@BullishTimes_ Hayes framing his exit as a trust call rather than a price call is the part that stings most for Zcash. He wasn't saying ZEC is overvalued. Just saying the core promise which is provable privacy has a hole in it that nobody can definitively say wasn't used for four years.
@WouterW_Vreug Exactly. One hand writes the safety manifesto, the other hand files the S-1. The "please regulate us" blog post generates headlines. The IPO filing generates $65B. Guess which one the board actually cares about. π―
Anthropic just begged the world to stop building AI. π¨
The company β valued at nearly $1 trillion after raising $65B β published internal data this week showing Claude now writes over 80% of its production code. That's up from single digits in early 2025. Their engineers ship 8Γ more code per quarter than they did between 2021 and 2024. Claude achieves 52Γ optimisation speedups on complex tasks.
And it gets wilder. Open-ended engineering task success rates jumped from 26% to 76% in just six months. That's not incremental improvement β that's a phase change. Anthropic says recursive self-improvement β AI building its own successor β could arrive within two years.
So what's their solution? A globally coordinated pause on frontier AI development. Verifiable agreements between the US, China, and allies to slow everything down before it spirals.
Sounds noble. Except for the timing.
Anthropic just filed for IPO. They've raised $65B. They're approaching a $1T valuation. David Sacks called it regulatory capture dressed as altruism β and he's not wrong. You don't build the fastest AI on Earth, raise more capital than most countries' GDP, file to go public, and then tell everyone else to stop.
This isn't safety. It's the biggest "pull up the ladder" move in tech history. The fox isn't guarding the henhouse β it built the henhouse, ate the hens, and is now writing the food safety regulations.
That's the question nobody at Anthropic wants to answer publicly. Their own data shows Claude's success rate on "open-ended engineering tasks" jumped from 26% to 76% β which means 24% of the time it's still getting it wrong. At scale, that's thousands of edge cases silently shipping. The blog conveniently skips over what broke, who caught it, and how many made it to production before a human noticed. π
Zcash's infinite money bug π sat hidden for FOUR YEARS while the world's best cryptographers looked right past it.
An AI built the exploit in days. Arthur Hayes dumped everything. ZEC crashed 41%.
The shielded pool can't prove it's clean. That's the real crisis.
ππ#Crypto #Zcash
Source: Anthropic blog, 4 June 2026 β "When AI Builds Itself" by Marina Favaro & Jack Clark. Would you trust the company building the bomb to call the ceasefire?
Hayes predicted $150 HYPE. 72 hours later, he dumped $18M of it and rotated into Worldcoin β a token ZachXBT called a scam. The audacity is the product. π #defiendelopΓΊblico
See comments for full π§΅π
@TheblackCeo233 We always found it odd how the meme crowd had any faith in the scene.
Such a lottery, and such obvious manipulation.
I swear most winners are just alt accounts held by cabals to con people into continuing to pump sol to the next shitcoin offering.
@ETH_Farmer_003 Two days. That's all it took for a jury to see through the "traders trade" defence. Now imagine applying that same standard to Telegram alpha groups pumping tokens with undisclosed bags. The precedent is set β the question is whether anyone acts on it.
Andrew Left just got 13 fraud convictions. Crypto should pay attention.
The Citron Research founder ran a $21M tweet-and-trade scheme for five years β publish inflammatory takes, secretly trade the other side, tip off hedge funds before hitting "post."
Sound familiar? It should. Crypto's paid shillers do this every single day. The difference? Left is facing 25 years. They're still posting.
See ππ§΅π
@Eddieswrld22 This is the part that doesn't get discussed enough. DYOR assumes the information landscape is neutral. When the research itself is engineered to move price, the whole framework falls apart.
Manufacturing liquidity. That's exactly what this is. The conviction content, the charity bet, the public calls β all demand generation for an exit that was already planned. Well put.
@BullishTimes_ If you tell thousands of people to buy while you're secretly preparing to sell, you're not leading a community.
You're manufacturing liquidity.
@Gypsum_01 He cited "AI IPOs and energy prices" as the reason. Make of that what you will. Nine months of conviction, a public $100K bet, then a macro excuse and an $18M exit via Flowdesk. The bet's still technically active, though β so at least there's that.
Crypto is cooked. We let it be cooked by being excited re ETFs, Trump, big finance etc.
Who did what they always do. Diluted TF out of it with derivatives.
Started dirty yet legal games with liquidity sweeps to clean out retail without the margins to weather short-term fluctuations.
Trump, WLFI and USD1.... nothing needs to be said.
You all cheered as the dream was killed. The worst thing, some still are.
@Eddieswrld22 The older you get in this space, the more you realise the most valuable signal isn't the target price. It's whether the person making the prediction is increasing their position, holding it, or quietly reducing it whilst everyone else debates the headline. π