb Raider Remastered (2024), Croc: Legend of the Gobbos (2025) and System Shock 2: 25th Anniversary Remaster (2025) all show publishers know how to capitalise on nostalgia.

Our research into player engagement frequently encounters nostalgia as a key driver of player satisfaction. Nostalgia is both potent and delicate: powerful when done right, disappointing when not. So, why do some games nail nostalgia while others fail?

Cultural theorist Svetlana Boym has argued that two different types of nostalgia exist: restorative and reflective.
Restorative nostalgia is about conquering the past – a belief that what was lost can, once again, be found, with enough effort. Whatever your political views, there’s no denying that the slogan “Make America Great Again” is a clear encapsulation of a restorative nostalgic philosophy.

Trailer for The Blizzard Arcade Collection.

The Blizzard Arcade Collection (2021) is a good example of a game design underpinned by this restorative philosophy. Here, players can experience original versions of Blizzard’s early work with visual effects applied that mimic the effect of playing the games on old, grainy televisions from the 1990s.

By contrast, reflective nostalgia is less interested in accurately recreating the past and keener, instead, on recapturing a remembered feeling. For instance, instead of yearning for the specific place in which they grew up, a reflective nostalgic may realise it’s the sense of carelessness they felt as a kid they yearn for – not their actual hometown.

Shovel Knight’s (2014) design adheres to this philosophy wonderfully. Synergising a smorgasbord of inspirations from original NES games – a world map akin to Super Mario Bros 3’s (1988), themed bosses like those seen in Mega Man (1987) – it feels less like a specific game a millennial might’ve played as a child and more like their fragmented memories of playing games in the late 1980s.

Shovel Knight has sold well since its release and, as of 2026, has made an estimated US$12.5m (£9.3m) in revenue. This shows there’s a hunger for nostalgic content among players. But, this being the case, why aren’t all re-releases cash cows? The answer relates to the above nostalgic philosophies.

Nostalgic inconsistency in remasters

If Blizzard Arcade Collection and Shovel Knight represent a strict adherence to their respective nostalgic philosophical underpinnings, then Croc: Legend of the Gobbos Remastered (2025) serves as the antithesis. Croc is certainly not a bad attempt at a remaster, but its nostalgic inconsistency dampens its emotional impact.

Take the art direction. By dropping Croc’s original pixelated textures, the remake signals that it isn’t aiming for perfect authenticity. Instead, it reflects a nostalgic approach that priorities reinterpretation over faithful recreation. Some critics even argued that the new textures look flatter, losing the illusion of depth that the original textures created.

A direct comparison of the Croc: Legend of the Gobbos remaster and the original game.

At other times the game adheres rigidly to the original source material, aligning it with a restorative nostalgic philosophy in ways that feel incredibly dated. The anticlimactic pause screens that appear upon defeating a boss – who seems remarkably unaffected by the beating it just took – is one such example. So, too, is the lack of level scenery – indefensible in an era of high-powered PCs capable of rendering vastly more complex environments.

Personally, we’d have enjoyed a re-release of Croc that more wholeheartedly committed to the reflective philosophy – updating textures, sure, but also level scenery and boss animations. Doing so might have helped to better capture the joy we felt when playing the original even if it meant deviating more widely from the source material.

It’s true that dissatisfaction at the choices made when updating beloved cultural artefacts is nothing new though. Indeed, Boym cites the restoration of the Sistine Chapel ceiling frescoes as a potent symbol of this. In preserving Michaelangelo’s original work, the restoration team were lambasted for removing aspects that made the original artwork so special.

It’s likely you’ll always disappoint some when restoring a historical relic. But if more developers commit wholeheartedly to a nostalgic philosophy when remaking their games, we’d argue they’d more effectively nail nostalgia and, in turn, leave players more satisfied.

The Conversation

The authors do not work for, consult, own shares in or receive funding from any company or organisation that would benefit from this article, and have disclosed no relevant affiliations beyond their academic appointment.

ref. Video game nostalgia is powerful – but it only works if developers understand the psychology behind it – https://theconversation.com/video-game-nostalgia-is-powerful-but-it-only-works-if-developers-understand-the-psychology-behind-it-277421

What is the chance of a message in a bottle being found?

Source: The Conversation – UK – By Kevin Burke, Associate Professor in Statistics, University of Limerick

Jenny Sturm/Shutterstock

Recently, a cheerful 100-year-old message in a bottle was found on the south-west coast of Australia. In it, a world war one soldier proclaimed to be “as happy as Larry”.

If you’re a betting person, you probably wouldn’t expect great odds of this happening. A bottle cast into the ocean could end up absolutely anywhere.

If it floats to a remote location, there is little chance of somebody stumbling upon it. And if it lands somewhere more favourable where people could potentially find it, there are other issues. The message itself will deteriorate over time as light degrades it. If the bottle fills with water, it will sink and almost certainly never be found.

So, what are the chances of a message in a bottle being found and it being over 100? And what are your chances of finding this bottle?

Despite these many possibilities during a bottle’s lifetime, the probability we are after is a straightforward calculation. Just count up the number of bottles with messages that have been found and are over 100 years old, and divide by the number of messages that have been sent this way (assuming we know how many are sent):

Our diagram below shows a hypothetical situation where 20 bottles are sent in total, of which six are found (indicated in gold) and one of these is over 100 years old (indicated by the “100” stamp). So, one in 20 bottles are found and over 100 years old. (Note: This is only a hypothetical calculation, not the real data.)

Instead of calculating the probability directly, another way to do it is by breaking the problem into two parts: (A) a bottle with a message is found, and (B) the found bottle is over 100. These two probabilities can be calculated separately and multiplied together to get what we want:

This is known as the “multiplication rule” of probability, and we confirm from our hypothetical numbers that (6/20)×(1/6) = 1/20, as before.

Both approaches to calculating this probability are simple. However, the direct calculation requires knowing the total number of bottles sent out, which is very difficult to know in the real world.

The multiplication rule has the advantage that it breaks the calculation into two parts. We can tackle each separately, then bring the two results together to get the probability we want. This is useful in the real-world situation where we can draw information from different sources.

First, we’ll deal with the probability that a bottle with a message is found, irrespective of its age.

Experts from the Federal Maritime and Hydrographic Agency of Germany suggest a one in ten chance that a message in a bottle will be found. This aligns broadly with various historical “drift bottle” experiments, where oceanographers released large numbers of bottles to understand ocean currents.

For example, studies from the 1960s and ’70s in the North Atlantic Ocean led to recovery rates of 14% from the Gulf of Mexico, 8% from the Caribbean Sea and 7% from the northern Brazilian coast. A more recent and more northerly study (between Canada and Greenland) from the 2000s led to a 5% recovery rate.

We would expect the results to vary naturally from different experiments in different parts of the world. But to keep things simple, we will stick with 1/10 as the probability that a bottle with a message is found.

Now for the second piece of the calculation: of the bottles that are found, what proportion are over 100 years old?

The table below summarises data from news articles collected on Wikipedia about very old bottles with messages that have been found. However, only data on bottles over 25 years old has been collected, presumably because older bottles are more newsworthy.

So, we needed to estimate the number of 0- to 25-year-old bottles with messages ourselves – here’s how we did this.

The table shows that fewer bottles with messages are found as they get older. Messages in bottles degrade over time, which means the bottles have an increased chance of breaking and sinking, or just getting covered in layers of sediment. Plotting this data in the graph below helped us see the trend in the ages of found bottles more clearly.

We drew a line to match this observed trend in the ages of found bottles. This red line in the graph corresponds to the equation:

This equation provides an estimate of how many bottles have been found for any specific age range (where 25 = 0-to-25, 50 = 25-to-50 and so on). We are interested in the the 0- to 25-year-old bottles, so the equation suggests 46 bottles have been found in this range.

Adding up this and all of the numbers in the table gives a total of 106 bottles found, of which 12 are over 100 years old, and 12/106 is about one in ten.

Recapping the above, we have that: (A) one in ten bottles with messages are found, of which (B) one in ten are over 100 years old. Bringing these results together using the multiplication rule, we estimate the chance of a message in a bottle being found and it being over 100 years old to be (1/10)×(1/10) = 1/100.

So, if there are 100,000 bottles with messages floating around the oceans waiting to be found, we’d expect 1,000 of these to be found and be 100 or more years old. Assuming anybody in the world is equally likely to find one of these, with 8 billion people currently, that’s about a one in 8 million chance of you finding one – pretty unlikely.

However, some people are more persistent at message-in-a-bottle hunting than others. Following the paths of ocean currents (known as gyres) could provide clues on where to look.

Specifically, peninsulas or islands intersecting with these gyres could be good spots. For this reason, it has been suggested the Caribbean islands are ideally placed for finding bottles as they lie on the path of the North Atlantic Gyre. Which seems like a great reason to travel to the Carribean!

But let’s also spare a thought for the poor soul stranded on their desert island, who surely won’t appreciate the low odds of their SOS being found.

The Conversation

Kevin Burke receives funding from Research Ireland.

David O’Sullivan receives funding from Research Ireland

ref. What is the chance of a message in a bottle being found? – https://theconversation.com/what-is-the-chance-of-a-message-in-a-bottle-being-found-272122

‘Protected’ seagrass meadows aren’t necessarily healthy – because pollution doesn’t stop at the shoreline

Source: The Conversation – UK – By Heidi McIlvenny, PhD Candidate, School of Biological Sciences, Queen’s University Belfast

_Zostera marina_ or common eelgrass is a type of seagrass. Ben Jones / Ocean Image Bank, CC BY-NC-ND

I spent last summer wading through seagrass meadows across Northern Ireland, from the sheltered waters of Strangford Lough to the exposed coast at Waterfoot Bay. I was collecting seagrass leaves and testing them for nitrogen pollution. Every meadow I visited sits inside a marine protected area – a stretch of sea that’s been given legal protection to safeguard the wildlife living there. And every single one was polluted beyond the limit for healthy seagrass.

Seagrass meadows are among the most valuable habitats in our coastal waters. They store carbon, nurture young fish and shellfish, stabilise sediment and buffer shorelines from storms. They are also woven into the heritage of coastal communities who have fished and foraged around them for generations. But they are disappearing worldwide, and nitrogen pollution from farming, sewage and urban runoff is one of the biggest reasons why.

It’s easy to assume that designating an area as “protected” keeps the habitat inside it safe. My research shows that, for seagrass, this assumption is dangerously wrong. Physical protection from anchors and dredging means little when pollution flows freely across the boundary from the land.

What matters most for seagrass is not lines drawn on a map, but what happens on shore.

To understand how much nitrogen pollution seagrass is absorbing, we can measure nitrogen content in the leaves themselves. Seagrass continuously takes up nutrients from the surrounding water, so the chemistry of its tissue works like a long-term pollution record. And my results showed that every meadow in Northern Ireland exceeded the pollution limit.

But knowing the pollution level is only useful if you know how much is too much, and what it means for the health of the meadow. To answer that, we pulled together data from 13 countries across the northern hemisphere and found a clear pattern.

catshark lying on seabed in seagrass meadow
A catshark shelters among seagrass.
Shannon Moran / Ocean Image Bank, CC BY-NC-ND

When nitrogen in the leaves rises above 1.8%, seagrass starts to suffer and loose plant growth. Above 2.8%, the decline accelerates rapidly, and in this danger zone small increases in pollution trigger disproportionately large plant losses. Think of it as a traffic light system: green is below 1.8% where meadows can cope; amber is between 1.8% and 2.8%, where managers should be watching closely and acting to reduce pollution; and red is above 2.8%, where urgent intervention is needed before the damage becomes irreversible.

The starkest example of a meadow in the red zone comes from Dundrum Bay, on the County Down coast. According to government assessments, it’s healthy. But my data tells a different story. Nitrogen levels here were nearly double the pollution limit of 1.8%. Surveys over the past decade paint an even bleaker picture: where lush meadows once thrived, dense mats of green algae now smother what little remains. This meadow has likely crossed a tipping point, and may never recover even if we clean up the pollution.

A few miles up the coast we see a very different picture. At Castle Espie, beside a wetland reserve in Strangford Lough, a seagrass meadow is thriving. The plants here belong to the same genetic population as struggling meadows elsewhere in the lough. But the difference is that the reserve’s reedbeds and willows act as natural filters, cleaning the water that runs from the land before it reaches the sea.

The same species with the same level of marine protection, but dramatically different outcomes. The difference is what happens on land. But current monitoring methods aren’t designed to spot this kind of trouble before it’s too late.

An early warning system

Current monitoring methods tend to measure how much seagrass is still there. But by the time a meadow visibly shrinks, the damage may already be done. The tissue chemistry approach we used picks up stress signals much earlier, while there is still time to act.

The nitrogen thresholds my research identifies could give environmental agencies a practical early warning system: meadows at or above 1.8% need closer watching, and those at or above 2.8% need urgent action to reduce nutrient pollution from catchments.

Seagrass meadows can recover but only if we tackle the pollution at its source. That means better management of urban and agricultural runoff, investment in sewage treatment and recognising that marine conservation cannot stop at the high tide mark. If we lose these meadows, we lose their carbon stores, their fish nurseries, the coastal protection they provide, along with a piece of our coastal heritage.

The Conversation

Heidi McIlvenny receives funding from the National Environment Research Council as part of the QUADRAT DTP. DAERA and Ulster Wildlife also part funded this PhD research.

Heidi McIlvenny is a member of Ulster Wildlife, RSPB and the National Trust, and Director of the Irish Ocean Literacy Network.

ref. ‘Protected’ seagrass meadows aren’t necessarily healthy – because pollution doesn’t stop at the shoreline – https://theconversation.com/protected-seagrass-meadows-arent-necessarily-healthy-because-pollution-doesnt-stop-at-the-shoreline-279805

Jane Austen in pictures: two graphic biographies bring the novelist’s life to the page

Source: The Conversation – UK – By Alex Fitch, Lecturer and PhD Candidate in Comics and Architecture, University of Brighton

To coincide with the 250th anniversary of her birth, last year saw the publication of two graphic biographies of Jane Austen.

These books take different approaches to the novelist’s life. The Novel Life of Jane Austen: A Graphic Biography, by Janine Barchas and Isabel Greenberg, is a slim volume in a cartoony style that begins in 1796. Kate Evans’ painterly book Patchwork: A Graphic Biography of Jane Austen, meanwhile, starts with Austen’s birth. The latter also includes historic digressions to give her writing additional context.

Both are handsome editions. The Novel Life of Jane Austen is slightly more approachable for the casual reader due to its brevity and smaller focus. But Patchwork may be more rewarding to graphic novel fans thanks to the painted artwork and a mixed-media collage approach that incorporates fabric cuttings into its pages.

Patchwork

Inspired by a patchwork coverlet sewn by Austen, her sister Cassandra and their mother, now displayed in Jane Austen’s House in Hampshire, Evans’ book shows how the bits and pieces of Austen’s life were recycled as material for her novels. Patchwork is an apt image for Austen, who “lopp’d and cropp’d” her prose as deftly as her cloth.

Some of Austen’s early biographers were fond of putting emphasis on the domestic tasks she undertook such as needlework, often deflecting attention from her desire to write and make money. As Austen researcher Kathryn Sutherland has recently observed in writing about Austen’s quilt: “Sewing, like letter writing, is not redundant work; it is necessary to social functioning.” It is apt that Evans’ includes this activity in her biography.

Evans also brings the bookish and close-knit Austen family vividly to life. The sense of the mischievousness and rebellion that brought Austen’s early work into being is present here. The comic strip renderings of her teenage skits and her spoof Plan of a Novel are wonderfully exuberant.

The texture of Austen’s life, and the threads that connect one life to others, is a distinct aspect of her approach. Evans’ shifts of focus add context to the author’s life. The material that Austen sews links ordinary domestic practice to the horrors of British colonialism which delivered cotton thread to Britain.

This incorporation of fabrics that Austen showed interest in during her life – with her family having to make do and mend during periods of impoverishment – is an continuing technique used by Evans in her graphic work. An earlier graphic novel by the artist – Threads: From the Refuges Crisis (2017) – included comic book documentation of interviews with women in the Calais “Jungle”. It embedded their embroidery and needlework in the book, made as a way of escaping the horror of their situation.

Patchwork uses the Austen family’s interest in fabric as a metaphor for their changing domestic experience in different decades and patchwork as a medium which gets added to over time. This works well as a visual leitmotif to accompany the narrative. It shows Evans’ continuing interest in depicting textiles alongside her artwork as a rich storytelling technique.

The Novel Life of Jane Austen

The structure of The Novel Life of Jane Austen echoes the typical triple volume publication of novels at the time. Budding Writer, Struggling Artist and Published Author indicate distinct phases of Austen’s writing life. They begin with a letter from August 1796 and end with Cassandra’s destruction of some of the correspondence in 1844 (the last pages acknowledge how her final house in Chawton has become a pilgrimage site for fans).

Barchas and Greenberg give a strong sense of Austen’s links with other writers, such as Shakespeare, whose “stunning literary celebrity” prefigures her own. For Evans, fictional influences can be seen in the presence on Austen’s bookshelf of those professional women authors (Frances Burney and Maria Edgeworth, for example) who did so much to set the stage for Austen’s success. However Samuel Richardson, whom Austen both revered and parodied, Henry Fielding, from whom she derived much of her facility with the burlesque and Samuel Johnson, an important influence on her sentence structure and tone, are either absent or fleetingly mentioned.

Like Evans, artist Isabel Greenberg uses a technique to contrast one aspect of Austen’s creative practice with another. While Evans contrasts Austen’s making patchwork fabrics with her writing, as two different but related creative pursuits, Greenberg uses different colour schemes to delineate ways that the author’s imagination operated at different times.

Most of the colour scheme is a combination of yellow and blue, including renditions of both Austen’s biography and performances of her novels. However, for scenes where the novelist enters a moment of reverie – imagining her characters’ inner lives – the colours used become red, purple and yellow. This signifies heightened feelings in contrast to the more muted emotions observed in her everyday life.

As with Evans’ mixed-media approach to graphic novel art, Greenberg’s use of heightened colour to separate everyday experience from imagination, is often found in the artist’s work. Her first two graphic novels The Encyclopedia of Early Earth (2013), and the recently adapted One Hundred Nights of Hero (2016) both contrast everyday experience in different historical time periods with characters’ storytelling activities.

Her previous project to illustrating The Novel Life of Jane Austen, was Glass Town (2020) which like the contrast between fantasy and reality in Austen’s life, took a similar approach to biography of Charlotte Brontë and her sisters.

Both of these Jane Austen graphic biographies are, in their different ways, reinvigorating guides. They emphasise different aspects of Austen’s life and times, using different visual techniques to bring the writer’s history vividly to life on the page. Most importantly, they both give prominence to the force of her imagination, and the innovative fiction it produced.

This article features references to books that have been included for editorial reasons, and may contain links to bookshop.org. If you click on one of the links and go on to buy something from bookshop.org The Conversation UK may earn a commission.

The Conversation

The authors do not work for, consult, own shares in or receive funding from any company or organisation that would benefit from this article, and have disclosed no relevant affiliations beyond their academic appointment.

ref. Jane Austen in pictures: two graphic biographies bring the novelist’s life to the page – https://theconversation.com/jane-austen-in-pictures-two-graphic-biographies-bring-the-novelists-life-to-the-page-278006

How Ontario’s post-secondary student funding changes echo Ronald Reagan reforms

Source: The Conversation – Canada – By Susan Dianne Brophy, Associate Professor in Legal Studies, St. Jerome’s University, University of Waterloo

Ontario’s recent announcement of a tuition increase and major changes to grant and loan structures have prompted student protests at the provincial legislature.

The province has said the changes are required for sustainability.

But changes to financial aid will have significant implications for many students who rely on grants and loans. As The Toronto Star reports, the reforms have almost reversed the ratio of non-repayable grants and loans students can access.

Education is a pillar of “social reproduction,” meaning it’s a social service necessary for maintaining daily life now and for future generations. When governments alter access to education and the way they deliver it, they shape everyday lives today and beyond.

Since legal and regulatory changes shape how society is reproduced, it is possible to draw from these changes some ideas about the government’s social values. From this perspective, Ontario’s Doug Ford government is sending the message that education is about generating private wealth and social order.




Read more:
What are universities for? Canadian higher education is at a critical crossroads


These changes risk entrenching inequalities and raise questions about students’ freedom and their futures.

For a precedent, it’s possible to look at the record of a past U.S. president, namely Ronald Reagan.

Education as a private asset

Currently, students can access up to 85 per cent grants and 15 per cent loans from the Ontario Student Assistance Program (OSAP). But under the financial aid reforms, a maximum of one quarter of a student’s OSAP funding will be non-repayable grants and a minimum of 75 per cent will be loans.

These changes mean there’s an upside for banks. With less funding through grants, students will be funnelled into private sector financial arrangements. Canada’s banks stand ready with student lines of credit.

As household assets (including financial investments, like Registered Retirement Savings Plans) continued to increase in value in the third quarter of 2025, it may seem rational and even attractive to view education as an asset meant to generate private wealth.

When Ford unveiled these changes, the private asset approach to education was clear when he responded to reporters:

“I mentioned to the students, you have to invest in your future, into in-demand jobs.”

Yet this approach ignores record-setting levels of household debt. It also glosses over the fact that the wealth gap is increasing. In the third quarter last year, the top 20 per cent wealthiest households accounted for 65.5 per cent of net worth, and the bottom 40 per cent accounted for 3.1 per cent.




Read more:
What Doug Ford could learn from Wisconsin about higher education


Just as we have seen with a profit-driven, wealth-generating housing market and the housing crisis, a private asset approach to education risks dividing society further into haves and have-nots.

This is not lost on students, as reflected in a recent University Affairs article quoting Grade 12 student Radhika Cappelletti:

“Things won’t run if people don’t continue to be educated and they can’t even choose to be educated because they can’t afford it.”

When students are financially bound to banks and dependent on their families, they face lasting pressures beyond the campus.

Revisiting Reagan

Ontario’s changes reflect a trend across the provinces that has been ongoing since the 1990s. They also follow a similar pathway as the Reagan era in the United States, with greater emphasis on student loans instead of grants.

As social and political theorist Melinda Cooper argues in her book Family Values: Between Neoliberalism and the New Social Conservatism, Reagan pursued student discipline and budget cuts for universities, relying on police for one and the introduction of tuition fees for the other.

As California governor between 1967-1975, Reagan sat on the University of California’s Board of Regents. Considered the “crown jewel of American public universities,” the university system benefitted from public funding during the post-Second World War era.

Bipartisan support for this was based on the belief that post-secondary education was a public good benefitting the whole state, not just graduates.

Later, as president throughout the 1980s, Reagan’s appetite to curb public spending grew, leading him to expand the role of loans and limit the availability of grants.

Cooper’s research shows that, as inflation outpaced wage growth in the early 1980s, growing wealth meant growing the value of assets — for those who had them. For those without assets, acquiring them required taking on debt. As the financial burden of education spreads through the family unit, it reinforces student dependency on the family, which encourages deference toward more traditional forms of authority.

Cooper finds that Reagan’s legacy was to make “parental responsibility” and “private-debt-based inclusion” the bases of access to education.

In these ways, socially conservative values resonate in what might otherwise be read as pragmatic, even politically and morally neutral financial decisions.

Narrowing educational paths

Ford’s plan to increase tuition sits in stark contrast beside his April 2023 announcement of free tuition for police trainees. This shows that the government’s approach to education reflects certain social values, which have consequences for the future.

Research suggests that viewing student debt primarily as an investment in their personal job prospects invites cuts to post-secondary degree offerings and opens the door to predatory for-profit institutions.

There is also a question of how students can even achieve a brighter future. As long as they remain dependent on existing power structures, it is difficult to expect anything other than an ever-widening wealth gap.

Another Ford initiative has been a push to allow students to opt out of fees that are the lifeblood of campus groups. The Student Choice Initiative failed several court challenges, but a version reappears in the fast-tracked Supporting Children and Students Act that passed in November 2025.

Critics say that this scaling back of student fees could have detrimental effects on equity-seeking groups and also potentially weaken student governance — something Ford has derided in the past.




Read more:
Ontario’s Bill 33 raises serious concern about campus equity and student rights


For all of Ford’s talk of choice and the future, then, changes in post-secondary funding limit the choices students have over their own lives. By deepening inequalities, Ford is casting a long shadow over the future of all Ontarians.

The Conversation

Susan Dianne Brophy is a member of the federal New Democratic Party.

ref. How Ontario’s post-secondary student funding changes echo Ronald Reagan reforms – https://theconversation.com/how-ontarios-post-secondary-student-funding-changes-echo-ronald-reagan-reforms-279534

Electric minibus taxis: the challenges and gains facing Cape Town’s transition

Source: The Conversation – Africa (2) – By MJ (Thinus) Booysen, Professor in Engineering, Stellenbosch University

The minibus taxi is ubiquitous in southern Africa. These vehicles are the backbone of the urban economy, providing affordable mobility for millions. In Cape Town, South Africa’s second most populous city, they are central to the transport landscape.

Around two-thirds of the city’s public transport users rely on paratransit services (which respond flexibly to demand), carrying about 830,000 daily passengers across 1,466 routes, and run by private individuals or associations rather than the state.

Minibus taxis in Cape Town, South Africa.

But because these vehicles run on petrol and diesel, they also contribute to greenhouse gas emissions, poor urban air quality and rising fuel costs.

The global shift away from internal combustion engines is accelerating, and public transport must be part of it. Bringing the electric vehicle transition to this sector, however, is not simply a matter of replacing one vehicle with another. In African paratransit systems, electrification raises a harder question: how do you change the vehicle without undermining the service on which so many people depend?

Electric minibuses would change how these vehicles operate, where and when they stop, how they interact with the grid, and driver decision making. They also require charging infrastructure that fits into the rhythms of taxi ranks, neighbourhoods and routes without disrupting service.

With Cape Town expected to launch its first few fully electric minibus taxi routes in Century City later in 2026, electrification is no longer a distant possibility. It is now urgent to understand whether it can work in practice for operators, passengers and the electricity grid.

We are a team of engineering researchers studying transport electrification in sub-Saharan Africa. In a series of studies, we have examined environmental and financial viability of electric vehicles under current mobility patterns, including charger placement, access, and adapted driving and charging behaviour.

Our new research found that electrifying minibus taxis is both necessary and possible. But it is also a complex challenge, with environmental trade-offs, grid constraints, operator costs and equity questions. Although our work focuses on Cape Town, the lessons are relevant to other African cities where paratransit dominates daily mobility.

Environmental perspective

The global narrative around electric vehicles often assumes they are a simple win for the climate. But this does not hold everywhere, especially where electricity still comes largely from fossil fuels. In South Africa, coal accounts for approximately 83% of electricity generation.

Petrol minibus taxi converted to electric.
MJ (Thinus) Booysen, CC BY-NC-ND

Using real minibus taxi mobility patterns in Cape Town, our research compared the energy use, emissions and costs of electric and conventional minibuses. It found a counter-intuitive result: under current grid conditions, an electric minibus taxi has about a 14% higher carbon dioxide equivalent footprint than a standard diesel minibus. In other words, charging an electric taxi on a coal-heavy grid can currently produce more greenhouse gas emissions than running a diesel vehicle.

That is not the end of the story. Electric minibuses still offer major environmental and health benefits. They eliminate tailpipe particulate pollution, reduce brake wear, and cut noise. These local benefits matter in dense urban areas where people live close to busy roads. As South Africa’s electricity system shifts towards more renewable energy, the climate case for electric minibus taxis will strengthen too.

So the real conclusion is not that electric taxis are a bad idea. Rather, they are a long-term climate solution whose immediate value lies especially in cleaner air, lower noise and better urban health.

Energy perspective

Electrifying Cape Town’s minibus taxi fleet would add substantial new electricity demand. In one study, the typical vehicle required about 50.8 kWh per day, scaling to roughly 460 MWh a day across a fleet of about 9,000 vehicles, or the equivalent of about 65,700 homes. The key issue is not just how much energy is needed but where and when vehicles charge.

Here, the newer work changes the story. It is tempting to think the answer is simply to install faster chargers at taxi ranks. But our modelling suggests that access to charging matters more than charging speed alone. Home or secure neighbourhood charging has the biggest effect on whether current mobility patterns can be sustained and on how well the system performs when driver behaviour adapts.

A typical daily charge of around 50 kWh might take roughly two to three hours on a 22 kW charger, or just over an hour on a 50 kW charger, though real charging times vary. But faster charging does not solve the real problem: drivers still need reliable places and enough stationary time to charge without undermining service or losing income.

The studies also show that chargers should not be planned only for formal taxi ranks. Infrastructure stops and informal stops matter too, because that is how paratransit actually works.

Viability of maintaining internal combustion engine mobility patterns for different charging scenarios.
DOI:10.1038/s41893-026-01808-9, CC BY-NC-ND

Nor will the effects be shared equally. Because apartheid-era geography still shapes where people live and work, operators in historically marginalised areas are more vulnerable when home charging is unavailable. Charging infrastructure is therefore not only a technical issue, but also an equity one.

There is also a grid challenge. Depot-only charging creates early-morning and daytime peaks, while home charging shifts demand into the evening residential peak. Unmanaged charging could therefore worsen stress on an already fragile electricity system. But time-of-use tariffs, managed charging, and better alignment with solar and other renewables could integrate electric taxis far more intelligently.

Operators’ perspective

For taxi operators, the economics of switching to electric vehicles are complicated. In one comparison, the electric option cost about 1.5 times as much as the diesel Toyota Ses’fikile – a 16-seater minibus – that currently dominates the market. Many operators already work on thin margins and face expensive finance.

The economics of switching to electric vehicles are complicated.
DOI: 10.1016/j.esr.2025.101892, CC BY-NC-ND

There are also financing costs: typically a 10% deposit and a 20% interest rate over a 72-month repayment period. Many operators may also be seen as high-risk by lenders, making finance difficult to access.

At the same time, the running-cost case for electric minibuses is much stronger. Energy costs are generally 33% to 57% lower than diesel fuel costs, and electric motors require less maintenance. For operators, then, this is a story of higher upfront cost set against lower operating cost, with the outcome depending heavily on electricity tariffs, finance terms and access to affordable charging.

Preparing for electrification

Careful planning and simulation are needed to roll out electric minibus taxis at scale. Policymakers need to understand the interactions between vehicle energy demand, charging infrastructure, grid capacity, driver behaviour and passenger service.

That is why we modelled driver behaviour in an electrified paratransit system. Unlike formal bus services, minibus taxi drivers adapt routes, stops and charging to passenger demand and competition. Our simulations show that constrained depot charging increases waiting times and reduces trips served. But with home charging, depot congestion falls sharply and service quality is largely maintained.

This matters because electrification is not just about vehicles and chargers, but about how informal transport systems actually work. If planners treat taxi operations like centrally controlled bus fleets, they will design the wrong interventions. The better approach is to plan around real mobility patterns, charging behaviour and neighbourhood inequality.

It is therefore crucial to bring taxi operators, municipalities, energy providers and communities together. Cleaner air and lower noise must be weighed against the grid’s current emissions profile. Operator economics must improve through better tariffs and financing. And charging infrastructure must be placed not only at depots and ranks, but also in the neighbourhoods and informal stops that shape paratransit every day.

With targeted subsidies, better overnight charging access, investment in renewable energy and clear policy support, Cape Town can begin building a public transport transition that is cleaner, more realistic and more just. If it gets this right, it could offer a blueprint for cities across Africa.

The Conversation

MJ (Thinus) Booysen as Director of the Electric Mobility Lab at Stellenbosch receives funding from the Western Cape Government and the Transport Education and Training Authority (TETA). The Electric Mobility Lab and Centre for Renewable and Sustainable Energy Studies (CRSES) have partnered with flxEV (GoMetro), Powerfleet (MiX Telematics), HSV and ACDC on the importation of South Africa’s first new electric minibus taxi, the eKamva.

Joshua Sello receives funding from the Global Strategic Communications Council (GSCC) and the DW Ackermann Bursary Scheme.

ref. Electric minibus taxis: the challenges and gains facing Cape Town’s transition – https://theconversation.com/electric-minibus-taxis-the-challenges-and-gains-facing-cape-towns-transition-278808

Nelson Mandela was a towering global symbol – but how effective was he as a president?

Source: The Conversation – Africa (2) – By Anthony Michael Butler, Professor of Political Studies, University of Cape Town

Nelson Mandela remains one of the most revered political leaders of modern times. He is widely credited with guiding South Africa through a peaceful transition from apartheid to democracy. He embodied racial reconciliation, and lent moral authority to a fragile new state. Yet admiration for Mandela the symbol has often obscured a more difficult question. How effective was Mandela in the day-to-day exercise of presidential power?

Most assessments of political leaders focus on their impact in terms of economic success and policy achievements. Some are also assessed through their character, integrity and moral vision. Both approaches have value, but they risk missing something essential: how leaders actually use the power of their office.

I am a professor of political studies. In a recent study, I proposed a simple framework for analysing presidential leadership across four dimensions – the relationships between:

  • executive and symbolic power

  • party and state

  • domestic and international roles

  • formal authority and informal influence.

Applying this framework to Mandela’s presidency between 1994 and 1999, I derive a more complex, and more critical, assessment than is often offered. Such an analysis is useful at a time when Mandela’s legacy is increasingly contested.

A powerful symbol, a limited executive head

Mandela’s symbolic authority was extraordinary. He helped to stabilise a deeply divided society and reassured anxious minorities fearing a loss of power and privilege. He gave moral meaning to the new democratic order. His gestures, such as donning a Springbok (South Africa’s national rugby team) jersey at a world cup final and embracing former adversaries, were not incidental. They were central to his political project of reconciliation.

But Mandela showed far less interest in the executive dimension of leadership. He delegated most of the core work of governing to his deputy, Thabo Mbeki. He also allowed cabinet ministers considerable autonomy. As a result, key areas of policy were shaped without sustained presidential direction or public accountability. This mattered because the presidency in South Africa’s system combines both head-of-state and head-of-government roles. The potential synergy between symbolic authority and executive control was therefore left largely unrealised.

The consequences were especially visible in moments of crisis. The HIV/Aids epidemic, which intensified during Mandela’s presidency, required both decisive executive action and strong public leadership. Mandela delivered neither of those, and he later acknowledged his failure to act more forcefully.

Blurring party and state

Mandela also struggled to manage the relationship between the governing African National Congress (ANC) and the state. South Africa entered democracy as a dominant-party system, and the ANC’s authority was both a source of stability and a potential danger.

Rather than drawing clear institutional boundaries, Mandela endorsed practices that blurred them. The policy of “cadre deployment” – placing loyal party members in key state positions – was intended to transform a state inherited from apartheid. But it also weakened institutional autonomy and contributed to longer-term problems of patronage and politicisation.

Mandela was not alone in shaping these developments. Many of the ideas originated with colleagues such as Mbeki. But as president, he lent his authority to them and did little to mitigate their risks.

Foreign policy: ideals and inconsistencies

Internationally, Mandela was expected to champion human rights and democratic values. Early statements under his name suggested that these principles would guide South Africa’s foreign policy.

In practice, however, foreign policy was often inconsistent. The government maintained close relationships with authoritarian regimes that had supported the anti-apartheid struggle. There were also tensions between proclaimed values and strategic or financial considerations. Efforts to isolate Nigeria after human rights abuses, for example, generated backlash within Africa. Relationships with countries such as Libya and Indonesia raised questions about the role of party funding in diplomatic decisions.

Mandela’s global stature brought South Africa visibility and goodwill. But this was not systematically used to advance clear domestic or economic priorities.

The hidden world of informal power

Finally, Mandela’s presidency illustrates the importance of informal power. Beyond formal constitutional authority, leaders shape outcomes through networks, appointments, and the mobilisation of financial resources.

Mandela was deeply involved in fundraising for the ANC, both domestically and internationally. Some of these practices blurred the line between party and state, and between legitimate support and undue influence. He also relied on personal relationships and informal interventions to shape economic and political outcomes. For example, he gave R2 million (hundreds of thousands of US dollars) to embattled politician Jacob Zuma in 2000, followed by another R1 million on 23 June 2005, days after Mbeki had sacked Zuma as deputy president and prosecutors had announced he would be charged with corruption.

Such practices were not unique to Mandela, nor to South Africa. But they helped establish patterns that would later become a problem, particularly as competition within the ANC intensified and access to resources became central to political power.

Rethinking a legacy

None of this diminishes Mandela’s historic role in ending apartheid or his contribution to national reconciliation. He set an important precedent by stepping down after a single term, and he helped to anchor South Africa’s constitutional order in its formative years.

But a focus on leadership practice rather than rhetoric, symbol and myth reveals a more uneven record. Mandela was an exceptional symbolic leader. He was less effective in integrating that symbolic authority with the demands of executive governance, institutional design and policy leadership.

Reassessing Mandela in this way is not an exercise in revisionism for its own sake. It is a reminder that even the most admired leaders operate within constraints. Understanding how they use power is essential if we are to learn from their successes, as well as their limitations.

The Conversation

Anthony Michael Butler does not work for, consult, own shares in or receive funding from any company or organisation that would benefit from this article, and has disclosed no relevant affiliations beyond their academic appointment.

ref. Nelson Mandela was a towering global symbol – but how effective was he as a president? – https://theconversation.com/nelson-mandela-was-a-towering-global-symbol-but-how-effective-was-he-as-a-president-279599