Extended intelligences¶
Daniel Pitarch & Pau Artigas
28 November - 5 December 2023
Contents¶
Deliverable¶
Majority of the content we consume is heavily influenced by a series of biases. Nowadays we are constantly bombarded with facts, news, articles that attempt to persuade readers and mystify the reality of the facts. It is not easy to extrapolate the news as it is because it is constantly influenced by the geopolitical conditions of a country, its current of thought, and which side the journalists are on.
To adress this issue we created a tool that can use AI to identify and highlight biases.
Steps to use : Tell me which words of this text are biased?
1. Upload an article/text
2. Generate the Response
3. Visualize the Biases
Demonstration Video
Final Reflection¶
I was very pleased to be able to change my view on artificial intelligence. Initially, due to lack of interest and information, I conceived AI as a single, extremely powerful global tool; the idea that machines would advance faster than us and that soon the world we live in would be transformed into something similar to a futuristic movie from the 1990s.
The seminar with Pau and Daniel has helped me understand that AI is just another tool and, at the end of the day, it is created by people. This revelation has two remarkable implications.
On the one hand, since it is created by us, it cannot surpass us in knowledge. AI takes pre-existing data and organizes it in a way that automates actions we all know how to perform.
On the other hand, the fact that the AI’s creation is human implies that many of the judgments that reside intrinsically in it transfer, just as they do in us.
What I found most interesting about the seminar was the most debated and thoughtful dimension: the bias that AI can have, the stereotypes it can promote, to what extent it can be used, as well as the ethical and moral issues surrounding this topic.