Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

I’ve joined Anthropic

twitter.com

Post reply on HN
Print view View original

481–490 of 673 posts

Re: I’ve joined Anthropic

#481

Post by xp84 »

CamperBob2 wrote: post #377
dwa3592 wrote: post #167

Earlier quoted context omitted.

yeah but i don't think there's any large org which is 'good guys'. Anyone who wants to become a monopoly or very very large is probably suffering from some sort of a neural condition (psychosis, if plural) which we will study 100 years from now. Right now they are rewarded but I think our little minds forget to take the negative externalities into account. I am working on a short story on this topic which is set in 2…

yeah but i don't think there's any large org which is 'good guys'. There are several. They're in China, releasing competitive open-weight models on a regular basis.

Ah yes, the famous altruistic China, no profit or geopolitical/national security motives, doing it all purely for the love of the game.

Re: I’ve joined Anthropic

#482

Post by nikcub »

xmcp123 wrote: post #224
nashadelic wrote: post #165

Earlier quoted context omitted.

they invented MCP, Skills, made these standards open so anyone could build the harness around them.

MCP is barely an invention. It's a fuzzy spec detailing a pretty obvious design pattern.

I'm trying to think of a standard this doesn't apply to

Re: I’ve joined Anthropic

#483

Post by politician »

bugufu8f83 wrote: post #450
scottyah wrote: post #120

Earlier quoted context omitted.

Anthropic has drawn lines with the most powerful organization in the world, that OpenAI capitulated on within hours for a small contract.

Their statement on this issue opened by emphasizing how eager they are to help kill people: >I believe deeply in the existential importance of using AI to defend the United States and other democracies, and to defeat our autocratic adversaries. There is no universe where this can be described as anything close to ethical.

"ethical" is not a word that carries the connotation of a universally agreed upon set of behaviors. Different peoples, groups, and cultures vary in what they consider acceptable behavior.

Re: I’ve joined Anthropic

#484

Post by snakajima »

Karpathy is so valuable as an educator to the whole world, and I wish he continue to communicate with us after joining Anthropic. Just concepts like "Software 2.0" and "Vibe Cording" are priceless!

Re: I’ve joined Anthropic

#485

Post by HarHarVeryFunny »

ricardobeat wrote: post #464
HarHarVeryFunny wrote: post #431

Earlier quoted context omitted.

The auto labeling work (which has been partially described/presented at Tesla AI day events) seems more like engineering than research, a grab bag of techniques that I would guess the whole team must have contributed to. For example, they auto label low resolution/indeterminate objects (image segments) by temporal continuity... Something that is a low-res blob in the distance becomes a hi-res and easy to identify obj…

Not back in 2016.

You don't think that tracking objects from frame to frame is obvious ?!

I can guarantee you this was built-in from day #1

I'm guessing you're not a developer if you don't then automatically think of end cases like "what if car # 1 isn't in the preceding frame" ... (then you look at some relevant test data and see it was there, unlabelled ...)

Re: I’ve joined Anthropic

#486

Post by solenoid0937 »

siren2026 wrote: post #442
solenoid0937 wrote: post #441

Earlier quoted context omitted.

You should chat with some people involved in AI safety, if you really think it's a farce.

Imagine being gullible enough to think any of those companies would ever chose AI safety over losing their monopolies in AI. AI safety is important. My point is: you should have zero trust in those companies to actually care about AI Safety besides the marketing and PR aspect of it. Incentives matter.

Dunno, Anthropic delayed Mythos and refused to break their red lines for the DoW. But you seem to harbor this irrational hatred for them and the AI safety crowd so I'm not sure this discussion has much value.

Re: I’ve joined Anthropic

#487

Post by wodenokoto »

the_arun wrote: post #372
meetpateltech wrote: post #210

Karpathy will start this week on Anthropic's pre-training team, which is responsible for the massive training runs that give Claude its core knowledge and capabilities, according to Anthropic. Source: https://www.axios.com/2026/05/19/anthropic-openai-karpathy-a...

This is good branding move for Anthropic. Karpathy is well respected among ML crowd.

Speaking of, how did he not lose credibility at “full self driving next year, better buy it now”-Tesla?

It might be Elon who went and said that and said they don’t need lidar, but as director of AI and auto vision Karpathy bears the responsibility for those features.

Re: I’ve joined Anthropic

#488

Post by solenoid0937 »

siren2026 wrote: post #445
solenoid0937 wrote: post #443

Earlier quoted context omitted.

They didn't backpedal at all, you're spreading FUD. Their red lines are still in place. They are the only AI company with those red lines.

Ah yes, right [1] https://www.obsolete.pub/p/exclusive-anthropic-is-quietly-ba... [2] https://edition.cnn.com/2026/02/25/tech/anthropic-safety-pol... [3] https://www.wsj.com/tech/ai/anthropic-dials-back-ai-safety-c...

You didn't read your own sources. The red lines are still in place.

Re: I’ve joined Anthropic

#489

Post by stingraycharles »

ollin wrote: post #246
meetpateltech wrote: post #210

Karpathy will start this week on Anthropic's pre-training team, which is responsible for the massive training runs that give Claude its core knowledge and capabilities, according to Anthropic. Source: https://www.axios.com/2026/05/19/anthropic-openai-karpathy-a...

Specifically it looks like he's planning to extend the ideas from https://github.com/karpathy/autoresearch into a larger effort towards recursive training improvement [1]: > Excited to welcome Andrej to the Pretraining team! He'll be building a team focused on using Claude to accelerate pretraining research itself. I can’t think of anyone better suited to do it — looking forward to what we build together! [1] https:/…

Am I the only one who wasn’t particularly impressed by AutoResearch? If you looked at what the agent was actually doing, it was just tuning parameters mostly, not really trying different novel approaches.

I couldn’t help myself but consider this mostly a very inefficient variant of hyperparameter optimization, but someone correct me if I’m wrong, I may be looking at this too pessimistic.

Re: I’ve joined Anthropic

#490

Post by 4ashz »

raincole wrote: post #230
nightski wrote: post #228

Anyone else fearing Anthropic more and more each day? Not from a perspective that they are doing so well, but rather that it's like an industry tornado, sucking up and destroying everything in it's path.

Name three things they destroyed?

A girl school in Iran (potentially, together with Maven/Palantir).
Post reply on HN