I didn't care how meat was made. Then I learned about the tech behind it.
Last week in the AI x Animals fellowship, I learnt about something called Precision Livestock Farming (PLF). It is basically what happens when you put a Fitbit on a cow and let an AI run the farm. Before this, I didn’t really pay much attention to how our meat is actually made, but the tech behind it is actually pretty wild.
Think of it as a stack of gadgets. You’ve got cameras watching how chickens walk, microphones listening to vocalizations, and sensors tracking how much water everyone is drinking. All that raw data goes into a machine learning model that looks for patterns. If a chicken starts blinking differently because it’s stressed, the AI spots it and sends an alert to the farmer’s dashboard. It is a way for one person to manage thousands of animals at once without ever having to step onto the farm floor.
This is the part that really clicked for me. Think about your own life: if someone gave you plenty of food and kept you in a room where you couldn't catch a cold, but you never saw the sun, never went for a walk, and never saw a friend, you’d be healthy on paper, but you’d be miserable.
Animals are the same. Chickens don't just want to walk around - they need to. Cows form friendships and get stressed when separated from them. Even fruit flies have been shown to experience something like depression, they can actually show learned helplessness and respond to human antidepressants.
The problem with current PLF is that it is great at measuring if an animal is getting fat, but it often ignores if that animal is lonely or bored. We are building smart farms that are genius at counting calories but can be totally blind to happiness.
The good version of this is called Precision Welfare. Since humans can't possibly watch 10,000 animals individually, the AI acts like a doctor that never sleeps. It can catch a disease days before it spreads, or even use sensors to stop a mother pig from accidentally crushing her piglets. If we build this tech the right way, using open-source code that anyone can check, it could bring a level of transparency to farming that we’ve never seen before.
The scary part is that AI is a multiplier for whatever goal you give it. If a big corporation builds the system just to make money, the AI will only care about things like weight gain and survival. It is like the Efficiency Trap: if the AI says the animals are healthy enough not to die, a farmer might use that as an excuse to pack them even tighter together. We risk turning sentient beings into just another data point on a spreadsheet, ignoring their need for things like sunshine or play just because happiness doesn't have a profit margin yet.
Right now, we are in a Wild West phase where the rules for this tech are being written. If we let the industry hide their data in black boxes, we lose the chance to hold them accountable. But if we push for Welfare-First tech, we might actually be able to use our coding and engineering skills to help the billions of animals that we usually ignore.
If you want to see the actual math and the debates behind this, you should definitely check out the original post by Max Taylor or the 80,000 Hours profile on factory farming.
Its really good post to get a understanding of PLF.