Homie chill. I use Opus every day and I love it. I’m not saying it’s all hype, just that these companies are here to make money and that every advertisement should be taken with salt yeah?

Also maybe consider what this kind of visceral reaction indicates on a personal level :/

I am tired. So very tired. Of seeing "it's all a marketing ploy" shouted by flocks of stochastic parrots under every single fucking "concerning AI capability or behavior discovered" report.

Yes, of course, it's all hype and not at all a really fucking concerning thing happening in cybersecurity. Totally wouldn't enable actors that wouldn't be able to hire top of the line hackers normally to cause chaos. Totally wouldn't cause old poorly maintained systems to start crawling with exploitable bugs at an ever-greater rate. Wouldn't affect "open source/closed source software" security dynamics in weird and hard to predict ways.

And, of course, the only reason for a company like Anthropic to bring it to attention is: marketing! Because it's not something that should get a "brace for impact" otherwise! Because trying to figure out whether you're about to fuck up something major in advance is only something that's ever done by UTTER FOOLS and MARKETING PEOPLE!

I am tired.

I mean if it helps I support the move to not release mythos right off the bat yeah? That makes sense, treat new models like new vulnerabilities and give companies time to scan with them etc.

But you have to admit it does serve a savvy business purpose of creating a moat where one wasn’t by getting these tech companies on board and the threat does make for good marketing yeah?

Would it actually just straight up kill you to take things at face value for a total of five seconds?

There is NO shadow wizard marketing gang pulling the strings on every single thing those companies do! Ridiculous, right? But it's true.

Nonetheless, every time an AI company drops a "mechanistic interpretability" paper, or an "extreme AI behavior" paper, or a "concerning capability" report, or says something sensible like "we don't actually know if those systems are conscious or capable of suffering in any meaningful way", or does just about anything really? There's an entire horde of people shouting "it's MARKETING it's HYPE" materializing out of nowhere and drowning entire comment sections in this thought-substituting bullshit.

"There is NO shadow wizard marketing gang pulling the strings on every single thing those companies do"

Yes there is they are called marketing departments and it is their exact purpose.

Normalcy bias is human nature. If human nature bothers you then you're going to be annoyed all the time.

We have been hearing this since GPT-2. They’ve been crying wolf for too long, that’s on them (the model providers). That and the fact they never publish anything interesting around their claims. It’s the ffmpeg thing all over again (very old bug in a decoder for a format used by one game from the early 90’s sold as some major breakthrough).

Is Anthropic lying about model capabilities? If not, where is the overselling?

March 2025, Anthropic was claiming that 90% of code would be written by LLMs in three to six months, and "essentially all" code within twelve months. This was one week after closing a Series E round for $3.5 billion. When they began working on their Series F round for $13 billion. You shouldn't need more than that to understand what's going on here.

The Claude Code leak revealed that Anthropic runs Claude-operated bots on the internet. One should be very cautious in getting swept up in the fund-raising process if they are not seeing first-hand the fruition of all of the flattering claims being presented by strangers on the internet.

You're talking about marketing predictions and I'm talking about data presented in a whitepaper. They are not the same thing.

>March 2025, Anthropic was claiming that 90% of code would be written by LLMs in three to six months, and "essentially all" code within twelve months.

There's a pretty big difference between "We predict in X time frame our model will be capable of Y" and "Our model did Y."

This is like watching someone measure the size of an object and saying "I don't believe you because you guessed it was X before you pulled out your tape measure."