Skip to content

Arkady Zgonnikov

Humans, AI, and Oxford commas

  • About
  • Bio
  • Publications and presentations
  • Research

About

I am an interdisciplinary cognitive scientist, with a background in applied mathematics, computer science, and cognitive psychology. Currently I am a tenure-track Assistant Professor in Human-Robot Interaction at the Department of Cognitive Robotics, TU Delft.

E-mail: a.zgonnikov at tudelft dot nl
TU Delft | open science framework | github
google scholar | twitter | orcid | researchgate

Arkady ZgonnikovFollow

Assistant professor @tudelft. Cognitive modeling, human-robot interaction, automated driving, meaningful human control over AI

Arkady Zgonnikov
4esterArkady Zgonnikov@4ester·
17 Oct

This

mike cook@mtrc

Many problems in AI stem from the broken link between human responsibility and machine action. Automated vehicle deaths, racist classifiers, autonomous weapons, ethical simulators.

“Be willing to take responsibility for everything your AI does” should be a core principle of AI.

Reply on Twitter 1449852231969349637Retweet on Twitter 1449852231969349637Like on Twitter 14498522319693496371Twitter 1449852231969349637
Retweet on TwitterArkady Zgonnikov Retweeted
xkcdRandall Munroe@xkcd·
16 Oct

Unsolved Math Problems http://xkcd.com/2529

Reply on Twitter 1449218820938293248Retweet on Twitter 1449218820938293248934Like on Twitter 14492188209382932485960Twitter 1449218820938293248
Retweet on TwitterArkady Zgonnikov Retweeted
m_guggenmosMatthias Guggenmos@m_guggenmos·
13 Oct

1/ Happy to share a new preprint on a computational model of metacognition (and an accompanying toolbox): https://www.biorxiv.org/content/10.1101/2021.10.10.463812v1

Reply on Twitter 1448173864396206086Retweet on Twitter 144817386439620608619Like on Twitter 144817386439620608661Twitter 1448173864396206086
Retweet on TwitterArkady Zgonnikov Retweeted
wgilpin0William Gilpin@wgilpin0·
12 Oct

Is chaos actually hard to predict? For NeurIPS this year I made a database of 131 known strange attractors, and trained state-of-the-art forecasting models on each one, to try to figure this out (1/N):
Paper: https://arxiv.org/abs/2110.05266
Dataset + Code: https://github.com/williamgilpin/dysts

Reply on Twitter 1447980535549075457Retweet on Twitter 1447980535549075457405Like on Twitter 14479805355490754571876Twitter 1447980535549075457
Load More...
Copyright © 2021 Arkady Zgonnikov. All Rights Reserved.
Fashify Theme by FRT