matthes

Karma: 1,231

I work as an engineer on cool stuff to improve the world. Currently, my focus is on finding and solving R&D projects in farmed animal welfare (academia and industry). Stay tuned for a big post on this.

I restarted High Impact Engineers. Check out and engage with the new forum!

I run EA Sheffield. Come along to our next social if you’re in the area!

I organised EA North 2025 and might make it a regular thing. It was very cost-effective.

I have a PhD in computer science and an undergrad degree in physics.

matthes 3 Jul 2026 22:59 UTC
2 points
0 ∶ 0
in reply to: matthes’s comment on: most EAs should probably not be living in high cost-of-living (HCOL) areas most of the time
I don’t think I have fully developed my opinions on compensation yet and very interested in hearing other people’s thoughts on the topic. I am surprised that it’s not being discussed more.
I think transparency about pay is great, and I hope I am not discouraging it with this post.

matthes 3 Jul 2026 21:34 UTC
9 points
2 ∶ 0
in reply to: Henry Stanley 🔸’s comment on: most EAs should probably not be living in high cost-of-living (HCOL) areas most of the time
I feel like not many people really have “a job that can mostly be done from anywhere”, and a lot hinges on that assumption here.
I am uncertain about this and it probably heavily depends on how you define “EAs”. My impression might be very biased.
I think your piece treats being in a high COL place as solely an EA move; in reality EA hubs are not picked from a random list and are attractive places in themselves
Hm, I don’t think it’s solely an EA move. I do prefer London, mostly for the EA community, but also because of access to art, food, architecture, and other niche communities. But I would consider choosing to live there a luxury, similar to taking expensive holidays. It seems strange to me that some employers are willing to subsidise and in some cases even financially reward the former.
This heavily depends on how exactly the COL-adjustment is made, and many orgs are not transparent about how exactly they do it. In most cases it seems like the adjustment is proportionate to your base salary. This means that in some cases, COL-adjustments encourage junior people to live in cheaper places while encouraging more senior people to move to more expensive places (because other expenses like housing aren’t relative to their salary.)

matthes 3 Jul 2026 6:06 UTC
10 points
2 ∶ 0
in reply to: Tristan Katz’s comment on: most EAs should probably not be living in high cost-of-living (HCOL) areas most of the time
I agree! It’s not that surprising that big cities are natural hubs. I also think that people who work on policy are a great example of a group that can’t really do their job from everywhere. So are people who earn-to-give in certain fields if they can get disproportionately high salaries in expensive cities. (Although the maths doesn’t work out like that in every role.)
In the UK, we probably also have a self-fulfilling prophecy, where almost all community building investment seems to go into Loxbridge. EA UK has historically almost exclusively focused on London. At one point while I was there, EA Oxford had 3 paid organisers with multiple free retreats a year. It’s not too surprising that you end up with lots of highly engaged EAs in places that get that kind of attention.
(I don’t think that all major cities and unis are completely equal in all ways, but I do think the differences are often overblown, at least in the UK.)

matthes 29 Jun 2026 19:31 UTC
2 points
0 ∶ 0
on: Looking for a co-founder in biosecurity startup (automated biosurveillance)
I recommend cross-posting this to the High Impact Engineers Forum, the Impact Forge. Good luck!

matthes 9 Jun 2026 20:09 UTC
1 point
0 ∶ 0
in reply to: Seth Ariel Green 🔸’s comment on: animal welfare has an evidence problem
Thanks for flagging this!
I should have double-checked this reference. Have made an edit to redirect people to this comment.

matthes 7 Jun 2026 11:12 UTC
13 points
1 ∶ 0
in reply to: matthes’s comment on: animal welfare has an evidence problem
I don’t want to be overly dismissive of WFI here. They are trying to do something really difficult and important. They have spent years on this work.

However, I think many have been overly confident in acting on their conclusions when they themselves highlight a need for more research.

matthes 7 Jun 2026 10:52 UTC
9 points
0 ∶ 0
in reply to: Vasco Grilo🔸’s comment on: animal welfare has an evidence problem
Thanks for engaging!

Re: cage-free welfare estimates

As I say in the post, I think that the WFI analysis does not include some potentially significant harms (such as chronic stress/pain from violence or parasites—which are likely higher in cage-free systems).
I also think that it’s not obvious how to score pain from behavioural deprivation, which accounts for a majority of the difference:
More generally, I think we currently just do not have enough data to make strong claims around the total cumulative pain chickens experience. Under “research gaps”, WFI itself states the following:
Surprisingly little research has been dedicated to the understanding of the impact of different welfare challenges at the individual level (where suffering occurs). Little is known about clinical evolution of the various welfare issues affecting commercial layers (e.g. healing times, duration of exposure), the likelihood of different clinical outcomes, rates of recurrence, and how adversely welfare harms are perceived by the individuals affected. Similarly, knowledge on case-fatality rates, comorbidity patterns and the prevalence of different conditions over the laying cycle is scant.
The independent assessment that Rethink Priorities organised only looked at two layer hen harms: peritonitis in conventional housing; and fractures during depopulation and transport in conventional housing. Neither was related to behavioural deprivation.
Additionally:
Raters were restricted to the references that Welfare Footprint Institute cited in the chapter where the original estimates appeared. Therefore, disagreement among raters is due to how different scientists use the same evidence, rather than which evidence they examine in the first place.
Re: shrimp stunning
The review you link relies heavily on data from other species. My post addresses all the published evidence for Whiteleg shrimp.

matthes 5 Jun 2026 22:05 UTC
34 points
4 ∶ 0
on: animal welfare has an evidence problem

matthes 22 Mar 2026 16:49 UTC
1 point
0 ∶ 0
in reply to: _Madeleine’s comment on: Open thread: 2026 Q1 (January—March)
Hi Madeleine, great to have you here! If you want to chat to other engineers, consider checking out the High Impact Engineers Forum: https://tinyurl.com/ImpactForge

matthes 17 Feb 2026 19:20 UTC
1 point
0 ∶ 0
in reply to: matthes’s comment on: Where should we hold EAGxUK 2026 (September)? Respond by 20 Feb
I also expect that people voting for Oxford and Cambridge will be more likely to apply and get accepted to EAG London. So they may benefit less from an EAGx than others.

matthes 17 Feb 2026 19:15 UTC
3 points
0 ∶ 0
on: Where should we hold EAGxUK 2026 (September)? Respond by 20 Feb
I am strongly in favour of Manchester for a conference. Oxford and Cambridge are quite annoying to get to from almost anywhere except London. People who haven’t been before might not realise this.
I also worry that this form will mostly reach people in the south of England, since they will be more likely to be signed up to the EA UK newsletter as it has historically mainly catered to London.
I will try to circulate this form with groups and individuals in the North of England, Wales, and Scotland before the deadline.
Also, Oxford and Cambridge are likely going to be more expensive than Manchester, where we could maybe even get a free venue.

matthes 12 Nov 2025 17:13 UTC
12 points
0 ∶ 0
in reply to: Dave Banerjee 🔸’s comment on: The Protein Problem
Yes, this table does not include data on this. I don’t have columns for values related to that. It’s a lot of work to track these numbers down and for many foods they are just not available.
When people online talk about the “bioavailability” of protein sources, they seem to mean one (or both) of two things:
One is the digestibility. So how much of the amino acids end up absorbed your body. That number is ~90%+ for most soy products (that I have found numbers for) as well as vital wheat gluten.
The other thing that people talk about is the amino acid profile. As one would expect, an animal muscle has the amino acids that one needs to build animal muscle, in about the right relative amounts.
Soy also has all essential amino acids, the profile is just a bit different. But none is particularly low or missing.
Vital wheat gluten is low in lysine. So I wouldn’t recommend relying only on seitan for protein. But if there is a bit of soy or some beans in your diet, I wouldn’t worry about it. (It’s also easy and cheap to supplement but that’s probably usually not necessary.)
Protein quality scores like PDCAAS and DIAAS try to account for both. That’s why, on its own, vital wheat gluten ends up with a poor score. But soy products tend to still score highly.
Basically, I am not worried about it and have personally been building muscle without problem on a vegan diet with lots of soy and some seitan. But you are right that accounting for this would probably change the table ordering a little.

matthes 10 Nov 2025 20:58 UTC
3 points
0 ∶ 0
in reply to: Catherine Low🔸’s comment on: The Protein Problem
Thank you! My base recipe is just:
- mix vital wheat gluten with some nutritional yeast (if I have any) and a bit of soy sauce
- add water until it comes together in a dough that I can fold in on itself until there are no more crumbs/flour
- rest for 15 min
- rip out (table tennis ball sized) pieces, pull them long, knot them
- boil the knots in vegetable broth (maybe 10-15 min)
- then cut into slices
I make these in bulk and find that the slices freeze well. I find that they work best in curries (Thai and Indian style) and chillies (especially if you cook them in a pressure cooker with all the beans and spices).
I also just snack on them on their own, but that’s maybe an acquired taste, like raw tofu or tempeh. (I also sometimes make my own tempeh. It’s quite fun! Making tofu turned out to have a really low yield, so I stopped doing that.)
These will turn out quite chewy. Some people don’t love that. To make them less dense, I have in the past added baking powder which makes them puff up when boiling—you’ll need a big pot for that!
Currently working on seitan sausage recipes, but not happy with any of those yet. If anyone has recommendations, please let me know!

matthes 9 Nov 2025 23:12 UTC
18 points
0 ∶ 0
on: Lack of Diversity Within EA
[Edit: This post had negative karma when I made this comment]
I’m sorry that your first post on the forum is getting a bit of a negative reaction. It’s great to have you here and I hope this isn’t super off-putting!
If you are interested in some actual numbers on demographics, check out the EA Survey posts, like this one from 2024.
My thoughts on your questions:
I am curious on the communities thoughts on this lack of diversity.
I personally don’t mind much. Almost any community will have demographic weirdness of some kind, and I don’t inherently value diversity. I know others find it concerning, but I am not aware on survey data on this. You can definitely find quite a bit of writing on this topic if you search the forum.
if there are any initiatives being taken to improve diversity within the EA community
I am unsure what you mean by “improve”. But most EAG events I have been to have meet-ups for various demographic groups. I have never attended these (despite “qualifying” for a few). But I know others really like organising and attending these. The most recent EAG London had meet-ups for the following: neurodiversity, people of colour, Christians, women and NB people, people from low and middle income countries, Jews, socioeconomic diversity, LGBTQ+.
(You may be getting downvoted for the line “This demographic has historically been disconnected from social impact”. Perhaps consider elaborating why you think that.)

matthes 6 Nov 2025 22:08 UTC
9 points
0 ∶ 0
in reply to: matthes’s comment on: The Protein Problem
Here are the top picks sorted by protein per kcal and price (based on Feb 2024 prices).
This is why I make my own seitan.

matthes 6 Nov 2025 22:01 UTC
1 point
0 ∶ 0
in reply to: matthes’s comment on: The Protein Problem
sadly, KoRo no longer delivers to the UK

matthes 6 Nov 2025 21:59 UTC
19 points
0 ∶ 0
on: The Protein Problem
For (potential) vegans that are depressed at the sight of the protein/kcal table. Here’s a screenshot of a similar table I put together last year (for myself—UK based). I like sharing this with gym bros that scoff at veganism for lean protein reasons.
Beware that it has a small number of duplicates as I was also tracking prices from different suppliers to compare.
Most of the things on this list have one or two ingredients if that’s something you care about.

matthes 28 Oct 2025 7:47 UTC
1 point
0 ∶ 0
in reply to: ethai’s comment on: High Impact Engineers is back
Thanks for the feedback! Maybe I overestimate how common GitHub use is.
Having said that, GitHub Discussions as a forum space requires no knowledge of git or coding. I think the interface is even simpler than the EA forum.
Contributing to the website is definitely going to be more accessible to people with some software experience. Although I am totally willing to onboard people with no prior experience if they’re excited to learn!
I’ll look into making the contribution guidelines on the website sound more welcoming.

matthes 27 Oct 2025 22:03 UTC
1 point
0 ∶ 0
in reply to: Peter Drotos 🔸’s comment on: High Impact Engineers is back
I’m open to collab suggestions! Did you have a particular idea in mind?

matthes 27 Oct 2025 22:02 UTC
5 points
0 ∶ 0
in reply to: Peter Drotos 🔸’s comment on: High Impact Engineers is back
Thanks! Some reasons I chose GitHub Discussions:
- threads stay up and visible forever for free (Slack is almost $10 per user and month if you want to keep post history beyond 90 days)
- there is an upvote feature making browsing more interesting because you can sort posts (once there is lots of content)
- people can close discussions when their post is resolved which declutters the default main page (for simple question + answer posts)
- no user management necessary
  - if I made a Slack joining link public I open the doors to bots—GitHub is natively pretty great at spam/bot prevention (at least so far)
  - if I made a Slack joining link private I would have to spend some of my time going through applications
- anyone can easily check it out without having to set anything up
- everything related to High Impact Engineers is on GitHub, making it easy to onboard new contributors/maintainers/admins which is great for a volunteer-run org
- many engineers have a GitHub account already and are familiar with the interface
Slack seems expensive and/or more appropriate for real-time conversation.

I do appreciate that this is a bit unusual, but it seems like the perfect platform to me!