Welfare tech should be developed by welfare people — EA Forum

Comments

Welfare tech should be developed by welfare people — EA Forum

10

Fwiw I think one of the main barriers to this work is having good engineers who want to work on this. If you are a great engineer, consider working at Amodo - they are hiring!

7

I strongly agree! I'll also note that it can be very difficult to start new companies developing tech in this space using the VC model for reasons I outline here - low margin industry, slow moving incumbents, and high capital costs for physical tech developments. Therefore, some corollaries to this are:

Consider working at some of the companies developing welfare tech such as Optimar, Baader, Orbem, HatchTech and many more. This will have high direct impact (as often these companies have a harder time attracting top-tier talent), as well as hugely useful learning potential.
If you're an impact investor interested in this kind of tech development, especially one that is comfortable with non-VC return structures, considering getting involved in the space. Currently, lack of interested capital means there aren't a ton of investment opportunities, but having more activity will help start this flywheel, especially if you're public about your investment thesis.

I'm happy to chat with folks that are interested in getting more involved here, and I'd encourage folks to reach out to me, or Aaron, or any of the other groups mentioned here like Amodo to better understand the landscape.

Marcus Abramovitch 🔸

4

Strongly upvoted. I think APRI is doing this at least in the sense of developing species that feel less pain. I think this is a promising direction

4

Thanks for this post. I thought it was really interesting and I think you are probably right.

My main objection - however - would be that technologies developed by welfarists would struggle to compete in the market as it stands because we'd be trying to meet the often competing demands of improving welfare and improving economic outcomes for farmers (in order to have our tech adopted in the first place).

Companies that develop technologies that are purely delivering efficiency gains for farms are more likely to succeed in a profit focused market than these welfarist technologies and so welfarist animal tech companies might be competed out of the market.

It might be more efficient to focus our attention on creating the kinds of market conditions that incentivise profit-making companies to develop the tech we want to see.

What this means is focusing on what we're already doing - corporate campaigning, lobbying etc. - to create the conditions where farms/food companies have to care about animal welfare to some extend and then let the profit-seeking (counterfactually cheaper) money be spent on R&D to deliver the tech that meets those objectives.

4

Cool post!

Misc thought that this seems analogous to some of the points/ideas/arguments in https://www.forethought.org/research/ai-tools-for-existential-security (albeit for different tech and to primarily address different problems)

2

Wholeheartedly agree, and I think that the principles apply more widely across a range of cause areas. The right people not being involved in design solutions can mean external people coming in with "clever" solutions can lead to Play Pump type problems, as well as the more general suboptimal "making the best of what's already out there" type issues you highlight.

I'm pretty new to EA, so maybe I've missed it, but I've not seen any discussion of User Centred Design in an EA context. UCD feels like an approach which helps to make sure that solutions are solving the right problems. Does EA need to embrace UCD more?

Simon Newstead 🔸

1

Thanks for sharing.

Reminds me a bit of the work Innovate Animal Ag are doing, not sure if the consulting part also covers the R&D / engineering or just go to market/commercialization.

It makes sense though, we've seen in other fields mission driven folks being the ones to develop all new technologies / try out their own ideas eg- climate tech, alt protein

Vasco Grilo🔸

0

Thanks for the post, Aaron! Relatedly, I estimate Optimar has been counterfactually responsible for electrically stunning 12.0 billion shrimp via selling 16 stunners to the Shrimp Welfare Project (SWP), which I calculate is as beneficial as the benefits to humans of saving 10.0 M children.

Entrepreneurship

This is a lightly-edited memo I wrote for the 2025 Animal Advocacy Strategy Forum, which were encouraged to be highly opinionated to generate strategy discussion.

It seems strange to me that animal advocates rely on animal ag to come up with the solutions that we want to see. We identify a problem, we look to see what exists in the world, we try to find the best of those things, and then we try to get them implemented.

I would like to see welfare-oriented engineers actively develop the tech that we want to see deployed on farms.

For example, Shrimp Welfare Project is trying to get new stunners developed, which I'm excited about ^[1]. But I think there are a lot of other examples that this would work for. In-Ovo sexing seems like another good example of technology that I'm kind of surprised that the animal movement wasn't more involved in the development of ^[2].

I'm also spending a lot of time nowadays looking into Precision Livestock Farming tech, which seems like a pretty interesting new Wild West of factory farming.
Rather than just looking at what's being developed, why don't we be the change we want to see in the world and actually design and build the Precision Welfare tech we want to see?

It also seems possible that the money for these developments could come from non-philanthropic sources.
For example, as part of McDonald’s eight Broiler Welfare Commitments in 2017, they committed to:

Innovative On-Farm Animal Health and Welfare Monitoring Technologies

Partner with technology companies, producers, and suppliers to develop on-farm monitoring systems to automate the gathering of key animal health and welfare indicators, including behavioral measures. Once established, these technologies will highlight potential areas for improvement in real time and will be among the first of their kind available at a commercial scale.

McDonald’s invested nearly $2m (and $2.6m of matching funding) into their (SMART) Broiler Initiative, to automate gathering their health and welfare indicators.

The projects that are continuing to receive funding through this program are:

OPTICFLOCK (Marian Dawkins) - The system automatically monitors the behaviors of broiler chicken flocks around the clock to deliver real-time information on key welfare indicators such as mortality, walking ability, leg health and infection. The tool uses inexpensive and commercially available closed-circuit television cameras linked to small on-farm computers.
FlockFocus (Niamh O’Connell) - This camera-based technology can automatically monitor and track activity patterns and the weight of individual birds within large flocks on commercial farms. It can also monitor feeding behavior and the distribution of birds in key areas of the chicken house.

Whatever people think of McDonald’s broiler commitment (and to what extent they are humanewashing) - I think these projects being funded seems pretty positive. And it seems like more projects like this could have been funded if we were ready for them ^[3].

To summarise, I think we should be actively trying to develop novel welfare technology that we want to see on the farms rather than expect it to be developed by the industry. And I also think this could be financed non-philanthropically.

I'm aware of some engineers in the EA space who have thought about this, such as Koen van Pelt and Amanda Matthes.
So if you're interested in developing welfare tech (or getting welfare tech developed!), I think Amanda would be a great person to get in touch with (Koen too, but he's just co-founded Scale Welfare, so may not have the capacity to engage as much).

^{^}
For those interested, we're working with Amodo Design on this.
^{^}
Though I think Innovate Animal Ag has done great work popularising this technology among producers.
^{^}
EU Horizon's calls for projects also seem promising sources of large counterfactual funding, such as: Towards modern, integrated, and effective fisheries monitoring, control and surveillance (MCS) systems