Twin and adoption studies consistently show that parenting choices have minimal effects on a kid's eventual intelligence, personality, or happiness (except in cases of extreme neglect or abuse). This should revolutionize how we raise children, yet almost nobody knows or cares.
Matthew Barnett
10.4K posts
- A new LLM truthfulness benchmark just dropped. (Context: Alabama in fact has a higher per capita GDP than Japan.)
- There's been a lot of low quality GPT-4 speculation recently. So, here's a relatively informed GPT-4 speculation thread from an outsider who still doesn't know that much. 🧵
- If you give people cash and they choose to spend it on a bunch of junk rather than on education or healthcare, one way to interpret that result is that marginal spending on education and healthcare is worth less than a bunch of junk.The "cash can replace" a strong social safety net people taking a real L this morning.
- I have no dislike for philosophers, but the profession did not prepare us well for AI. The field is full of muddled thinking, abysmal takes like the Chinese Room Argument, a focus on pointless vague inquiries over big picture questions, and is often detached from actual AI.
- I wish people made their predictions falsifiable. Robin Hanson has been saying that the current AI boom will bust since at least 2016, but AI has rapidly gotten better over that entire time frame, with correspondingly more investment and attention. When can we say he was "wrong"?Replying to @robinhansonThe current burst of AI activity will likely fade, as have many bursts before, before the future burst when AI actually takes over the world. Something else will be the "big thing" between future AI bursts.
- I personally think $20 a month is cheap when the benefit is knowing whether a fundamental claim in your field is valid (and in this case, the claim is approximately not valid).
- Why do so many people think that humans don't trade with animals because we're way more powerful than them, rather than because they can't talk or keep agreements? Do they think cats, mice, and ants really couldn't do anything useful for us if we could coordinate with them?
- I currently think this open letter is quite bad, and possibly net harmful. The proposed policy appears vague and misguided. I want to explain some of my thoughts. 🧵
- Why are some people still treating AGI as a thing that will at some point "be invented"? At this point, doesn't it seem pretty clear that AIs will just get continuously more general and capable with no clear finish line?
- Most rankings of the top causes of death can be misleading, since they count someone dying at 20 the same as someone dying at 90. When you weight by life-years lost, you get this ranking (as of 2015). From ncbi.nlm.nih.gov/pmc/articles/P…
- How it feels to read stories about how an AGI can take over the world.
- It's frustrating when people say "AI progress is too fast" while over 100,000 people still die from aging per day, with no sign of abating. It's like we're in a huge, deadly war and people say our leaders are rushing to agree to a peace settlement. No, they should go even faster.
- Here's a line of reasoning for AI doom I've seen before that seems bad: 1. The first AGI will be able to end the world via nanotech 2. I can't explain exactly how it could do that 3. But (2) doesn't matter, because an AGI will be much smarter than me, and will figure it out















