Opens profile photo
Follow
Click to Follow igilitschenski
Igor Gilitschenski
@igilitschenski
Assistant Professor in Computer Science at . Also Research Scientist at . Previously at , , .
Science & TechnologyToronto, Canadatisl.cs.toronto.eduJoined January 2009

Igor Gilitschenski’s Tweets

This week, 's #ChatGPT had its debut on the floor of the European parliament. used it to write a speech on the regulation of political advertisement. I wonder how ChatGPT would do in political negotiating. 😉 English translation: europarl.europa.eu/plenary/en/vod
Quote Tweet
Ich habe diese Woche im Europäischen Parlament eine von #ChatGPT geschriebene Rede gehalten. Und keiner hat‘s gemerkt! 😅 Da könnte man jetzt sagen: Cool, die erste von #KI-geschriebene Rede im #EU-Parlament, ein quasi historischer Moment. Mir geht‘s aber um etwas Anderes. 🧵👇
Show this thread
Embedded video
1:27
26K views
1
6
Show this thread
My train should win an award for friendliest staff with the most dry humour. 😂 Approaching one of them about forgetting to tap my card, I was informed that food in the Go train jail is much better than it used to be! 🍷🍱🧑‍🍳 I’m SO relieved.
6
Besides developing new technology, I have always also been passionate about its societal impact and broader public policy. Thus, I am excited to have joined as a faculty affiliate and very much look forward to interdisciplinary exchange opportunities beyond CS.
2
33
If you are working on HCI and looking for a cool postdoc position, my colleague is doing inspiring work. This is an amazing opportunity!
Quote Tweet
Please RT! Looking to hire an #HCI postdoc at University of Toronto in the broad areas related to Human-AI Interaction and Intelligent Programming Tools. Contact me or fill this form: forms.gle/sTYh5kwhvjDyLi #chi2023 #uist2023 #iui2023 @sigchi @ACMIUI
1
Need a new, easy-to-use RL algorithm? (Which is essentially DQN but for continuous control Tasks!) DecQN was accepted at #ICLR2023
Quote Tweet
In our recent paper, we show that a minor variation of DQN actually solves many continuous control problems from state or pixels on par with state-of-the-art actor critic methods such as (D4PG, DMPO, SAC, DrQv2, DreamerV2, ...). arxiv.org/abs/2210.12566
Show this thread
Image
2
92
Show this thread
Thanks, and team for inviting me to give a talk at the Robotics Seminar this morning. It was a lot of fun to meet all of you! Unfortunately, I was remote: the weather in Toronto ❄️☃️ is very different from the weather in Haifa☀️🌴 this time of year.
2
16
Happy to share the third talk 📢 in preparation for the Scaling & Models Panel at the #NeurIPS2022 Workshop on Robot Learning. 🦾 's and sharing their perspective on "Language as a Connective Tissue for Robotics"
25
We have from visiting the institute today. Tune in for his talk at 3pm ET.
Quote Tweet
Don't miss it! Nov 11, 2022 @ 3PM - Seminar with @NikolaiMatni Assistant Prof at Uof Pennsylvania @Penn & Google Brain Robotics @GoogleAI Visiting Faculty Researcher. Meet & greet follows 4 - 5PM for in-person. Virtual option youtube.com/watch?v=nwBq92 🤖🦾
Image
4
Can't wait for the #NeurIPS2022 Workshop on Robot Learning? 🤖👨‍🏫 We got you! In the upcoming weeks, we will post talks of our panelists starting with & who have cool insights on "Language as Robot Middleware" 🤖💬🚀
45