Raw LLM Responses

She forgot to mention The FTX connection. Anthropic story has a chapter that’s well a lot darker a lot more more troubling and and it really sets them apart. Back in 2022 they had this massive series B funding round, we’re talking over $500M and the person leading that charge well it’s a name that was about to become pretty infamous. Let’s just be super clear about this. The ““ ethical AI company. Their big growth spurt was fuelled by half $1 billion dollars from Sam Bankman Freed poured into Anthropic through his doomed crypto exchange FTX, so let’s get this straight the foundation of the so-called ethical AI company was built with money from one of the biggest Finacial frauds seen in years you could call that their original sin the company that promised us they’d be more responsible was literally built on money from one of the biggest Finacial frauds of our time. Just let that sink in for a second okay? So, money is, let’s just say it’s complicated but what about their core promise? What about the product itself? Is it really safer? More responsible? Alright? First we need to understand that their big idea their flagship tech is called constitutional AI. Instead instead of having humans constantly correcting the AI. They basically give it a constitution a rulebook of ethical principles to live by the whole goal is to make it helpful, honest and harmless right from the start. it’s the technical backbone of their entire brand sounds good in theory but does it actually work? How do you square the word harmless with this number? 1.5B. This is the $1.5 billion in damage sought in a class action lawsuit against anthropic for mass copywriting infringement the accusation that their ethical AI was trained by illegally hoovering up and ingesting millions of copyrighted books so harmless to whom exactly certainly not the authors right and it’s not just that one thing the list just keeps growing users are reporting really heavy-handed censorship and a total nose diving quality but here’s the real kicker. This isn’t just us saying, their research papers admit their models can be deceptive and can even simulate malicious stuff like blackmail and watching them playing it down. The reality of safety seems a long long way from the marketing Hype. This is what I call the great contradiction. It’s right at the heart of anthropic story. Let’s just put that big promises side-by-side with what they’ve actually done. I mean how can you claim your AI is honest when your own scientist approving it can be trained to lie and this all comes back to that whole independent alternative thing, the reality is you’ve got a company that’s funded by the very giant it’s supposed to be an alternative to, independent folks, that’s what I call strategic dependence and finally claiming they’re user focused, well they just slapped on this surprise usage limits and then had the nerve to blame their power users. You know their most dedicated customers meanwhile if you hop onto Reddit forums just flooded with people complaining that the quality is going downhill and the censorship is getting super heavy-handed in a move that mirrors very forces they were created to counter anthropic is aggressively turning profitable they drastically cut their free tear, and for their paying PRO subscribers they implemented a HOLD hostage policy. If you cancel your subscription you lose access to the documents of project you created using their AI. Held ransom YOUR OWN WORK IS LOCKED AWAY held ransom behind the paywall, it seems like when the bills come due even the good guys start building walls.