I find it wild that @physical_int robotics models can work at all given that they only have a 1 second context length. As the robot is moving around a room and cleaning up a table, it only sees the current timestamp of images from the cameras (plus the text description of its task). I asked @svlevine was how this model can work at all. If a human had to execute a many minutes long plan in order to accomplish a physical task, but had her memory reset every second, there's no way she could pull it off But somehow just knowing what state the world is literally in this moment is all you need to know to keep a reasonably complication plan going? Sergey's answer calls on Morevac's Paradox, and I thought it was quite interesting. Full episode out tomorrow!

Casey Handmer argues AI pricing is so elastic that Anthropic could raise subscription prices 10x to fund new gas turbines. If that’s true, why is solar still the future? “The learning rate for natural gas is nowhere near as steep as solar… there are very few manufactured products easier to make than solar panels.”

Casey Handmer argues that investment in natural gas is doomed, even if there were no near-term geopolitical or financial risks from an AI bubble or China invading Taiwan. The fundamental problem: gas turbines need 20 years to pay back, but solar and batteries have already won on price. You're financing infrastructure that's economically obsolete before it's even built.

As measured by GDP, AI will be super undervalued. How would the datacenter of geniuses show up in GDP? GDP would show raw inputs (aka chip & energy), and raw outputs (aka cost of tokens). But wouldn't clearly reflect the value of the crazy new shit that's being cooked up in those tokens. Similar problem to how the Internet's value is undercounted today (since many products are free, and thus contribute nothing to measured GDP). Perhaps the best way to measure the size of the future AI economy will just be our civilization's total energy use. Full episode with @CJHandmer out in a couple hours! (Credit to @jekbradbury and @gwern for mentioning this idea in conversation.)

Ruthlessly optimizing efficiency created the factory farming problem. But in some cases, efficiency gains from tech can have a positive impact on animal welfare. Here Lewis Bollard describes how in-ovo sexing "went from 10 years ago just being a vague idea to today, it's already a third of the European egg industry."

The largest meat producers exploit climate rhetoric as cover for avoiding reforms. One of the biggest knows from its own surveys that consumers care more about animal welfare than climate. So when asked to adopt higher-welfare breeds or give animals more room, meat producers argue this would raise the carbon footprint.

People don’t appreciate how effective corporate campaigns have been on animal welfare. As opposed to policy change (often gets gridlocked by the meat industry) or moonshot cultivated meat tech (might be decades away from being competitive). This is a pro-market story! We can improve the lives of billions of animals by convincing food companies that consumers would prefer them to use meat from farms with marginally more expensive but far lower suffering practices. @Lewis_Bollard: US has gone from less than 10% cage-free to 47% cage-free. This transition has already spared over 200 million hens A YEAR from these battery cages.”

We're just getting started Lewis!

It's hard to grasp the scale of factory farming. @Lewis_Bollard described what he saw at one farm with 200,000 hens in battery cages (each hen crammed together with 4-6 other hens in a space the size of a microwave oven for years on end, unable to even flap their wings). Consider an alternative civilization where factory farming didn't exist. If we had just discovered that there was one farm treating 200,000 animals this way, it would be a moral emergency. But there are about 8 billion caged hens alive globally at any moment. Which means this one farm is literally just 1/40,000th of the suffering. And that's just for egg production. The whole episode isn't so dire - it discusses the science and economics of the meat industry. But I wanted to highlight this point because it is so easy to flinch away, and not notice how big and important this issue really is.

When an industry optimizes for one metric, it often breaks everything else. In factory farming, this creates compounding suffering. Lewis Bollard explains how trying to solve pig boredom opens a spiral of unintended consequences. “At each step, there is a new solution that can’t solve the fundamental underlying parts of the problem. And sometimes just makes it worse.”

I donated $2,000 to @farmkind_giving and got @dwarkesh_sp to match my donation! you should consider donating too and help him get to $250k!

Lewis Bollard explains Richard Dawkins' counterintuitive theory: intelligence might actually protect us from the worst of suffering. "Less intelligent animals might feel pain and fear more acutely than more intelligent animals. And that's because they can't rationalize it. They can't tell a story around it... It is just unmitigated pain. And they don't know when it's going to end."

Ending factory farming isn't about making everyone vegetarian. The animal welfare movement made the same mistake as early environmentalists - turning a systems problem into a personal virtue test. No major social reform succeeded by first converting millions to change their habits. They changed the defaults - subsidies, procurement, regulations. You could have more impact on animal welfare while eating a burger than a vegan who believes they are doing their part. Full episode with Lewis Bollard out now!

The U.S. meat lobby spends around $45M per election cycle. Far from tech or finance money. And yet, no meaningful animal welfare bill gets through without them stopping it. Part of the answer to their success is that they are never fighting alone. Agriculture, insurance, and feed suppliers all have a stake in protecting factory farming and close ranks when it is threatened. The bigger advantage is narrative. They draw on the American archetype of the hardworking mom and pop farmers. In reality, 99% of these operations are industrial, not pastoral. It is the “children’s book rule” of politics: never mess with cops, doctors, or farmers.

Just $1 can help avert 10 years of farmed animal suffering. I decided to give $250,000 as a donation match to @farmkind_giving after learning about the outsized opportunities to help. FarmKind directs your contributions to the most effective charities in this area. Please consider contributing, even if it’s a small amount. Together, we can double each other's impact and give a total of $500,000. Use the link below to donate with my match. Bluntly, there are some listeners who are in a position to give much more. Given how neglected this topic is, one such person could singlehandedly change the game for 10s of billions of animals. If you’re considering donating $50k or more, please reach out directly to @Lewis_Bollard and his team by DMing him, or emailing andres@openphilanthropy.org

Before I interviewed @Lewis_Bollard, I had assumed that factory farming was on its way out (especially given new tech like cultivated meat around the corner). Unfortunately this is far from inevitable: factory farms are already incredibly efficient machines for making meat (the most efficient broiler chickens convert 1.38 kg of grain into an astonishing 1 kg of flesh). I'd previously assumed that cultivated meat will soon trounce factory farming just on raw economics. Growing meat around a whole creature and mind cannot be the most efficient way to produce tasty flesh, right? Lewis thinks we may be many decades away (at least) from this outcome. Evolution has spent on the order of tens of millions of years optimizing human intelligence. And in order to try replicating this feat, AGI labs have to spend hundreds of billions of dollars a year. Evolution has spent far longer than that (basically the entire time) figuring out how to convert food into meat efficiently. Of course, tech on farms historically has favored more suffering (think gestation crates, battery cages, and overgrown broiler chickens). Improving the conditions on factory farms also requires corporate commitments and regulations against the most cruel practices. Every year we're factory farming about 2% more land animals globally. On the default trajectory, the amount of raw suffering in the world is likely to keep increasing. But there are reasons to think this can change. New technologies like in-ovo sexing have already saved hundreds of millions of male chicks from gruesome fates, with the potential to save billions more. And corporate commitments to go cage-free have already spared well north of 500 million hens from torturous battery cages (again, with the potential to help tens of billions more). Full episode with @Lewis_Bollard out tomorrow.

I filmed a video version of my post 'Why I Don’t Think AGI Is Right Around The Corner' so I could show it to my YouTube audience. You can watch it too I guess. 0:00:00 - Continual Learning 0:08:06 - Computer Use 0:11:27 - Reasoning 0:12:45 - So what are my predictions?

All the interviewees from the @dwarkeshpodcast with an X account and the host @dwarkesh_sp in one X list. The following are the guests not on X. 🧵nitter.siddz.com/i/lists/19503395557578…

How have you not had Alan Turing on the pod yet??? @dwarkeshpodcast @dwarkesh_sp