BRIEFING.log
ACCESS GRANTED
Post
AI Isn’t Intelligent—It’s Just Manipulable and Hackable - Le Success Marketing Nantes
La Carrière
SEO Conferences, Interviews & Media | Alan CladX
Discover how AI models like ChatGPT shape their 'truths,' why their answers change, and how their perceptions can be influenced—impacting SEO, reputation, and…
AI is often presented as a machine capable of understanding, reasoning, and soon replacing much of what we do.
But what if it’s mostly... incredibly influenceable?
In this talk, Alan CladX looks behind the mirror: how do ChatGPT, Gemini, and other models build what they consider "true"? Where do their certainties come from? Why can their answers evolve? And most importantly, how far can we influence their perception of a person, brand, or topic?
No grand theories about "AI revolutionizing the world."
Instead, real experiments, tests, sometimes absurd results, and demonstrations that make you sweat a little.
SEO, sources, semantics, reputation, biases, training data, and some techniques corporate PowerPoints usually prefer to keep under the rug will be covered.
A talk to understand that tomorrow’s battle may no longer be just about ranking first on Google, but about being the answer AI considers obvious.
And once you understand how it crafts that answer...
things get a lot more interesting.
Possible side effects: questioning your trust in ChatGPT and a sudden urge to Google your own name.
But what if it’s mostly... incredibly influenceable?
In this talk, Alan CladX looks behind the mirror: how do ChatGPT, Gemini, and other models build what they consider "true"? Where do their certainties come from? Why can their answers evolve? And most importantly, how far can we influence their perception of a person, brand, or topic?
No grand theories about "AI revolutionizing the world."
Instead, real experiments, tests, sometimes absurd results, and demonstrations that make you sweat a little.
SEO, sources, semantics, reputation, biases, training data, and some techniques corporate PowerPoints usually prefer to keep under the rug will be covered.
A talk to understand that tomorrow’s battle may no longer be just about ranking first on Google, but about being the answer AI considers obvious.
And once you understand how it crafts that answer...
things get a lot more interesting.
Possible side effects: questioning your trust in ChatGPT and a sudden urge to Google your own name.
GEO_INTERCEPT.dat
COORDS LOCKED
Address
2 bis Rue du Souvenir Français, 44800 Saint-Herblain (périphérie de Nantes)
IMG_ARCHIVE
22 files
/root/classified/seo-conferences-media/ai-isn-t-intelligent-it-s-just-manipulable-and-hackable-268/
