Hey there! I’m Gergő, the founder of EA Hungary and Budapest AI Safety.
If you would like to connect with the Hungarian EA/AIS communities please feel free to message me! We host lots of events! :)
Leave anonymous feedback on me here:
Anonymous feedback to EA Hungary here:
Thanks for sharing these! Would you recommend any podcasts for people who want to get into US AIS policy?
Will send you a msg!
EA Hungary Head of Communications/Operations, early 2024
We are currently looking for highly motivated individual to join the team of EA Hungary. In this form you can apply for the position. Please make sure to read the full description before applying.
Selection process: We expect to have 2 or 3 rounds. First round being this application form. Applicants continuning to the next stage will have one or two additional interviews/work tasks.
Salary: 9-12 USD per hour (pre-tax). For exceptional candidates we might be able to offer a higher rate up to an additional 1-10 USD per hour.
(Income tax in Hungary for people under 25 is 0%, above is 33%. For students above 25 years old it is 15%, but you will have to double-check this with your accountant.)
Workload: This is a part-time position. We expect to make an offer for between 15 to 30 hours per week. If our courses get an exceptionally high application rate, we might be able to offer additional hours.
A full time role has 25 days per year of paid vacation, accordingly a part time position would have between ~9 to ~19 days of paid vacation per year.
Contract period: Starting around early March, you’d have a contract for 6 months with a 2 month-long trial period after which your performance will be evaluated. Around month 4-5, we will evaluate your progress again, based on which we can offer you a contract for an additional 6 months.
More about this role:
EA Hungary is currently run by its founder, Gergő Gáspár, and its current Head of Operations, Milán Alexy, as well as a number of committed volunteers. We are looking for an additional person to the team and would be happy to tailor the role towards their strengths, hence the possibility of joining either as an operations or communications person.
Example tasks for the operations role:
- Preparing and running meetups and events
- Coordinating our fellowships (e.g., evaluating applications, sorting people into groups based on their availability, supporting our mentors, booking and preparing venues with food and reading materials, etc.)
Example tasks for the communications role:
- Communicating with our partners via email/calls
- Marketing our programs (e.g., marketing on social media, reaching out to professors and academic dormitories, giving a short “pitch” of our program to university classes, tabling, distributing leaflets, etc.)
- Writing newsletters
- Public speaking (desirable but not essential)
Additional example tasks for both roles:
- Giving career advice
- Facilitating fellowships
- Supporting volunteers
We expect to see some additional overlap between the two roles as our organisation evolves.
An ideal candidate:
- Is passionate about making the world a better place using EA principles
- Has a good understanding of EA concepts (at least equivalent to having done the at-home reading version of the introductory course and the in-depth course)
- Knowledgeable about the main EA cause areas and the organisations working on them
- Has great attention to detail, has good interpersonal skills
More about this role:
You will be working closely with EA Hungary founder Gergő Gáspár and other members of the team. We expect that about 50-60% of your work hours will be “on-site” (either holding sessions or coworking with team members), with the rest of your work hours being possible to do online. The first 2 weeks of onboarding will be 100% in-person.
Location: Budapest, Hungary.
Deadline to apply: 2024. February 12.
Apply here!
With any questions, please write to gergo@eahungary.com
Budapest AI Safety Technical Lead application form early 2024
We are currently looking for a highly motivated individual to join the team of Budapest AI Safety. In this form you can apply for the position.
Selection process: We expect to have 2 or 3 rounds. First round being this application form. Applicants continuning to the next stage will have one or two additional interviews/work tasks.
Salary: 9-12 USD per hour (pre-tax). For exceptional candidates we might be able to offer a higher rate up to an additional 1-10 USD per hour.
(Income tax in Hungary for people under 25 is 0%, above is 33%. For students above 25 years old it is 15%, but you will have to double-check this with your accountant.)
Workload: This is a part-time position. We expect to make an offer for between 0.25 FTE to 0.6 FTE, but are unlikely to offer more than 15 hours per week for the first 6 months. If our courses get an exceptionally high application rate, we might be able to offer additional hours.
A full time role has 25 days per year of paid vacation, accordingly a part time position would have between ~6 to ~15 days of paid vacation per year.
Contract period: Starting around early March, you’d have a contract for 6 months with a 2 month-long trial period after which your performance will be evaluated. Around month 4-5, we will evaluate your progress again, based on which we can offer you a contract for an additional 6 months.
Skills/experience/responsonsibilities for this role:
Essential:
- Ability to give valuable career advice and help people new to technical AI Safety reseach orient in the field
- Facilitating AGISF
- Supporting volunteers
- Leading student-level research projects
- Contributing to our newsletter section on technical AIS
Desirable:
- Being able to support the post-AGISF part of the AI Safety pipeline by facilitating advanced courses and leading small research projects
- Research experience in Alignment
- Public speaking experience
Depending on the amount of your work hours and the capacity of the rest of the team, you might also be asked to do some operations-related work in a smaller capacity, such as:
- Evaluating applications, sorting people into groups based on their availability, supporting our mentors, booking and preparing venues with food and reading materials, etc.)
- Marketing our programs (eg. marketing on social media, reaching out to professors and academic dormitories, giving a short “pitch” of our program to university classes, tabling, distributing leaflets, etc.)
An ideal candidate:
- Is passionate about AI Safety
- Knowledgeable about the main threat models and technical research agendas within AIS and the organisations working on them
- Has good interpersonal skills
Location: Budapest, Hungary or remote (please read the additional information below)
If we are unable to find a suitable candidate we will consider hiring someone fully remote, so feel free to apply for this role even if you are only interested in a remote position. In this case, we will hold onto your application until we have evaluated candidates who are based in Budapest or are willing to move here. If we don’t find a suitable candidate we will start processing fully remote applications. If you are only interested in a remote position, for now it’s enough if you answer the first 4 questions of this form. If we start evaluating remote candidates, we will notify you to answer the rest of the questions.
More about this role:
You will be working closely with Budapest AI Safety founder Gergő Gáspár and other members of the team. We expect that about 50-60% of your work hours will be “on-site” (either holding sessions or coworking with team members), with the rest of your work hours being possible to do online. The first 2 weeks of onboarding will be 100% in-person.
Deadline to apply: 2024 February 12.
Apply here.
With any questions, please write to gergo@eahungary.com
Great post! Thanks for writing it, this is now on my default list of the resources I send to new organizers! :)
Hey DoGooder!
Welcome to the forum! This is indeed a super important, as well complicated question. I don’t have a confident opinion, but thought I would point out some potential cruxes that might help you to form a view on this!
The specific type of the animal that people are eating is going to be pretty important, see this video for a quick summary as to why. Reading and thinking about moral weights is going to be relevant here too.
How would you balance suffering caused to animals vs. reducing poverty by purchasing products vs. harm caused by the emissions of such products. Huh this is a difficult one, perhaps reading “Important Between-Cause Considerations: things every EA should know about” can offer a start into thinking about this.
How bad do you think it is if a e.g. chicken is killed if they had a decent life? Reading about population ethics will be relevant here. On specifically the badness of death, I think HLI has a very good article that shows how different theories think about this.
There is probably a bunch of other considerations that don’t come to mind at the moment, but hope some of these will be useful!
Also I just wanted to mention that I know a couple of EAs who are originally from Pakisan but live in Hungary, as well as a person living in Pakistan—if you would like to be connected to them ping me a private message!
Wow, even more impressive then haha!
Hey Bella, thanks for your questions!
What targeting settings did you use on the ads?
We targeted university-age people (so around 18-26) in Budapest (+~40km radius). In the future, I also plan to include people up to 35 years old (or more).
Here are the tags I used. Many of them come from EA CZ’s @Jakub Sechter who did structured interviews with their members about what stuff they follow on Facebook—so I think those must be pretty good. I also added a bunch on my own which may or may not be useful. (Unfortunately, I can’t find Jakub’s original list, but what I have in the docs includes all of his suggestions)
As a side note, I think you can also set up ads specifically to one university if you set the radius very small—but I haven’t tested this yet. If you do that though, you shouldn’t use any tags as that would filter out too many people from an already small sample.
What creative (images/videos) did you use?
I used either general posters that are available on CEA’s Canva (such as this and this) - or pictures of our group photo. My general impression was that a friendly group photo gathers more clicks /$ spent, but I don’t know if that also translates to getting more motivated applicants, unfortunately. I haven’t used videos before, but I was told that they tend to perform better (ofc the cons being is that they are more time to make)
One mistake I made though was that I always used just one poster/picture for a campaign in the past. Since then I learned that you should upload several posters, group photos in different angles/proximity etc. - and then after the first hours/days, Meta’s algorithm will automatically prioritize those that perform best. (this is true for different descriptions, headlines etc. too)
What copy (text that goes in the ad) did you use?
I added them to the second half of this doc, both for the EA and AIS group :)
Did any of those do especially well?
Unfortunately, I didn’t get to compare different messages yet, apart from what I wrote about the pictures vs posters stuff! I recommend using several descriptions etc!
Edit: After writing the initial reply I realized most of the additional “tips” I wrote will be pretty obvious to you, as I just saw that you work in marketing haha x) - I guess I will leave the comment as it is, in case it’s helpful for others!
Thanks for sharing, this is amazing stuff! May I ask why the name of the software you used to illustrate the results?
Can you share a bit about what’s the reasoning behind making this list?
I’m relatively new to AI policy, and all I know about Politico is that a while ago they published an article on EA that seemed to be written in pretty bad faith. (Looking at the titles of some of the articles, I imagine this is true for some of the others as well.) I imagine many other readers of the forum have even less context than myself.
I think it’s pretty amusing that much of the mainstream media thinks that EA is some coordinated illuminati-like movement, even though that is incredibly far from being the case. However, dropping a list of EA-critical articles on the forum and speculating on journalists future behaviour without any further explanation actually does make that impression—even though I think most readers of this forum will be just as confused by seeing this list as any journalist from the manisteam media would be.
(Sorry if this comment comes off as confrontational, I actually didn’t mean it that way, I’m sort of just thinking out loud)
That’s amazing, thanks for sharing!
The courses on https://aisafetyfundamentals.com/ are a common starting point for people!
I would also share this on LessWrong if you haven’t already!
That’s amazing, thanks for creating this!
Hey Ezra! Thanks for writing this up and sharing your thoughts. I wish you lots of strength and endurance throughout these challenging times.
Thanks for asking, that’s really something I should have mentioned in the post!
As for in-session readings of Intro to EA fellowship—we did it with overall ~40 people in 2 rounds.
As for the intensive fellowship from the previous post, we did it with overall ~45 people in 2 rounds (EA fellowship), and another ~20 people in 1 round (AGISF).
Hey Andreas! Thanks for writing this up, it was a really interesting read and I’m glad you shared it!
Some quick rambling thoughts after reading:
I think some of the distinctions might be semantic—some of what you describe would fall under misuse risk/malicious use, which could indeed be a real problem (If an AI is causing harm because it its values are aligned with a malicious human, is it aligned or misaligned overall? I’m not sure, but the human alignment problem seems to be the issue here) - and I’m not sure how to weight that against the risk of unaligned AI. I think given that we are nowhere close to solving the alignment problem, people tend to assume that if we have AGI, it will be misaligned “by definition”. In terms of s-risks, I would really recommend checking out the work of CLR, as they seem to be the ones who spent most time thinking about s-risks. I think they also have a course on s-risks coming up sometime!
Thank you for sharing!
Sorry if I’m asking something obvious, but is it fair to assume that all of these internships require US citizenship? (According to GPT there could be exceptions so I thought I would ask you just to be sure)
Thanks for flagging these great opportunities!
This sounds really amazing, thanks for creating such a service. I would love to apply and take part, but with my current workload, I feel I wouldn’t be able to 4-6 hours per week to the program and I suspect many others are in a similar situation.
Do you think it would be possible to share the materials/exercises for those who would be interested in doing the program at their own pace, asynchronously, kind of like it was a workbook? (I realize sharing it would probably be not this straightforward as it was developed for groups, but I thought I would ask:) )