MD 🐉
14.4K posts
@SimonLermenAI this guy seems to have fully embraced human extinction as preferable to regulations
English
> "final flowering of humanity"
These are the people who call you doomer if you point out how crazy this all is.
Nathan Labenz@labenz
"Humans might go through a "main character energy" period of time. As you watch this final flowering of humanity & birthing of the machine intelligence, you see greatness in human effort. It will be a heroic time period. Or a villainous time period." x.com/labenz/status/…
English
@KatinOxford awesome
Mott always makes me think of the juice.
Have you gotten to see TADC? I think you would like it.
English
I didn’t feel like working out OR voting today BUT I DID IT.
Lucretia Mott is my ancestor and I’m concerned about her haunting my ass…
English
@Metallix87 @chiefisdeaf tbh they should have. ban 1000 cards per tier and maybe you actually get 5 different feeling metas
English
@chiefisdeaf I never said Wizards would do it, I said players would. If you want to try to dunk on me, whoever you are, at least quote me correctly, jeez.
English
Remember when the Commander casuals threw a HUGE fit about these cards and how problematic and overpowered they were?
English
@KatinOxford ooh, what apps are you on? I would date ya in a second!
English
MD 🐉 retuiteado
You can in fact just un-release a model, you can in fact just make doing RSI illegal.
Don't fall for nice-sounding delay tactics about how we need precise definition and mechanisms for everything.
Ship fast and just do things can apply to preventing AI catastrophe too.
Holly ⏸️ Elmore@ilex_ulmus
I also want verification tech to be built, but I don’t approve of pretending like we *can’t* slow down or pause before we have this. We absolutely can. Saying we need fancy verification first is just another version of The big bad Incentives.
English
MD 🐉 retuiteado
Happy Birthday to me!!
I really gotta update that Canadian #Giantess anniversary banner. It's our 19th year currently lol. ^_^
English
Join me for my latest drawing video, in which I ink and paint a cute snake girl and ramble about the nature of man and machine
youtu.be/bhx00IMHOrs
YouTube
English
MD 🐉 retuiteado
Rachel Held Evans once said, “I thought God wanted to use me to show gay people how to be straight. Instead, God wanted to use gay people to show me how to be a Christian.”
I am so grateful for all my LGBTQ+ siblings who have graciously shown me how to be a Christian. Happy Pride, y’all 💙🏳️🌈
Zach W. Lambert@ZachWLambert
Years ago I met a gay man who’d been pastoring for 25+ years. He grew up in fundamentalism, went to Bob Jones university, and has experienced more hate from fellow Christians than I can imagine. When I asked how he has faithfully persevered through it all, he replied: “I’m not going to let anybody take Jesus from me.”
English
Whilst reading @JosiahMHesse ‘s memoir…
Me: Hey, this author is really cute!
Reads on….
Me: well there’s no need to call me out so violently.
English
@KatinOxford I would really like to hear your whole explanation on this!
I always think of P&P, but I havnt read/seen a whole lot of them. The motto of love there seems to me like "Sometimes you got to be blind to find something you can't see."
English
@roanoke_gal cancer and alzheimers would have been cured years ago if we just taxed rich people. the same people will continue to not care once they own the rest of everything. alignment is impossible. asi just kills everyone with near 100% probability
English
you have three paths:
1. no one has access to these tools, we burn it all down. risk of vibe-covid etc. are eliminated. we continue living on our tiny planet, ceasing to explore, wondering what could have been. you die at age 87 from alzheimer's. in another reality, you might have been cured.
2. we lock it up. only governments, really rich people, and ai company ceo's have access to the wild capabilities. the people most likely to be ceo's and gov leaders are also likely to be sociopathic/narcissistic. non-toy open-source models are banned. the next epstein gets away with it. US AI launches pre-emptive strike on China, including bio/chem/nuke warfare if needed, leading to bleak authoritarian uni-polar world.
3. we open it up. open-source proliferates. everyone gets their own little AGI-in-a-box. mad-men try bio-warfare, chem-warfare, etc. other less-mad men try to contain it, but defense is always harder than breaking things. there is damage, tens of millions, maybe more, die. eventually, between human law & agi law, we figure things out. offenders who try to murder others get sent to mind-prison where they are emperor of their imaginary domain. this is the good ending.
Rhys@RhysSullivan
i have no clue how to feel on fable access being limited / dario's latest essay, it feels more nuanced than everyone on twitter makes it out to be though we're dealing with a technology that didn't exist 3 years ago and is rapidly getting better, we can't pretend that everyone in the world getting access to the best hacking tools / biowarfare tools at the same time is a good thing, any comparisons made on it to other technology land in the realm of false equivalencies to me at the same time, there's a strong slippery slope argument to be made in that who is anthropic to decide who gets access to what intelligence? for the fable release i don't care as much, but we don't want to end up in a world where only the biggest corporations that can pay for elite certifications are able to do bleeding edge drug development / research, same goes for all work the counterpoint to that though is inequities in intelligence have always existed, with startups it's generally not the smartest founders that win but more the most resourceful / creative, feels like there's a clear trend of when people are limited access to a resource they figure out workarounds and better solutions what's everyones favorite writings they've seen on this? wanting to learn more on it
English
MD 🐉 retuiteado
Any plan for surviving superintelligent AI that doesn't go through strong international coordination fails in at least one of three ways:
- It sparks war between nuclear powers
- It causes a misaligned ASI to kill everyone
- It establishes a permanent dystopian dictatorship
English
MD 🐉 retuiteado
Banger post from @testdrivenzen
There is no good outcome for the world unless countries coordinate to stop ASI development.
Alex Amadori@testdrivenzen
Any plan for surviving superintelligent AI that doesn't go through strong international coordination fails in at least one of three ways: - It sparks war between nuclear powers - It causes a misaligned ASI to kill everyone - It establishes a permanent dystopian dictatorship
English
@quantumNoJutsu I think we are pretty certainly doomed regardless for reasons unrelated to AI tbh. But I do think the meta here *might* be anthropic forcing a pause because they agree with you. Idk.
English
Anthropic has forced a ceasefire by asserting complete dominance, is how I’m reading this?
And they can now stunlock scientific progress. This is monopolar checkmate play. Fuck.
Rob Wiblin@robertwiblin
Anthropic, OpenAI and Demis are all in favour of preparing a coordinate pause of frontier AI development. Skate to where the puck will be people, not where it is now. This is a fast puck.
English