GeorgL0ngGamma

trading, science, sports

[GeorgL0ngGamma @options_g]

one group of #traders is operating on a purely
quantitative basis, often with solid background in science/engineering
another is only using qualitative heuristics
about price action, market mechanics,
forced buyer/sellers
can the 2nd group survive?
@mikeharrisny @Stefano_Peron

[josh levine @bigjoshlevine]

So humans should stop publishing papers since they are just going to get read by AIs anyway. Instead we should build systems where researchers and publish raw information. It is also important that there be a way that some of the value generated by the information makes it back to the information creator.

[GeorgL0ngGamma @options_g]

@priceactionlab @tweetatweir @hughesanalytics now shredding physical copies of scanned books so they can't be sued again

[Citrini @citrini]

I give it a year until we see a new breed of AI native private equity firms that acquire companies just so they can move their workflows from Claude to open source Chinese models and flip them.

[John Scott-Railton @jsrailton]

NEW: malware developers added nuclear & biological weapons text to to their spyware.

Goal? To trigger LLM safety refusals... so that their spyware wouldn't be analyzed by an AI security scanner.

Cleanest practical example I can think of for why over-indexing on first order safety alignment is risky.

When closed (and open) models ship with aggressive refusals, they will be sprinkled with second-order blindspots that attackers will discover...and exploit.

We are only in the earliest days of attackers leveraging these features, and it wouldn't surprise me if users systems that need to handle complex cybersecurity issues demand that models be less safety-blunted.

In the weeds: @SocketSecurity's post also shows why intention matters in how you design a malware analysis pipeline to avoid prompt manipulation.

H/T to colleagues that shared this with me https://t.co/f3Aj9TYxU4

[mattparlmer 🪐 🌷 @mattparlmer]

Fable assessment at end of first workday using it is that this is an incremental step forward on capabilities and a significant step backward on alignment and steerability, and relatedly a serious disaster in terms of policy choices and taste

[FinancialJuice @financialjuice]

SoftBank's effort to secure $6 billion OpenAI margin loan falters

[GeorgL0ngGamma @options_g]

@goodalexander how good is it once they rugpull the pre-IPO subsidies?

[BowTiedBull.eth - Read Pinned or NGMI @BowTiedBull]

Every year there is one Casino, no need to hate on the narrative they eventually sell and put it elsewhere. Be there before them

https://t.co/Hg96mIydtd

[GeorgL0ngGamma @options_g]

@ambient_xyz of course they made it butterflies ...

[illiquidity providooooor @skyquake_1]

We're not quite there yet but its getting close with the mega raises...

[Agustin Lebron @AgustinLebron3]

Overhedge and tank the game?

[ak0 (semis/acc) @annanay]

I am a big fan of exchanges allowing execution by RFQ, but this has HUGE caveats:

Mainly, care must be taken to ensure that the likely direction of the trade is not 'leaked' to the RFQ providers.

Otherwise the MMs can and will front-run the flow.

For example, the exchange can disable requesting one-way RFQs.

They can also try to ensure that the user's current position and balance are not leaked to the MM, but this can be very difficult with a fully on-chain matching engine.

Some other points

> The trade price slippage must be very easy to see against some benchmark (eg a naive market order).

> In the FX world, some RFQ streams are earmarked as 'full amount'. This allows users to get tighter pricing on a given size, on the promise that they will not TWAP the RFQ session (or they would enforce this by making RFQ unavailable for x minutes).

[OnlyFund ( top 0.01% on Bloomberg IB) @OneHotCode1]

@larpcapitalwc

[fiskantes ⭐️🩸 @Fiskantes]

Is there anyone who relies mostly on AI tools/agents for trading/investing who outperforms on a non trivial sample size?

[Haralabos Voulgaris @haralabob]

Looking at the ZEC price action has me excited for the upcoming @naval tweet explaining how financial ruin is just the universe teaching you non-attachment.

[Haralabos Voulgaris @haralabob]

Sure, but the book sets the line. Are you saying sportsbooks intentionally set lines that systematically remove their own edge? Because if so, that sounds like a sportsbook problem.

You’d be better off just embracing your actual role: a parasitic goblin whose job is to extract maximum value from customers, even if it destroys their lives, while simultaneously banning anyone with the slightest chance of showing a profit.

[Quantum @Quantum_Sport]

Bitcoin being sold to fund korean/jap semi and AI-adjacent industry buying whenever there is the absolute top they are always there literally never fails

[Price Action Lab @priceactionlab]

@GestaltU New dark ages coming. Secret societies will prosper.

[686 Ebullient Prism @686Prism]

I think the 2000s telecom analogies are flawed mostly cause the companies now are actually making money, in some cases hand over fist. The better analogies in my view are the 1840s UK railways and the 1907 panic with all the trust structures that controlled ungodly amounts of shares. In both cases the entire move was retraced before it started going up again. But mkt doubled or tripled in these cases from its starting base before the move ended. SPX is near double what it was in 23' so we're getting there.

[sysls @systematicls]

This is probably good news for younger, leaner organizations.

Behemoths will not simply be able to propagate AI throughout a machinery where most of the important context is inside numerous, disparate human heads.

There is a chance for meaningful disruption here.