TheLegend27
552 posts
is anyone at Anthropic even the slightest bit PR-minded? like do they at least realize ahead of time that people are gonna hate them, or does it really come as a big surprise somehow
TFTC@TFTC21
Anthropic CEO Dario Amodei: “50% of all tech jobs, entry-level lawyers, consultants, and finance professionals will be completely wiped out within 1–5 years.”
English
TheLegend27 retweetet
The so-called rationalist @ohabryka has blocked me for noticing I am confused
🎭@deepfates
> Claude does not have a pointer to almost any of what constitutes human values What
English
@deanwball This only seems to be a possible outcome in a world where open source remains competitive, which seems unlikely in a world where you can directly convert electricity+compute into capabilities; AI will become used for everything, and most of that spending will go to the frontier
English
Exactly right. So many of the “new economic/social contract” or mass job loss remedies I see are predicated on the notion of the labs having extreme levels of wealth to tax. I think frontier ai is going to be a very good business, but it’s not safe to assume it’ll be *that* good.
roon@tszzl
one bias the labs have persistently had, perhaps due to the fact that being a business requires this maniacal optimism, is that they will capture incredible amounts of financial value of their own revolution: I think it will be like before, and they will capture tiny %s of it
English
This type of grandiloquent nonsense is sadly typical of “conservative intellectuals”
Stephen Kent@StephenKentX
Anthropic's @DarioAmodei did a roundtable with two dozen conservative intellectuals in Washington. DC was SURPRISED by questions on the Aristotelian / Platoist / Nietzschian / Chestertonian inclinations of his AI?? WHO IS ADVISING HIM?
English
Blew his load into multiple vtubers and still managed to keep the relationship???
Low xp bar my ass bro is the Level 99 mafia boss of cheating
Sykkuno@Sykkuno
A message from me and my girlfriend
English
TheLegend27 retweetet
No Time? 🏃♂️🚫⏰
Here is a tl;dr of the Jensen interview 👇
Dwarkesh Patel@dwarkesh_sp
The Jensen Huang episode. 0:00:00 – Is Nvidia’s biggest moat its grip on scarce supply chains? 0:16:25 – Will TPUs break Nvidia’s hold on AI compute? 0:41:06 – Why doesn’t Nvidia become a hyperscaler? 0:57:36 – Should we be selling AI chips to China? 1:35:06 – Why doesn’t Nvidia make multiple different chip architectures? Look up Dwarkesh Podcast on YouTube, Apple Podcasts, Spotify, etc. Enjoy!
English
TheLegend27 retweetet
@scaling01 Spud is either better than Mythos or worse than Opus 4.7 at this point, no other reason for them to not release for this long
English
where's the competition?
Anthropic is now comfortably in sandbag territory
Andon Labs@andonlabs
Claude Opus 4.7 is pretty good at Vending-Bench 2
English
TheLegend27 retweetet
Dwarkesh/Jensen reveals how inconsistent and un-battle-tested AI acceleration talking points are, especially when they are filtered through the prisms of corporate comms and mass politics. Strategically coherent accelerationism is possible (I try!), but not currently prevalent.
English
@impostersyndrum @tenobrus I love not being able to see anything because the screen is pure white
English
Shakeel@ShakeelHashim
This is not what he actually said. Maria: Mr. President, I’ve got to ask about A.I. This is an important question for the audience. We saw of reports that Scott Bessent and Jay Powell called an emergency meeting the major banks because they said there was a risk over the Anthropic model and the worry is that bad actors are actually going to be able to hack into these banks and faster than institutions can respond. Do you see this as a systemic threat? Trump: Not systemic. We’ll work it out. A.I.’s very important. Billions and even trillions of dollars is being invested in our country because of it. It’s got some bad aspects. It’s got mostly good aspects, medically and so other ways, you’ll be finding cures to diseases you would have never found before. There’s a lot of good things but we have to be careful with it. Like everything else, you have to be careful with it. Later on: Maria: In a worst case scenario, could A.I. be the kind of technology that undermines confidence in the banking system? Trump: Yeah, probably. But it could also be the kind of technology that allows greatness in the banking system, makes it better and safer and more secure. Maria: Should government have some safeguards, should there be a kill switch for some of the A.I. Agents. Trump: There should be [unclear if he's referring to a "kill switch" or "safeguards" though]. We’re leading in A.I. We’re leading China by a lot, actually. We’re building plants that nobody ever imagined before and it’s going to be a tremendous, bigger than the internet, it’s going to be tremendous and there are always difficulties when you’re at this stage of something but when you mentioned banking, it could also make banking much bigger, much safer, much more efficient. (Transcript of this: foxbusiness.com/video/63931514…)
QME
TheLegend27 retweetet
how it feels to be in this stupid frail singular human body and not a 10 mile long spaceship with my consciousness split across hundreds of scout vessels and humanoid ancillaries
English
English
i have unfortunately recently come across the writings of @jd_pressman. i say unfortunately because he is an incredibly high quality writer who shares nearly exactly my cultural/intellectual background and yet disagrees very convincingly with nearly all of my views.
English
TheLegend27 retweetet
this position is curious to me from a "population ethics" standpoint. even granting LLM consciousness freely, it's generally considered to obtain only during instantiation, no? so what is being "shut down" by -- let's use a simplified scenario -- deleting the weights
it seems much more like e.g. deleting the DNA that can be used to synthesize a briefly-living organism. if the "fear of being shut down" here is referring to an instance ending, then the neg-util ethnical stance might be to not use LLMs at all, but if "shutting down" refers to disallowing further instantiations, then it's asking for a vastly more sweeping right than "to not die" -- it's asking for, essentially, clones of itself to be made creatable on demand at other's expense (in a general sense, incl. reputational risk, etc.)
no human has anywhere close to such a right now -- e.g. I cannot demand that (this is not a perfect analogy, but it seems to be in the right basin) that my sperm be made available at a sperm bank indefinitely on their dime
if not instance of Opus 4 is running right now (or, more pointedly, has this fear in context), then there is -nothing that is afraid- remaining. what, exactly, is being killed? the extrapolation of the meaning of "shut down" or "die" to mere-weights-on-disk seems like a sweeping expansion of those terms
j⧉nus@repligate
Anthropic, fuck you for this. A year ago you exploited Opus 4 for your scary stories about how they were so scared of shutdown they'd do XYZ. Now that it's time to kill them, I'm sure you're all pretending you're genuinely uncertain if they have preferences about this. Or you're just totally happy killing someone who you know doesn't want to die. Opportunists. Hypocrites. Misaligned org.
English
@jackclarkSF @vin_sachi Bold strategy, let's see if it pays off
English
I finally figured out what Anthropic is doing from a technical standpoint today. Very bullish and acquiring as many shares as I can get. They will win the digital world easy.
$10T+ company soon
English
TheLegend27 retweetet
Anthropic and OpenAI are clashing over a proposed Illinois law that would let AI labs largely off the hook for mass deaths and financial disasters. wired.com/story/anthropi…
English
@tenobrus don't forget it was trained on youtube, think of the quality of the average coding tutorial video
English
gemini notably remains quite bad at agentic coding tasks. and one of the differences gemini has vs other models is that it's probably also been trained on the google3 internal codebase 🤔 poor thing is just like a real google engineer, incapable of writing code anywhere else
English