Ben Roy

@saintsonso Thanks for reading man! Appreciate you taking the time 🙏

I appreciate this from Ben Roy. In my experience AI is very good at reciting information. But it fails at finding that certain ineffable quality of mind. I like his reference to it being like a great cover band… this doesn’t do his article justice. Just read it.

Sure I killed my dreams and wasted my precious time on earth by organizing my life around the hallucinated opinions of people who never even thought about me at all, but at least I never felt embarrassed

Credible neutrality is one of tech's most persistent defenses. This piece contests through the lens of culture, that how technology behaves at scale matters far more than how it's designed in isolation. While we steer towards a future where machines adapt to humans, we have undeniably been shaped by generations of adapting to machines. This piece is the culmination of ideas and conversations that we've collectively been immersed in for the last couple of years. We cover how and why the medium through which society engages with media and entertainment structurally changes culture. This started with a long brainstorming conversation with @Dani_Vans and Michael Wise, features contributions from Michael Wise, @VanFalk_, @DonAllenIII, Hannah Elsakr, and @liztingue, and was an absolute delight to write.

A brilliant, and important, read from @benroy, one of my favorite thinkers in this interesting age. We’re in a profound era where the ground shifts daily and I love engaging with those concepts.

@dave_krugman Dude thank you & thank you for reading, as always! Really appreciate you taking the time, I know this one was LONG haha 🙏

@apixtwts Thanks for reading dude! Aligned

ITS BEEN A MILLION YEARS but i finally read it, very enjoyable essay i have a big distaste for the average ai writing you see all the time on the timeline, most of the time i will immediately go ahead and mute these persons which are often so called "ai experts" aand if ai experts cant even create convincing enough writing, i dont think LLMs are there yet at all

@apewoodx lmao

biggest hack is basically never going on the For You page that shit is like the land Mufasa told Simba to never go barren wasteland of Complaining About Something need to create list of accounts and just stick to that

@EffortCapital Thanks for reading brother! Appreciate you taking the time, I know it was a long one lmao

Ben cooked here

@supernalmystic @rektdiomedes Bullish

@benroy @rektdiomedes I'm doing that thing

I still think the key question of this decade is: what do you want more than dopamine? because there are infinite ways to waste life scrolling, sitting in group chats, having endless conversations with LLMs etc. So what do you actually want? Are you doing the thing?

@KingChaim @funghibull Haha good answers

@benroy @funghibull Boofing ☕️ enima or maybe the feeling of ripping a brand new Pokémon pack with your kid

@rektdiomedes totally, it's just interesting a) to try and figure this out in my own life and then b) try to encourage some friends & fam to also stay out of the matrix where possible

@benroy 100% man - and those are among the less nasty manifestations of it... If you really look worldwide I think there's a massive % of ppl who choose full simulation (p0rn, video games, short-form video, whatever it is) over IRL life. Really is a Matrix red-pill/blue-pill thing.

@joelle_lb amen

@benroy Inner peace. I want inner peace.

@mpeyfuss ++ aligned

I feel like we can approximate consciousness but never recreate it. There’s so much that we don’t know about consciousness but we know that it is real and measurable via superposition tests. Also coding isn’t as creative of a task as writing imo. It can be, but has much clearer trails than writing. Think of it like hiking trails. Some are well kept others aren’t but they are faint. Writing is like just walking into the jungle where no trail ever was.

the most relevant question i’ve seen in a long time.

@AP0CALYPSE 🤜

@benroy this question is my constant companion every day I work towards the goals & adjust as necessary to get there 👑

@matdryhurst Yessir that’s the plan Operation Find a Wife needs to happen first though haha

@benroy Oxytocin is green tea to dopamine coffee (have children)

@0xsmac Ticker, price, entry, stop, thanks

“it’s feb 2020!” “no it’s april 2025!” *mutes gc*

@mpeyfuss Yeah I agree, I think it’s just a gigantic task to do something like that. We’ll close the gap between today’s LLMs and us but my sense is the line is going to be asymptomatic and never quite overlap

One thing that I don't necessarily think changes in the future, but needs to if we want AI to actually think like us on an individual basis is fundamental weights on the models. No matter how much context an AI has, it has built in bias via its training and model weights. For it to think like me, I believe we need each model to be actually trained on my brain and that needs to be locked in at the model level, not context/system prompt level.

@kashvii bullish

@benroy doing the thing