HAL will be down for maintenance from Friday, June 10 at 4pm through Monday, June 13 at 9am. More information
Skip to Main content Skip to Navigation
Preprints, Working Papers, ...

Tournesol: A quest for a large, secure and trustworthy database of reliable human judgments

Abstract : Today's large-scale algorithms have become immensely influential, as they recommend and moderate the content that billions of humans are exposed to on a daily basis. These algorithms are the de-facto regulators of the information diet of billions of humans, from shaping opinions on public health information to organizing groups for social movements. This creates serious concerns, but also great opportunities to promote quality information [Hoa20, HFE21]. Addressing the concerns and seizing the opportunities is a challenging, enormous and fabulous endeavor [HE19], as intuitively appealing ideas often come with unforeseen unwanted side effects [EMH21], and as it requires us to think about what we truly and deeply prefer [Soa15]. To make progress, it is critical to understand how today's large-scale algorithms are built, and to determine what interventions will be most effective. Given that these algorithms rely heavily on machine learning, we make the following key observation: any algorithm trained on uncontrolled data must not be trusted. Indeed, a malicious entity could take control over the data, poison it with dangerously misleading or manipulative fabricated inputs, and thereby make the trained algorithm extremely unsafe. We thus argue that the first step towards safe and ethical large-scale algorithms must be the collection of a large, secure and trustworthy dataset of reliable human judgments. To achieve this, we introduce Tournesol, an open source platform available at https: //tournesol.app. Tournesol aims to collect a large database of human judgments on what algorithms ought to widely recommend (and what algorithms ought to stop widely recommending). In this paper, we outline the structure of the Tournesol database, the key features of the Tournesol platform and the main hurdles that must be overcome to make it a successful project. Most importantly, we argue that, if successful, Tournesol may then serve as the essential foundation for any safe and ethical large-scale algorithm.
Document type :
Preprints, Working Papers, ...
Complete list of metadata

https://hal.archives-ouvertes.fr/hal-03390514
Contributor : Vlad Nitu Connect in order to contact the contributor
Submitted on : Thursday, October 21, 2021 - 1:59:42 PM
Last modification on : Wednesday, March 16, 2022 - 3:52:05 AM
Long-term archiving on: : Saturday, January 22, 2022 - 7:04:59 PM

File

Tournesol__White_Paper_.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-03390514, version 1

Citation

Lê-Nguyên Hoang, Louis Faucon, Aidan Jungo, Sergei Volodin, Dalia Papuc, et al.. Tournesol: A quest for a large, secure and trustworthy database of reliable human judgments. 2021. ⟨hal-03390514⟩

Share

Metrics

Record views

36

Files downloads

1