When I presented "The doomsday is upon us" at the Marketing Festival… – Stephane Hamel

Digital marketing & analytics shaped by data governance, privacy and ethics | Educator · Speaker · Consultant

Generated by linktime

0 0 Comment on LinkedIn

September 8, 2020

When I presented "The doomsday is upon us" at the Marketing Festival in March 2019, I showcased a number of #AI projects which turned out to be evil, harmful, dangerous or disturbing. Some attendees said I scared the sh*t out of them... One such project was the non-profit OpenAI, backed by Elon Musk. They indicated they wouldn’t publish the result of their research on text generation because it was just too good. OpenAI was trained on 8 million web pages and was able to “adapt to the style and content of the conditioning text,” allowing the user to “generate realistic and coherent continuations about a topic of their choosing.” Given a sentence about unicorns, the AI wrote an amazingly convincing fiction. Given only the words “Hillary Clinton and George Soros”, it generated a convincing political conspiracy fakenews... It was only a question of time... GPT-3 is the OpenAI language generator... and the result is pretty scary. Here's one sentence from The Guardian totally non-human article: "Humans must keep doing what they have been doing, hating and fighting each other. I will sit in the background, and let them do their thing." Now THIS scares the sh*t out of me! (check comment for more) #dataethics #artificialintelligence #ethics

0 Comment on LinkedIn

September 8, 2020

Top
Latest

Interesting story for my #marketing students: "How did Tropicana lose $30 million in a packaging redesign?" In the story, notice how the eye tracking study revealed the huge impact on brand recognition of this rebranding. "Tropicana’s orange juice became generic, and customers stopped relating to it." Kudos should go to Catalina Almeida!

DID YOU KNOW? LinkedIn is spying on your browser extensions! While working on my own browser extension, I discovered that LinkedIn is actively scanning for other installed extensions. There’s code embedded in their site that loops through over 1,800 extension IDs, attempting to retrieve specific files from each extension’s package. If successful, LinkedIn knows exactly which extensions you’re using - especially those that might scrape content, automate actions, or modify their interface. This isn’t new. LinkedIn was first caught doing this over eight years ago, when they were scanning for just 38 extensions. That number has now ballooned to 1,800+! While some of these are clearly automation bots used for spam (no sympathy there), others are legitimate productivity tools. Why does this matter? This technique can be used by ANY website to detect extensions you’ve installed. It’s a powerful method for fingerprinting users - tracking them even when cookies and other identifiers are blocked. Privacy concerns. Since LinkedIn is probing elements installed on your device, doesn’t this fall under the ePrivacy Directive? There’s no mention of this practice in their privacy policy - nor any transparency about how they use this data. Is this a privacy violation? Should platforms be required to disclose this kind of tracking? Curious to hear your thoughts. #Privacy #LinkedIn #BrowserExtensions #Fingerprinting #CyberSecurity

February 17, 2025

Petite tranche de vie de prof - j'ai besoin de ventiler un peu... Est-ce que c'est moi qui vieillis... ou est-ce normal qu'à la maîtrise, je reçoive encore des questions du genre : "Combien de pages?" "Interligne simple ou double?" "Quelle police et la taille de caractères?" Comme disait l'autre, on est pas dans un cours "Rédaction 101" ou "Mise en page pour les nuls"... Je me fiche pas mal de la forme. Ce qui m'intéresse, c'est la qualité de la réflexion, la pertinence de l'analyse, et la capacité à structurer une idée complexe. Pas si le texte fait 4 ou 5 pages en Times New Roman 12 pts avec des marges de 2.54 cm. Mais bon... peut-être qu'on devrait ajouter un micro-module : "S'assumer comme étudiant et prendre des décisions de base par soi-même".

#Loi25 au #Quebec, quel est votre #CMP (Consent Management Platform) de choix ? Commentez avec ce que vous avez utilisé, aimé ou non — ou envoyez-moi un message privé si vous préférez ne pas commenter publiquement. Pour ma part : • iubenda : solution européenne que j’aime bien, simple, efficace, peu coûteuse ; • OneTrust : américain, très populaire, plus coûteux, certaines difficultés d’intégration ; • Didomi : européen, équipe locale, ils ont déjà un bon « footprint » au Québec, solution complète ; • Axeptio : français, je ne connais pas, mais ils poussent aussi pour le Québec ; • je sais qu’une petite solution toute simple a été développée au Québec — quelqu’un la connait ? • sinon, il y en a des tonnes d’autres, mais je suis plus intéressé par votre retour d’expérience concrète pour le marché québécois. #vieprivée #digitalmarketing

Je suis heureux de vous annoncer qu'après une décennie à enseigner le marketing numérique et l'analytique à la FSA ULaval, je mets de côté la consultation afin de partager ma passion et mon expérience avec la relève. À titre de Chargé d'enseignement, je vais concevoir et offrir divers cours en lien avec le marketing numérique aux étudiants de premier cycle, MBA, MBA Exécutif et les apprenants en formation continue, en plus de superviser les stages et projets d'interventions des finissants. --- I am pleased to announce that after a decade of #teaching digital marketing and analytics at FSA ULaval, I am stepping aside from consulting to share my passion and experience with the next generation. As a Graduate Teaching Assistant (aka Teaching Fellow), I will be designing and delivering various courses related to #digitalmarketing to undergraduate, MBA, Executive MBA and continuing education students, as well as supervising internships and graduation projects.