Singularity Institute

Target date: 2 weeks from start date.

Step 3: Publicize
Research best practices used by other video contests, and by similar contests by relevant communities.  Send "come enter our contest" advertisements to relevant websites, to marketing-related degree programs at good universities, and to other communities that may be interested. 

Target date: 4 weeks from start date.

Step 4: Judge

Award winners will be selected based on the product of two scores:

• Videos' accuracy and relevance in presenting the key idea, as assessed by current SIAI researchers and interns.
• How memorable the video is, and how likely viewers are to reference the video during dinner conversations or to recommend the video to their friends, as assessed by votes of viewers.  

Step 5: Assess

Assess effectiveness of contest, as detailed below. Use this assessment to update our picture of how best to promote rationality and, by so doing, reduce existential risk. 

[1] The "starting date" is the date (guaranteed to be within six months of the receipt of grant money) when we have skilled people to allocate to the project.  Extra donations increase our base of skilled people and thereby increase the number of projects we can get to; the lagged start date allows us to find new people, bring them here, and train them.

Budget: $6,000 for the initial contest, counting both direct expenses and labor.

Budget breakdown:

• $3,200 prizes
• $2,300 labor.
• $500 in AdWords, hosting, etc., including some slack for unexpected needs

Prizes:

• 1st prize of $2,500
• 5 2nd prizes of $100 each
• 10 honorable mentions with $20 SIAI T-shirts

How this project, if funded, will reduce existential risk:

Benefit 1: One or more videos that memorably present a piece of core SIAI content (as well as many videos that make some attempt).

Benefit 2: Many people who were already interested in SIAI, such as Less Wrong readers, will spend at least a bit of time brainstorming memorability, and/or will read linked marketing material such as "Made to Stick". Some will get better at communicating core SIAI content.

Benefit 3: Some new people will spend time on the issues, either because they hear about the contest and it directly appeals to them, or because their friends enter the contest and they watch their friends' videos.

Benefit 4: We may attract some good marketing folk to AI risks and other existential risk.

Risk 1: Using YouTube videos may impair academic respectability around SIAI or existential risk.

How actual impact will be assessed, so that we'll know whether to repeat YouTube contests and similar efforts:

An assessment will be written up afterward which records:

• How useful are the best resulting videos? Do they draw any new researchers, donors, or other useful folk to be involved with AI risk (i.e., do we ever later hear someone say they became involved because they saw video X, and then got more engaged)?
  
• Do we hear people citing this video, or ideas/metaphors used by this video, in conversation on existential risk? Do we find ourselves referring people to this video, as part of a set of orienting readings?
• How many Google search results does the contest website have?  How many Google search results does the winning video have?  To what extent do Google search results for the contest website, and for the winning videos, include people likely to be useful in AI risk reduction efforts?
• The number of page views the contest page received, and the number and type of people who entered, or who read AI risk material because they were thinking about entering, as indicated by entries into the contest website's form.
  
• The number of YouTube views and comments obtained by the total set of entries.
• Whether anyone with exemplary marketing skills makes their way into the AI risk reduction community and cites this contest as their point of entry.
• Whether we spontaneously receive any negative comments about the contest's image from academics, and whether FHI academics or others who care about AI risks or related topics give positive or negative comments on the contest when asked.