Over June I have been evaluating entries of the essay contest organised by Riesgos Catastróficos Globales, an organisation supporting Spanish-speaking work on Global Catastrophic Risks (GCRs). The contest ran from 4 March 2022 to 30 May 2022, and we received 141 (valid) applications. The applications were written in Spanish. 

I am immensely grateful to our support team (Cristina Schmidt Ibañez, Claudette Salinas, Alison Díaz and Emilio Bazan) for running the operations of the contest and helping me evaluate the entries, to the rest of the RCGs team (Juan García, Ángela Aristizábal and Pablo Stafforini), the Spanish Speaking EA community (especially coordinators Sandra Malagón and Laura González) and many university professors for their support and help promoting the contest; and to the FTX Future Fund Regranting Program for financing the contest.

The essay contest has been a great opportunity to promote Spanish content and writers. Also it has been an interesting exercise to learn how Spanish-speakers relate to GCRs. In this article I share some observations about the contest.

In particular, I talk about the reception of the contest, the content of the entries, the format of the entries and the quality of the entries. I also include an overview of the winning entries.
 

Key highlights

• The writing contest seems to have been a very cost-effective way of promoting engagement with GCRs. We spent $10k and got ~39k visits to our website and received ~141 valid essays.
• Unintuitively, the contest drove a lot of web traffic from Latin American countries with little EA presence like the Dominican Republic, Guatemala, Venezuela, Perú, Honduras, Ecuador and Nicaragua, and relatively little from places with moderate community presence like Spain, Mexico and Colombia.
• People mostly wrote about Global Catastrophic Risks in general and climate change. There were very few entries on other specific GCRs.
• Most essays were of rather poor quality. There were 26 essays that met a minimum bar of quality and and from those 11 that I considered good quality.
• You can read the winning entries of the contest (in Spanish) here.

Reception and geographical information

• I am very pleased with the reception of the contest. I was expecting to receive twenty-odd entries. Instead we received 141 valid entries.
• While the contest was up (March, April and May) our webpage received 39,377 visits. In comparison, in January and February we received <200 visits.
• Running the contest cost around $10k including prizes, salaries and ads. That's about $0.25 per visit or $70 per essay. Googling around I found cost-per-clicks between $0.15 and $0.75. I now believe that writing contests are a cost-effective way to promote engagement with a topic.
• The ads we ran through Facebook were not a big part of the traffic, though they were particularly cost-effective. We spent $100 in Facebook ads, and got 840 visitors through the social media site.
• Most visitors came through Twitter and direct search. We ran a campaign of emailing University professors in Latin American countries to promote the contest, though we don't have good data on how effective that was.
• Most visitors came from Dominican Republic (7355 visits, 18.7% of all visits), Guatemala (5792, 14.7%), Venezuela (5364, 13.6%), Perú (5012, 12.7%), Honduras (4895, 12.5%), Ecuador (3795, 9.64%) and Nicaragua (2383, 6.05%). This was surprising to me as these are countries where Effective Altruism and the GCR community has very little presence. The next most popular countries were Colombia (1758, 4.46%), México (806, 2.05%), Chile (597, 1.52%), United States (379, 0.96%) and Spain (290,0.74%), where the Spanish Speaking EA community has a moderate presence.
   

Content

• The most popular topic of the contest was writing about Global Catastrophic Risks in general (54 entries, 38% of all essays). One popular format was listing and briefly explaining different GCRs. Another were appellations to human nature to contextualise Global Challenges. One common theme was human greed and the failure of capitalism to address global risks. Most of them were poorly argued, and it makes me think that more people should read Inadequate Equilibria and/or take introductory economics classes.
• The second most popular topic was climate change and environmental disasters (31 entries, 22% of all essays). Concerns about climate change have permeated public conscience; this makes me very optimistic about progress in the problem in the coming decades.
• Other topics covered include Artificial Intelligence (5 entries, 3% of all essays), food crises (3 entries, 2% of all essays), natural risks (4 entries, 3% of all essays), biological risks (7 entries, 5% of all essays), nuclear risks (5 entries, 3% of all essays). I am surprised we had so few entries on biological risks, given the COVID-19 pandemic.
• We received many entries (45 entries, 32% of all essays) on topics that were not related to GCRs. The high volume suggests that the concept is not very well known among Spanish-speakers.
• There were around 4 entries arguing that we were living through an ongoing collapse that is intrinsically associated with capitalism. This is a frame I have not engaged with a lot, which made it hard to assess the entries. Overall, I consider the frame overly simplistic and not very productive, but I recognize I could be dismissing it just because I am not familiar with it. Here is an example of a university course syllabus with recommended readings which I think illustrates quite well the vibe of the essays. I was surprised to learn about this school of thought, and positively surprised that they were willing to engage with the GCR community.
   

Formats

• The most popular format was the essay (67 entries, 47% of all essays). If we include opinion pieces then 76 essays, 53% of all essays. The line between an essay and an opinion piece was not very well delineated.
• Fiction was a fairly popular format (42 entries, 29% of all essays). Most stories focused on narrating the aftermath of a GCR. Writing quality varied greatly, as well as plausibility and engagement with GCR ideas.
• One of my favourite entry formats were summaries of papers, reports and books. I think this is a great way to engage with a topic and create a valuable resource for others who wouldn't read the original otherwise. Sadly, we only received 3 entries of this sort, one of which was awarded the first prize.
   

Quality

• Most entries were of low quality. We discarded 103 entries (73% of total) because they were not topical for GCRs, were very poorly argued, had very little content or were outright spam.
• Most entries failed to adhere to basic style norms like writing an introduction and a conclusion, dividing their essays in sections or connecting ideas from paragraph to paragraph. I also would have liked to see entries following Reasoning Transparency style norms.
• Many entries failed to address their relation with Global Catastrophic Risks. For many I could tell that they had been written for a completely different purpose, and sent in the hopes it would qualify for a prize.
• Because of the abundant quantity of low quality essays, it was easy to delegate an initial pruning of the entries to interns in our support team following some basic guidelines, and focus on evaluating the remaining ones.
• There were a few entries (less than 5 entries) that gave me pause, because they covered topics that I was unfamiliar with or proposed ideas that were totally novel to me. I dug into the topics and, for most of them, they turned out to not be well supported by the evidence. Scepticism is really hard.
• There were some entries (26 entries, 18% of all entries) that were reasonably topical and reasonably well written, but which I would not have felt comfortable giving a prize to, because they didn’t reach the threshold of quality I was hoping for.
• After the initial evaluation, we selected 11 entries as finalists. Of these, 7 were awarded a prize. You can read the winning entries here.

Winning entries

• "Catástrofes existenciales y continuación fallida dentro del modelo de las tres capas" by Santiago Ramírez Sáenz. A critical review of Defense in Depth Against Human Exctinction.  (1st prize)
• “El cataclismo inefable” by Jhojan Galindo. A sci-fi story narrating how alien explorer perform an autopsy of the then extinct Earth. (2nd prize)
• “La adolescencia de la humanidad: Riesgos existenciales, largoplacismo ético y el (vasto) futuro de la especie humana” by  Jon Rueda Etxebarria. An introduction to longtermism and existential risk.  (2nd prize)
• “Un gran poder y una gran responsabilidad” by Pablo Antonio Moreno Casares. An introduction to AI Risk. (3rd prize)
• “Un asesinato sin asesino: el uso de robots en la guerra” by Sara María Varón Echeverri. An essay outlining the risks of Lethal Autonomous Weapons (LAWs). (3rd prize)
• “Cambio climático versus guerra nuclear: el rol de los nuevos reactores nucleares en la mediación de dos riesgos catastróficos globales” by  Sara María Varón Echeverri. An essay outlining the implications of new generation nuclear energy for climate change and nuclear risk. (3rd prize)
• “Siempre lo supimos” by Denver Romero Páramo. A story about the aftermath of nuclear winter. (3rd prize)

In short, the essay contest has been more successful than I initially expected, garnering a lot of attention and producing a few thoughtful essays. It has also allowed us to peek into the attitude towards GCRs of the Spanish speaking world - where concerns about climate change are widespread, but not so much about other GCRs.

In hindsight, I wish we had asked for more information on the application form so we could have studied more. Particularly, I would have been interested in demographics of the participants and how and when they learned about GCRs.

Following up the contest, we have invited the winners to participate in some activities. Depending on how that goes I might start advocating more loudly in favour of essay contests. So far it seems to have been a promising way of gathering attention - it remains to be seen if its an effective way of finding talent as well.

Thank you to Cristina Schmidt Ibañez for editing this post.