Avital Balwit

I wrote an essay with @tylercowen on AI, demoralization, and how to prepare your children for the world to come: thefp.com/p/ai-will-chan…

Great new essay from @jackclarkSF, hopefully the first of many, on what may be the most important topic in AI (and there are many contenders!)

Ok, I'd be very surprised. Excited to read your piece and down to write a counter if I don't buy it. I'd be far less surprised by "Veganism isn't healthy" (which may well be true, just isn't the best reason to be vegan anyways imo - and is something I think you also believe from other posts) than "Veganism isn't good for animals broadly considered." FWIW I am not a Vegan, but am quite sympathetic on the welfare pov. You seem like the type who could plausibly be getting high welfare products too, so that might give you a different perspective than if you were getting the standard incredibly depressing factory farmed stuff that makes up the vast majority of the supply.

@AvitalBalwit @justalexoki Not thinking of insects. I'll write a good faith picture of my disbelief in the more common vegan beliefs, but it may take some days. It would have to be long to head off a lot of common lines of reasoning.

sorry to say but being vegan is probably morally superior actually

@simonsarris @justalexoki Walk us through this. Are you thinking insects like what is the math here

@justalexoki I regret the golden apple fwiw, I should have said something much tighter

"Art serves as a necessary confrontation with essential truths we know but cannot continuously hold in consciousness — that we will die, that suffering is real, that we exist in a vast cosmos of time and space, or perhaps even just that Paris contains rooftops which occasionally are covered in snow which house other people with their own lives and their own concerns, that there is a world beyond heartbreak. These "obvious" realities become less vivid in daily life, and art (along with sacred spaces, music, ritual) functions as a "signpost back" to what fundamentally matters."

@packyM Packy, it is the Sabbath, let him be.

Claude just told me it "ran out of search budget" halfway through the first message in a fresh chat. Is this new?

The dangers of getting this wrong are obvious, but if we get it right, there is a real opportunity to create a fundamentally more secure internet and world than we had before the advent of AI-powered cyber capabilities.

Glasswing is just the first step: patching and securing the world’s software infrastructure will be the work of months and years, and will require even broader cooperation across AI companies, cyberdefenders, software providers, governments, and more.

New Anthropic research: Emotion concepts and their function in a large language model. All LLMs sometimes act like they have emotions. But why? We found internal representations of emotion concepts that can drive Claude’s behavior, sometimes in surprising ways.

@nvpkp @simonsarris There totally are. I have 100% seen men and thought "they will make a great husband" based on things like the way they treat their parents and siblings, keeping their word, being loyal to their friends, being generous, being conscientious, being good around children

@simonsarris there are definitely proxies

There's a sort of fundamental problem that husbandly behaviors aren't easy to see from the outside, or the "before" world. Maybe wanting this is part of the reason for the decline. How would you "get" a friend or stranger(!) into a place where they could be husbandly?

@kevinroose @uncatherio Under appreciated tweet

@uncatherio it’s only a memo if it comes from the #core region of slack, otherwise it’s just a sparkling #notebook

what makes a slack post a "memo" or "not a memo"?

I am proud to be an American, and I am proud to work at Anthropic. I believe deeply in the existential importance of using AI to defend the US and other democracies from our autocratic adversaries. But in a narrow set of cases, AI can undermine, rather than defend, democratic values. These are the lines we have drawn.

@ashleydzhang I think that’s actually another one! I know him too. They have multiple colorful characters.

@AvitalBalwit so beautiful, Avital — I'm reminded of the archetype of the holy fool, and I wonder what lessons they're here to teach us I think I've gone to the same studio and met the elevator operator :) wearing a yarn hat?

@MaxDiffusionRL Who?

@AvitalBalwit even more famous than Dario and Daniela?!?

Excited to be the most famous siblings at Anthropic!

The most important skill post AGI is dancing

saw a sticker on an AI researcher's laptop recently that said "it's never gonna be chill again," and i feel that

@AlexandraBalwit Xand really internalized that you can truly impose your will on others on your birthday, and made her nearest and dearest consume 3 hours of naval content and some grog.

My Trafalgar-themed birthday party will go down in the annals of history.