Sessions endend just now, and it was very interesting. It gave me a brand new insight on how to understand the new age of data collection as a continuation of colonial dispossession, I will dig this up

I will particularly be attending today's session, about Digital Surplus and Data Colonialism.

Show thread

One of the researcher from my team is organizing a workshop on AI consequences on society: how those technologies can amplify our racial and sexist biases, or make jobs even precarious? Those are some of the questions this seminar aims to answer:

Inscription is free but mandatory, feel free to join. 15h-19h from today to saturday

I think there is this issue with PhD studies where no one, except the student, can assess the quality of the work. If you have either a low self-esteem or high expectations, things can go spiraling down real quick. The student is the only one deciding on a day-to-day basis when they should stop working or what is to be considered "a good working day". There are no objective, outside metrics that helps to assert that, because almost every research activity is peer-based.

Show thread

I have an intern since two month or so, and although it is incredibly rewarding to see him grow and become more and more capable, I still fear that he does not learn "enough". Of course there are goals for his internship (we will soon submit a paper), but I don't know what he should retain from this internship, how should a "good researcher" behave.

people know that academics don't get paid for publications right?

Heya. The materials for my tutorial at PFIA 2020, "Formal Verification of Deep Neural Networks", are available here:

The tutorial will take place monday morning at 10 a.m. CET

You can subscribe to the webinar here:

A paper, named " A Deep Neural Network Model to Predict Criminality Using Image Processing", will be published in a Springer venue. You might find the title unsettling. Well, guess what, it is. Using data produced by the criminal system will not "predict crime", it will predict who the police arrests. Which are two really different things.

More details there:

I'll be giving a tutorial on formal verification of deep neural networks at PFIA 2020. All material (slides, code) will be distributed freely.

A quite pessimistic point of view of french security and cryptograpy experts on a COVID-19 tracking app (in French only):

Fifteen simple abuse case scenario, a state of the art of the possible form such an app coud have... an interesting reading

it's gross and creepy to see patterns in sites like Netflix where I'll watch a few films in Spanish or watch a couple films with Black leads, or i'll watch a few LGBTQ romances, and then all of a sudden the whole site changes. film covers are changed to display BIPOC folx more often, i somehow unlock whole new categories of films, etc.

then i watch a couple of blockbusters one weekend and the site reverts to a homogenized safety bubble of middle class hetero white society

it's scary to think how many sites create these safe spaces for privileged people to erase everyone not like them, and no one even notices

Coucou Mastodon !
Jeudi 5 mars, je suis en grève pour l'action "La recherche s'arrête"
J'aimerais toucher les élèves à qui j'enseigne. J'ai l'impression qu'on les exclut souvent de nos luttes; ils sont de plus assez peu politisés (l'informatique, tout ça)
J'aimerais détourner les deux heures que j'ai pour discuter avec eux de la réforme, de politique en général...
Est-ce que vous avez des conseils de comment procéder ? Des supports ?
Boosts appréciés !!

Hello! I'm a researcher at Harvard's Shorenstein Center working on disinformation, censorship, and media manipulation in general. Glad this place exists :)

PJ Reddie, one of the most wholesome and cool scientists of this world, quitted research in computer vision due to ethical concerns:

A relatively large group of French scientists propose to add a fictional Alex to all our publications in all domains. This would make the Alex an extremely prolific scientific publisher:

I think (and the people behind this idea do as well) that this initiative may be interesting to researchers all over the world. Unfortunately, the page they provide is in French. Maybe they will set up another one in English later.


A cool conference on the deep learning hype, at 36c3:

Good questions here:
* is deep learning a sustainable technology in our resource-limited world?
* are those deep learning systems useful for anyone else than big companies and a restricted set of academics?
* What about the anonymous workforce used to label or mimic deep learning datasets?
* how the heck can we continue to publish to top-tier conferences with 10% of results that are not reproductibles

If I were a scholar on economical history and philosophy, I would say that it's just capitalism trying to devour another field. That after reaping everything it could from our diseases, our culture, our food, our emotions, it goes for our knowledge and spirit.

But I am none of those. I am just a poor PhD student raging against this state of affair, and hoping that things will change.

Show thread
Show older
Scholar Social

Scholar Social is a microblogging platform for researchers, grad students, librarians, archivists, undergrads, academically inclined high schoolers, educators of all levels, journal editors, research assistants, professors, administrators—anyone involved in academia who is willing to engage with others respectfully.