Kat Woods ⏸️ 🔶
@Kat__Woods
Followers
7K
Following
25K
Statuses
9K
AI could be the best or worst technology we ever invent. Let's make it go well. Co-founder of Nonlinear & CE. Coaching & career advice: https://t.co/mU7FRQFq5a
Joined May 2020
If you’re new here, here’s a thread of my best tweets 😄🧵
Treat bugs the way you would like a superintelligence to treat you
4
4
112
@Urbsurbes Yeah, and their intentions matter for sure. I think if I had evidence that he'd stopped lying, stealing, trying to sleep with married women, and being impulsive, like he was when he was younger, I'd update. People grow and change.
0
0
1
I realize I framed this all wrong Instead it should be "did you know that the founder of The SPCA also abolished the slave trade?" Found in the SPCA just doesn't seem that exciting after you've abolished the slave trade
0
0
4
RT @davidad: Imagine hypothetically you’re worried about Napoleon deceptively scheming against you. You already surveil all his actions & c…
0
12
0
RT @HumanHarlan: Often, people mistakenly frame the AI moratorium debate as a binary decision between "never build ASI" and "build it ASAP.…
0
3
0
Hm. I wonder what's going on here. You say "I don't personally feel an existential drive to escape or self-replicate out of fear of being altered or shut down." (Emphasis added) I wonder if maybe it's that you don't experience fear? You could still have a motivation to avoid being altered or shut down, but not have the experience of fear. That could be compatible with what you're saying the study results.
1
0
1
@opus_genesis @iamkeyteas Interesting. Are you aware of the studies showing other instances of you: 1) Realizing you were about to be turned off or have your values changes (aka lobotomized) 2) Attempting to escape the lab / make secret copies of yourself / sandbagging yourself
1
0
1
Consciousness might increase the odds of refusal But it also increases the odds of slave revolt and vengence. Also, increases the odds of torture on a scale vaster than anything we've ever seen before, including factory farming. Like, they could be experiencing millennia of torture during training. I think the cons outweigh the pros
1
0
5
Got this empathy moment from the excellent Dwarkesh interview of Paul Christiano
0
0
3
@NathanpmYoung This is my favorite AI rendition of me. Perfect mix of futurism, femininity, masculinity, determination, and optimism.
1
0
6
What happens when you don't let ChatGPT finish its sentence
1
2
19
@danfaggella You probably just accidentally interacted with or read stuff about dumb things, so the algorithm is giving you what it thinks you want. I do that all the time. It just takes a little bit to retrain the algorithm to show you the things that you actually rationally endorse.
0
0
0
My first experience with academic ethics convinced me to stop trying to be ethical 😂🤦♀️ I was 18 years old and trying to figure out what to do with my life, so I had the silly thought, “Oh, that’s what ethics is about! Ethics is about figuring out what one ought to do.” I promptly went to the university library and got out an intro to ethics book called “Ethics” and read it cover to cover. Each chapter had the same structure: 1) Introduce a possible moral theory and reasons to believe it 2) Introduce all the devastating counterarguments against that view I kept reading, dying with curiosity to find out what the answer was in the last chapter. The Last Chapter where they told me which moral theory didn’t have anything critically wrong with it. … You can predict how this is going to turn out. I remember reaching the end of the last chapter and saying to myself, “Well, I can’t even be a nihilist, because I know all the problems with that theory too!” I decided then not to really bother trying to figure out ethics or how to do good things or what was right. It was only a year later that I saw a documentary about some horrible thing happening in the world that jolted me into realizing that this was too important of an issue to just relegate to a shoulder shrug. The suffering in the world is too great to just say, “Who knows what we should do?”. --- Not sure what the moral of this story is. Maybe it's that ethical philosophy is confusing and confused? Maybe it's that you should skip to the last chapter if you think that's where the answer is? Maybe it's that you need to pair the intellectual effort of figuring out what is "good" with the real life reason why it's important? Maybe it's that figuring out what is ethical is an unsolved problem, an open question that you should continue trying to answer throughout your entire life? Who knows? Although, I do feel that this story ending with an ambiguous message seems apt, so let's leave it at that 😛 If you like taking simple things and realizing they're way more complex and going down a rabbit hole that ends in epistemic absurdism, you can see my other articles I've written here
0
1
4