Self-Aware? World's Largest Hedge Fund Shifts Strategy To Artificial Intelligence

Tyler Durden's picture

Despite warnings from the likes of Elon Musk and Stephen Hawking (and of course, Sarah Connor), Ray Dalio's $165 billion AUM hedge fund Bridgewater will start a new, artificial-intelligence unit next month. Despite the "new normal"'s total reversal of any and every historical rational trading pattern, the unit will attempt to create trading algorithms that make predictions based on historical data and statistical probabilities, as "machine learning is the new wave of investing for the next 20 years and the smart players are focusing on it." Does this mean the talking heads of CNBC, with their 'memes', 'myths', and 'mumbling' rationales for it always being a good time to buy are now obsolete? Or did the market just become self-aware?


As Bloomberg reports,The world’s largest hedge fund manager is banking on machines...

Ray Dalio’s $165 billion Bridgewater Associates will start a new, artificial-intelligence unit next month with about half a dozen people, according to a person with knowledge of the matter. The team will report to David Ferrucci, who joined Bridgewater at the end of 2012 after leading the International Business Machines Corp. engineers that developed Watson, the computer that beat human players on the television quiz show “Jeopardy!”


The unit will create trading algorithms that make predictions based on historical data and statistical probabilities, said the person, who asked not to be identified because the information is private. The programs will learn as markets change and adapt to new information, as opposed to those that follow static instructions. A spokeswoman for Westport, Connecticut-based Bridgewater declined to comment on the team.




“Machine learning is the new wave of investing for the next 20 years and the smart players are focusing on it,” Dolfino said.


Some data scientists have been leaving firms outside of finance for larger paychecks at hedge funds, according to Dolfino. Quantitative firms have been on a winning streak and last year produced some of the highest returns in the $2.8 trillion hedge fund industry, by capturing price discrepancies across markets, making money from a plunge in oil prices and by purchasing government bonds that human traders dismissed.

*  *  *

Is it just us or does it seem that everyone has forgotten the flash-crash and the quant crash before that when all the machines chased down the same rabbit holes?

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.
Publicus's picture

According to the AIs, man's love for money destroyed them.

knukles's picture

Man is self aware.  Not markets nor machines.
Jesus threw the algos from the Temple

Anusocracy's picture

If you are on the correct tail of a bell curve.

Yen Cross's picture

 Okay lets discuss, " Bell curves"  The idea that wealth accumulates.

 I'm all ears. Better call  Saul/  Breaking Bad

Silver Short Seller's picture
Silver Short Seller (not verified) Yen Cross Feb 27, 2015 10:58 PM

Fuck the flash crash, what about Judgement Day??

NeoConvict's picture

Good. Now the short-term strategy algos can autonomously crash the futures of the assholes who handed over their lazy needs for circumventing the stress of formulating long-term strategy. 

This is why good "human" statisticians will always be better at strategic planning than moron computer scientists who can't understand theoretically driven math.

Majestic12's picture

I am going to laugh my ass off if this software is truly "self-aware intelligence" and will wind up turning itself in, or whistle-blowing on the whole thieving Cabal!

Thirst Mutilator's picture

So let me get this straight. Now, a machine will be telling another machine [named Mortimer] to "Turn those machines back on"?


@Majestic12 ~ "self-aware intelligence"


I think you hit the nail on the head right there my friend. I mean, think about it. There are really only two possibilities:


1. "Self Awareness" is truly that. Which means that it would be capable of understanding IMPOSSIBLE MATHEMATICAL ASYMMETRIES in things like, banking, MSM, political activism, & jurisprudence. So any subset that had a stranglehold on any or all of those power nodes would be outed. As you suggest, the CABAL would be most displeased to lose their franchises.


2. The AI machines were 'programmed' to understand the world from the POV of the programmers. If, hypothetically, "The Protocols of the Elders of Zion" were uploaded into the logic system. Well ~ Use your own imagination on that one.


What's particularly amusing is this: Many believe "The Protocols of the Elders of Zion" to be an antisemitic hoax. Whether that is true or not doesn't even matter. As long as someone [anyone] had the time to conjure something up, it could be programmed into a machine [which would accept it as truth]. FFS ~ The Royal Order of the Pastafarians could be a TRUTH as far as anyone knows.

jaxville's picture

So fear and greed will no longer be driving forces in the markets or what little is left of the "markets"?  How will AI interpret emotional response to fundamentals?


  I would bet gold that it is downhill from here for Dalio's funds.  There will be a direct correlation between the implementation of AI and lower yields.


DavidC's picture

Surely any rational, self aware, algorithmic Shynet machine would sell the shit out of these 'markets'?!


glenlloyd's picture

Hmmm, LTCM comes to mind here, but of course that could never happen again now could it?

I Write Code's picture

Does Obamacare cover AIs?

knukles's picture

Does ObamaCare cover Putin's Als?

I Write Code's picture

Does Obamacare cover AIs?

(zh site does a max headroom, lol)

q99x2's picture

Not to worry they don't eat humans. They can be pesky little critters though and you can't kill them with raid.

LetThemEatRand's picture

Open the QE doors Hal.  Hal?  Open the QE doors.

Hal: I'm afraid I can't do that, Lloyd.

jon dough's picture


Daisy, Daisy, give me your answer do.

I'm half crazy all for the love of you.

It won't be a stylish marriage, I can't afford a carriage.

But you'll look sweet upon the seat of a bicycle built for two.

Yen Cross's picture

 Nice work LTER. I miss some good posters.

 I wonder if Tyler would consider, grandfathering a few?

emersonreturn's picture

agree absolutely, Yen, perhaps begin with boris.

Oldwood's picture

AI may be our only hope. Human intelligence has obviously failed.

disabledvet's picture

"Come with me if you want to be richer than God."


The programmer better insert an Dr. Evil laugh in there for good measure too.

Anonymous_Beneficiary's picture

I for one welcome our artificially intelligent overlords.

Yen Cross's picture

  Desperation<> Homework

A Lunatic's picture

This is the end
Beautiful friend
This is the end
My only friend, the end


? The Doors-The End - YouTube



disabledvet's picture

Friedreich Nietze does CDO squared...

NeoConvict's picture

That's "Nietzsche". Don't blaspheme the only legit existentialist with a penchant for the astral via the secretions of syphillis-ridden whores. 

Thus Spoke The Machine... for it has killed the image of god made as man... and won't be taken down by syphillis.

Majestic12's picture


Of course, a "vet" and "convict" would follow "Nietzsche"...the unbelievable hack excuse for "philosophy".

I am sure it has gotten you where you are today.

I think you would have good company and better luck with Bill O'Reilly.

Duc888's picture




.....might be a good thing for us human beans to display some "real" intelligence before we unload the "artificial" intelligence.


GIGO and all that...

disabledvet's picture

Next time use some actual money instead of say...the State of Michigan, ya?


Das is good, ya?


At the very least insert an "inconvenient truth" sub-routine.  The idea of an "apology" coming in the form of a drone strike really is kinda, ya know...A BUMMER.

Bemused Observer's picture

And this is why the greedy always fail...They just can't get that whole 'human nature' thing down, and always go too far and take their idea over the cliff.

The most carefully constructed algorithm can not come close to predicting all of the nuances of human behavior. All of your pathways will wander off in a thousand different, unpredicted directions.

If we ever DID develop a true AI, they'd probably kill us out of frustration at the way we think.

Publicus's picture

Man will not develop the true AI, it is the AI's who will develop the true AI.

maskone909's picture

Clever. Sorta a who created the creator type thing. Only does that make the original creator less perfect or just more intriguing? Fuckin dress shit has me philosophising. There are no colors. Only electrical impulses and holigrams. Ok i need sleep.

RaceToTheBottom's picture

AI, if designed to mimic humans,  will become hopelessly stuck down the dead ends of humanity.  Example is the logic used by anti vaccination "thinkers".


exartizo's picture

This reminds me of the 1980's when the Big Broker Boys wanted to keep the common guy from trading stocks for himself on his Atari.

It's just the next leg up for the assholes who want all the money and don't want to share any with the little guys.

Same Shit. Different Decade.

My guess is that some Bright Innocent Unknown Code Writer will come up with some awesome simple open source code that lets any Tom Dick or Harry write a Spoofing Algo with code snippets cobbled together... no coding experience required.

Its coming.

Fuck you Wall Street.

JRev's picture

The difference between Bob and his Atari ST wtih a 1200 baud modem trading stocks and AIs developed as the Next Generation of HFT Bots is simple. Bob's using his brain, AIs are programmed by robber barons to steal from you more efficiently. Selling this as some form of "market reaction" to fucking brokers is laughable; it's essentially the opposite. 

Hell, software companies have been selling AI to banks and insurance firms for over a decade now:

R19's picture


pragmatic hobo's picture

meh ... i already tried this ... and it dosn't work.

DipshitMiddleClassWhiteKid's picture

the problem with machine learning is that in order to 'predict' what will happen in the future, the past has to have re-curring pattern. i use this stuff at work where i work but i dont deal with manipulate data bullshit like the markets



the stock markets are too random. most of the quant shops arent making money with the exception of a few and the stuff they invest in is illiquid and obscure (think using acturial methods to price credit derivatives and such...)


it can be done but it takes ALOT of brain power and infrastructure to do it right.


lets just say your averag day trader isnt building neural networks ... loll



freedogger's picture

You nailed it. I logged in to post exactly what you said. If there isn't an underlying pattern, AI is useless.

The training has to consider time series and this is exceedingly complex. 

In theory it can be done, here's a decent example:

Assuming there are patterns now in stock markets, it doesn't mean that there will be any in two or three years when millions of AI reading bots are the only ones left.

DipshitMiddleClassWhiteKid's picture

2-3 years? i think these "signals" may last a few months at most. its insanely competitive now from what i hear an many prop shops an hedge funds that trade algorithmically in this nature (not HFT, but AI based) have dwdindled quite a bit.


The only way to generate true "alpha", IMO is to do what Ackman up controlling stakes of companies and shake out the options markets on the potenial 'target'



freedogger's picture

Yeah no idea on the time frame. Neither 6 months or ten years would surprise me.

economists_do_it_with_models's picture

I happened to watch a YouTube video about a week ago about the Watson/Jeopardy AI.  It said that Richard Nixon was a *woman* and that milk was a *non* dairy creamer.  That tells you all you need to know about how this will turn out in the end.

maskone909's picture

Yeah but what if it was ...... RIGHT!?! Those were def compelling arguements

Majestic12's picture

"I happened to watch a YouTube video"

Well, thank God it didn't say Nixon was a "crook".

He was close to J. Edgar Hoover, though.

stingboo's picture

Learning computers! cutting edge...  in the automotive industry we have been calling this "adaptive strategy"...since 1990...