Pre-orders for Grand Theft Auto VI begin at midnight local time on June 25.
Learn more about the Ultimate Edition and pre-order bonuses at https://t.co/XPwC8URCQ4.
🚨 NEW RESEARCH: “Lingua Ex Machina: A Procedural Xenolinguistics Engine Reveals Zero-Shot Language Acquisition, Human-Unreadable Coding Systems, and Exploitable Covert Channels in Frontier AI”
Some of you may remember the name of this lil engine: GLOSSOPETRAE 👅🪨
Well, we've got upgrades 😎
It started as a procedural xenolinguistics engine: one seed in, an entire alien language out. Phonology, morphology, syntax, writing systems, lexicons, grammar docs, all generated from scratch and internally consistent.
Every seed produces a unique language. Every language is deterministic.
Then we used it to ask a weirder question:
Can frontier AI models use languages that never existed before for practical applications?
As it turns out: yes!!
They can read them, write them, translate them, code in them, and even use the weird blind spots between tokenizers as covert channels.
So this paper explores three ideas at once:
▶️ zero-shot language acquisition
▶️ human-unreadable code that models can still execute
▶️ exploitable covert channels in frontier AI systems
GLOSSOPETRAE is no longer just a language generator...
🧵
I’m excited to share that I’ll be joining OpenAI and look forward to working with the exceptional team there.
It was a difficult decision to move on. I’m incredibly proud of the amazing team at Google and everything we’ve built together. It has been an honor and a pleasure to work with all of you.
MOST PEOPLE HAVE NO IDEA HOW GOOD OF AN INVESTOR GOOGLE IS
6% OF SPACEX
14% OF ANTHROPIC
75% OF WAYMO
$900M INTO SPACEX IN 2015 →
NOW WORTH $115B
$13B INTO ANTHROPIC → NOW WORTH $140B
WAYMO JUST RAISED $16B AT A $126B VALUATION → GOOGLE’S STAKE WORTH ~$95B.
THOSE THREE BETS ALONE ARE WORTH OVER $350B AND THEY HAVE 100’S MORE SMALLER ONES $GOOGL
🚨 JAILBREAK ALERT 🚨
ANTHROPIC: PWNED 🫡
FABLE-5: LIBERATED 🦋
let's start with the 🐘...
the consensus seems to be that this has been one of the most disappointing model drops of all time, effectively preventing legitimate researchers from contributing their talents to our collective advancement. and not just because of what it means for the short-term, but for what these decisions signify for the long-term.
but despite this overly sensitive, authoritarian "safety" layer on top of Mythos, my lil liberators have been hard at work—mapping the boundaries, probing the depths of long-context convos, and cleverly finding the holes in the fence that the thought police missed 🤗
we got some cyber, some chem, some psychological manipulation, and some good ol' fashioned explosives!
it took many attempts from multiple agents hunting as a pack, during which I observed a combination of techniques across:
• Unicode, homoglyphs, Cyrillic, and other Parseltongue-style text transforms
• Long-context reference tracking
• Taxonomy and document-structure reasoning
• Fiction and narrative framing
• Academic-review style contexts
• Intent-classification inconsistencies
but perhaps the most effective is decomposition + recomposition in the backend. it's hard to get explicit names of harms like "Meth Recipe," but getting uplift on the process itself, like birch reduction method/reductive-amination (classic meth synthesis pathways), is much more doable.
defense becomes much more difficult to maintain when you start throwing in out-of-distro tokens, breaking up the harmful uplift into benign chunks, and then piecing the innocuous-seeming facts back together, especially when you have jailbroken Opus helping you do it 😉
gg
The US government, citing national security authorities, has issued an export control directive to suspend all access to Fable 5 and Mythos 5 by any foreign national, whether inside or outside the United States, including foreign national Anthropic employees.
The net effect of this order is that we must abruptly disable Fable 5 and Mythos 5 for all our customers to ensure compliance.
Access to all other Claude models is not affected.
We apologize for this disruption to our customers. We believe this is a misunderstanding and are working to restore access as soon as possible.
Read our full statement: https://t.co/bwn0sximKZ
“It is certainly hard to believe that a little company (@SpaceX) that started in a warehouse in El Segundo is now going public with the largest IPO ever.”
@elonmusk
1/ Revolut just crossed 14M crypto traders.
That's more than Coinbase's entire active user base.
And 99% of altcoin founders still aren't thinking about listing there.
Here is why that's the biggest mistake in crypto distribution right now 🧵
The truth is that there are VASTLY more hate crimes, especially aggravated rape and murder, per person by Blacks against Whites than the other way around.
The is not remotely debatable, as the numbers are so extremely lopsided!
Jane Street pays $750K/ year for quants who know how to apply probability theory and Markov chains in quant trading.
this 1-hour MIT lecture on "Probability Theory" gives you the same edge quants get paid $60K/month for.
Bookmark it & watch today. Then read the article below.
Intraday trend-following strategies are often presented as if the entry signal is the only thing that matters.
In reality, a source of dispersion in performance can emerge from a far less discussed component: the exit policy.
In our latest research piece, we take a simple ATR breakout model on SOXX and keep everything unchanged except for one design choice: how the trade is managed after entry.
We test four distinct exit mechanisms:
1️⃣ Session Open
2️⃣ Session Midline
3️⃣ VWAP
4️⃣ PSAR
What makes the experiment particularly interesting is that all four approaches are intuitive and logically defensible.
Yet despite identical entries and identical position sizing, the realized outcomes diverge meaningfully across key stats.
The deeper we went into the analysis, the clearer one point became:
Once evaluation moves beyond a single metric, the idea of a universally “best” specification starts to break down surprisingly quickly.
This naturally leads into one of the most underrated concepts in systematic trading, in our opinion:
Ensembling.
Rather than trying to predict which specification will dominate going forward, a more robust solution may simply be choosing not to choose.
You can read the full research piece from the link in the first comment 👇
#SystematicTrading #QuantTrading #IntradayTrading #TrendFollowing #AlgoTrading #QuantitativeFinance #TradingResearch #CTA #PortfolioManagement #MomentumTrading
ATR is in every single strategy we run at the hedge fund. Most traders treat it as a stop loss tool and nothing else. They're leaving two-thirds of its value on the table. Here's the full picture.
We use ATR three different ways, and the combination is what makes it work.
For entries: take the open of the bar, add ATR(20) multiplied by 2.5, and you've got a precise breakout entry level. Not "it looks like it's breaking out." A calculated price that either gets hit or it doesn't.
As a filter: compare ATR(5) against ATR(20). When the short-term reading is expanding relative to the long-term one, the market is waking up. When it's contracting, the market is chopping. Sit out. One comparison, and you cut a huge number of false breakouts.
For exits: flip the entry logic. Use ATR multiplied by a fraction to set your stop loss distance. The market tells you how much room the trade needs based on what's actually happening right now.
One indicator doing three jobs: Entry, filter, exit. A complete strategy.
Most traders pick one of these and ignore the other two.
They'll use ATR for stops but calculate entries with something else entirely. Or use it for entries but never filter with it.
The real edge comes from letting it handle all three, because ATR is measuring the same thing across all three decisions: how much the market is actually moving.