Data

AfI/IAf #4: “AIs can be artists too” / “Les IAs peuvent aussi être artistes”

These past few years, we’ve had plenty of articles on AIs that had created amazing paintings, incredible texts or uncannily realistic music. While the results were sometimes not as truly unbelievable as the article’s buzz-title suggested, we have to admit that there’s an ever-growing trend of “creative AIs”, or AI artists. But this begs the question: can AI really be creative? Isn’t this ability to invent, to create, inherently human?

Read More »AfI/IAf #4: “AIs can be artists too” / “Les IAs peuvent aussi être artistes”

Data & Security/Donnée & Sécurité #2: Asymmetric Ciphers/Chiffrement asymétrique

Last time, I discussed how symmetric cryptography has long been in use for transmitting secret messages. Despite all of its advantages, we did point out some problems with symmetric ciphers: they require you to safely exchange keys and some can now be easily broken thanks to the newest technologies in computer science. Today, I’ll talk more about the counterpart: asymmetric cryptography. I’ll discuss its pros and cons compared to symmetric cryptography and I’ll also give a very rough picture of a possible game-changer: quantum cryptography.

Read More »Data & Security/Donnée & Sécurité #2: Asymmetric Ciphers/Chiffrement asymétrique

Data & Security/Donnée & Sécurité #1: Symmetric Ciphers / Chiffrement symétrique

Today, we live in a world of data. It is great, because it helps Google perform amazing searches and retrieve exactly the info you want; because it allows Facebook and Instagram to show you news that are relevant for you; because it makes it easy to connect to hundreds of applications and enjoy thousands of new technologies. But, at the same time, some scammers are currently take advantage of the Covid pandemic to trick people into giving out info and money with phishing, possible medical data leaks from the French to the American government have been pointed out and there are more and more scandals about hackers attacking hospitals or banks mixing up their clients data and having them stolen.

Read More »Data & Security/Donnée & Sécurité #1: Symmetric Ciphers / Chiffrement symétrique

Automated discourse analysis / Analyse automatique du discours

Over the past few years, artificial intelligence has been used in lots of applications. Its history is already paved of several milestones such as the golden years in the 70s, the fall in the 80s, the boom of big data, the rise of deep learning or more recently, better-than-human strategy game players. The past two years are considered by several AI prominent figures like Yann Le Cun to be the latest revolution in AI, with a focus on natural language processing (or NLP). Indeed, in 2019, OpenAI kicked the anthill once more on the topic of NLP when they released an AI capable of generating amazingly realistic text. They even decided to refrain from publishing the code – contrary to most of their projects – because they were worried it could be used to create fake news more easily.

Read More »Automated discourse analysis / Analyse automatique du discours

AfI/IAf #2: “Machines are objective” / “La machine est objective” (1/2)

In this second article in the Artifakal Intelligence / Intelligence Artifaussielle series, I’ll dive more into the famous phrase: “machines are logical, hence they are objective“. To which extent is it true that machines are unbiased? Can we really trust an AI to be completely “fair and neutral” when solving a problem?

Read More »AfI/IAf #2: “Machines are objective” / “La machine est objective” (1/2)