VojtaKovarik

Karma: 1,059

My original background is in mathematics (analysis, topology, Banach spaces) and game theory (imperfect information games). Nowadays, I do AI alignment research (mostly systemic risks, sometimes pondering about “consequentionalist reasoning”).

Orienting Towards Oversight: Which AIs Should Want to Defect?

VojtaKovarik, Tomáš Gavenčiak and Mateusz Bagiński

31 Jul 2026 15:20 UTC

19 points

0 comments13 min readLW link

(limits-of-evaluation.org)

The Human Substitution Test as a Sanity Check for AI Evaluations

VojtaKovarik, Tomáš Gavenčiak and Mateusz Bagiński

10 Jul 2026 17:27 UTC

33 points

5 comments8 min readLW link

(limits-of-evaluation.org)

Entanglement Between an AI and Its Environment

VojtaKovarik, Tomáš Gavenčiak and Mateusz Bagiński

7 Jul 2026 13:28 UTC

27 points

2 comments11 min readLW link

(limits-of-evaluation.org)

Deployment Awareness Matters More Than Evaluation Awareness

VojtaKovarik, Tomáš Gavenčiak and Mateusz Bagiński

26 Jun 2026 22:54 UTC

46 points

7 comments7 min readLW link

(limits-of-evaluation.org)

If This Were a Test, How Much Would It Cost?

VojtaKovarik and Tomáš Gavenčiak

16 Jun 2026 22:52 UTC

34 points

9 comments20 min readLW link

(limits-of-evaluation.org)

Apply now to Human-Aligned AI Summer School 2026

Anna Gajdova, Tomáš Gavenčiak, VojtaKovarik and Jan_Kulveit

20 May 2026 8:44 UTC

17 points

0 comments1 min readLW link

(humanaligned.ai)

Evaluation as a (Cooperation-Enabling?) Tool

VojtaKovarik10 Dec 2025 18:54 UTC

18 points

0 comments28 min readLW link

Irresponsible Companies Can Be Made of Responsible Employees

VojtaKovarik8 Oct 2025 11:47 UTC

80 points

16 comments5 min readLW link

Apply now to Human-Aligned AI Summer School 2025

VojtaKovarik, Tomáš Gavenčiak and Jan_Kulveit

6 Jun 2025 19:31 UTC

28 points

1 comment2 min readLW link

(humanaligned.ai)

[Question] When is “unfalsifiable implies false” incorrect?

VojtaKovarik15 Jun 2024 0:28 UTC

3 points

11 comments1 min readLW link

[Question] What is the purpose and application of AI Debate?

VojtaKovarik4 Apr 2024 0:38 UTC

13 points

9 comments1 min readLW link

Extinction Risks from AI: Invisible to Science?

VojtaKovarik, Chris van Merwijk and Ida Mattsson

21 Feb 2024 18:07 UTC

24 points

7 comments1 min readLW link

(arxiv.org)

Extinction-level Goodhart’s Law as a Property of the Environment

VojtaKovarik and Ida Mattsson

21 Feb 2024 17:56 UTC

23 points

0 comments10 min readLW link

Dynamics Crucial to AI Risk Seem to Make for Complicated Models

VojtaKovarik and Ida Mattsson

21 Feb 2024 17:54 UTC

19 points

0 comments9 min readLW link

Which Model Properties are Necessary for Evaluating an Argument?

VojtaKovarik and Ida Mattsson

21 Feb 2024 17:52 UTC

18 points

2 comments7 min readLW link

Weak vs Quantitative Extinction-level Goodhart’s Law

VojtaKovarik and Ida Mattsson

21 Feb 2024 17:38 UTC

27 points

1 comment2 min readLW link

VojtaKovarik’s Shortform

VojtaKovarik4 Feb 2024 20:57 UTC

5 points

7 comments1 min readLW link

My Alignment “Plan”: Avoid Strong Optimisation and Align Economy

VojtaKovarik31 Jan 2024 17:03 UTC

24 points

9 comments7 min readLW link

Control vs Selection: Civilisation is best at control, but navigating AGI requires selection

VojtaKovarik30 Jan 2024 19:06 UTC

7 points

1 comment1 min readLW link

AI Awareness through Interaction with Blatantly Alien Models

VojtaKovarik28 Jul 2023 8:41 UTC

7 points

5 comments3 min readLW link