Impact dashboard

This follows the /open practice as set out in [Kran, 2022] to share the impact of our non-profit projects as openly as practically feasible.

Alignment Jam hackathons

Last updated August 18th 2023. The following data is output from a single R Markdown file in the /open Apart Research Github repository.

Survey results

The following results are automatically gathered from the evaluation survey.

• Difference in prior knowledge (5.00) and posterior knowledge (6.10) about AI safety: 1.10
• Difference in self-reported AIS career probability before (4.90) and after (7.20): 2.30
• Promoter score ("would you recommend this to a friend"): 8.70
• Origin countries: Germany, India, Latvia, France, UK, Slovakia, Poland, Kenya, New Zealand, Czech republic
• Participation countries: Austria, Moscow, Mexico, Roorkee, Denmark, France, UK, United States, Poland, Kenya, New Zealand, Czech republic
• Self-reported probability (-100% to 100%) that the participant would not have worked on a similarly impactful research project: 82.50%. -100% means they would have worked on a more impactful project otherwise while 0% indicates they would have worked on something similarly impactful.
• Self-reported change in motivation to work on AI safety long-term (-5 to 5): 2.6

General participation data

Apart Research website visits

SimpleAnalytics is cookies-free analytics software. Click in the top left of the graph for in-depth visitor information.

Alignment Jam website visits

The Alignment Jam website is where all the fun hacking happens. This is where participants sign up, see information about the hackathons, and submit their projects.