TEDxNewWallStreet - Sean Gourley - High frequency trading and the new algorithmic ecosystem

[@hariramnarayanan5157]

The fact that this video is 8 years old is scary to imagine what those algorithm have grown up to by now.

[@ToptierGod]

Fr

[@hillbillymagick7964]

Enter covid lock downs and automatated trading while everyone is at home. Amc and gme get hit 150+% over shorted at least, who knows if not thousands. Wow!

[@hariramnarayanan5157]

@@hillbillymagick7964 you used a bot ?

[@hillbillymagick7964]

@@hariramnarayanan5157 no not me, the Hedgefunds. Thats the reason Gamestop and AMC have been shorted to infinity. Computers left alone, programmed not to lose.

[@jinxy7869]

shut up u sound like a kid, these algos don't make more than 50% an year

[@SouvikAswad]

This was one of my most liked Ted Talks. Great, illuminating storytelling.

[@nirajgupta9982]

Agreed!

[@goikofinanzas]

I'm currently reading Flash boys not so fast by Peter kovac. This video is a great companion to such a great book

[@leeles]

Thank you, Sean Gourley, for this insightful TedTalk, sharing insider views of stock market trading with algorithms and also being honest about the potential dangers and dark sides. I appreciate your vision of going towards an "augmented intelligence" by combining human and artificial intelligence wisely.

[@mikestoneadfjgs]

this guy is absolutely 95% ± 5% correct.

[@goikofinanzas]

I see what you did there ~2 standard deviation ☺

[@nothanniballecktor9633]

When he starts the talk talking about a human brine, you know it's gonna be good.

[@KP-dd2ci]

I can spot an Aussie accent like a fart nowadays.

[@redcenterau]

Bwahahaha... bloody oath! Goodonya, mate. 👍

[@RyanDavisSoftware]

Thank you

[@lionelshaneyfelt5607]

lmfao

[@SystemsTrader]

This video highlights my business world:A domain traversing philosophy and psychology as much as technology and markets.The challenge of technical fields such as applied genetics, neural networks, and other sophisticated mathematical algorithms is bringing life to artificial "intelligence" ... and so on and so on.It is complex, and it is fun, deliciously so, yet we do have a responsibility to maintain awareness of our actions at all times so as to realise benefit for sentient beings.

[@shandoticwa]

now it is 2021! wonder at what level HFT is happening now specially because AI has advanced so much in last few years.

[@boonedockjourneyman7979]

Insane. Thermodynamics an real estate along the fiber are the only limits.

[@forexwall3687]

Now it is 2024 and they are controlling everything.

[@MMPICT]

It has become so advanced that you, as a human trader, can get a significant edge in the market by only reacting to algorithmic patterns and understanding time x price delivery

[@jacobbranker159]

Jake Tran brought me here. Excellent talk. Thanks Dr. Gourley.

[@jamess6787]

3:15 ..... spot on. The algos caused flash crash, and one poor human sucker got the blame

[@lucarovi8622]

xd

[@Gamenetreviews]

Greg Williams And you will sell me your 500 dollar course on how to get rich quick by trading faster than the robots.

[@guilloutube]

You forgot to mention that those algorithms are financed, developed and used by humans. I'm not afraid of algorithms, I'm afraid of those humans

[@jinxy7869]

ok

[@shipitintheremate]

My favorite Ted Talk.

[@joamo7392]

Great presentation! KZbin algorithm more of these videos, please to me.

[@nads9298]

Yea

[@luszlips]

The spread between buy and sell orders is narrowing with algorithms, good for the investor bad for wall street. Go algorithms!

[@washingtonmultimedia5896]

Algorithm

[@coenraadloubser5768]

The spread is just one algorithm gaming another.

[@ajcook7777]

Yes! I knew it was an algo that dropped ADGI so damn quick on Dec. 14, No one had time to react...just immediately TANKED! definitely was an ALGO

[@wedding_photography]

The main point he missed at the very end is that Wall St will be there to take advantage of AI or AI augmentation way before an average person. So nothing will change. What we really need is a new exchange where each bid costs money, where the order of trades is randomized, where trades take random amount of time to execute (let's say, up to 1 minute). This will instantly stop all the high-frequency trading.

[@augiedoggie]

I like the idea of a dollar value for each bid, though if companies are willing to spend billions on cable to make their HFTs milliseconds faster they can find a risk/loss point on how many bids they can put in and still make money. I think the best way to avoid this mess is to invest rather than trade. The average person probably shouldn't be day trading.

[@chounoki]

Trading is never about fairness. You can easily think of 1 million ways to make trading more equal and just to all participants. Sadly, that is NOT the interest of this industry. It is all about congregating money from many people to only a few, not the reverse. So you do that by deliberately creating inequity.

[@augiedoggie]

Chounoki I guess so, I still just think its a drop in the bucket if you're long on a stock. Small day to day fluctuations haven't ever disrupted my portfolio beyond the inconvenience of having to hold the stock a little longer.

[@chounoki]

augiedauggie Right. But for these trading firms, they make a lot of money from this kind of short term fluctuation. In the end, if the stock price never changes, how do you make money from it? The more it changes, the more potential to make money from it. The faster it changes, the faster they make money from it. If you make $100 by holding the stock for a whole year, they can make the same amount of money in 1 day from the fluctuation.

[@IDNeon357]

Chounoki It's the man-in-the-middle problem of all markets. A wants to buy 100 of something B wants to sell 100 of something. A middle-man C either because of fore-knowledge of the event or other conditions takes the 100 from B instead and extracts 5% for himself and then transfers 95 to A and calls it a trade. The reality of the numbers though is fractions of fractions. A wants 100, B gives C 100, C gives A 99.9999. A is satisfied with that result. C produced and did nothing, and has no reason to exist.

[@scottab140]

FYI... Algorithms are mathematical and all of the Quants blew up in 2008. Fundamentals and Technical Analysis are totally separate as well.

[@Elaba_]

People use algorithms to replace humans. It still stays a zero sum game if you don't have a competitive edge.

[@MarkoKraguljac]

Important distinction: We are not slaving for machines and algorithms but their privileged owners, humans. They chose and were allowed by their bed fellows - politicians to automate finance instead of automating economy and making all lives better.

[@5P4C3V01D]

This is eye opening.

[@Poliglotaignoto]

great presentation

[@AsiaStreet365]

2023. I wonder how far we have come with from 10yrs ago with the altos trading. I'm finding the markets are more uncertain than ever.

[@kahisawheel]

Very cool presentation.

[@scottab140]

Yep, you are correct. It wasn't the point of the video only an observation that you rather observed or didn't, which was my original question. What do you think about my original question now?

[@christianevans3357]

Great Video. Very Informative!

[@michaelreyes2794]

Very good trading video. I personally make about 200 dollars each day net profit trading online. If you want a free consultation on how I do this send me an email here: toshtao@gmail. com

[@jaravta]

If you want to help someone, make it looks easy. If you want to impress someone, make it looks complicated. Well, This video should help us instead it makes me more complicated.

[@michaeltierney2539]

Great talk.

[@Aka_GUNDAM]

Just found my way to this video 🐺🔥🍾

[@MrPaulzeng]

Hi, Thanks for your video! I wonder if you've given thought to this strategy: often traders lose money because their stop loss gets hit and the price bounces back but the trader is no longer holding that position. What if you can write code that upon your stop loss sale, you place a limit order to buy back at the sale price? And then on the upside, you set a trailing stop loss (and if the trailing stop loss gets hit, you’d also place a limit order to rebuy the stock at the stop loss price) This strategy essentially limits your downside and lets you participate more fully on the upside, and takes the “art” out of guessing the optimal stop loss and take profit price levels. I wonder if you’ve ever considered this strategy and if you see any problems with it. Thanks in advance for your response :)

[@rajat0610]

this sounds great!

[@ImUrZaddy]

If you’re gonna buy back at the pice where you’re stopped out, then why bother using a stop to begin with ? A stop should be set where the trade idea is wrong, so if the stop is triggered then why get back into the same trade ?

[@MrPaulzeng]

@@ImUrZaddy let’s say you bought in at $105 and your stop is $100. Sometimes this is the right stop, and the price goes to $80, $70, $60… but sometimes this is the wrong stop, and the price goes to $99 but comes back up. In this situation, given that you had bought at $105, then you would wasted $5, because when the price came back up, you were no longer holding the stock.

[@paulceltics]

great talk

[@danielstapler4315]

High frequency trading is high speed front running. A computer housed in the exchange detects a order on its way to the market, lets say to buy 1000 shares of ABC. The computer then places an order in front of that order and buys the shares and then sells to the real buyer of the 1000 shares. If you did the manual equivalent this you could go to jail. But because the high frequency traders pay the exchange a lot of money its ok. This is just corrupt bullshit.

[@Evan-McM]

+Daniel Stapler Not all HFT is front-running, in fact most of it isn't. Do some research, there are many, many strats that are encompassed in HFT.

[@GabeClendenning]

They allow you to fill your order faster and priced better

[@CzechRiot]

Dark pools are private exchanges or forums for trading securities, not accessible by the investing public, they are so named for their complete lack of transparency.

[@GabeClendenning]

CzechRiot many brokers offer access to multiple ats' and dark pools as well as a range of proprietary and 3rd party SOR systems

[@das250250]

How is the algorithm intercepting your contract before the exchange ?

[@yasiradnan9051]

Great talk indeed

[@srebrnimedved]

👍 thanx, Sean

[@JohnMcLaughlinStockCoach]

Excellent, informative video - good work, thanks. John McLaughlin, StockCoach

[@illoominatedfool3815]

Is it our world or the machine's? Which century do you want to live in? If you expect to use your credit card at wal-mart, or to pay your bills online, or to get real-time and ACCURATE exchange rates and commodity prices, you need algorithms. Do we work for the machines? No; they work for us- and when crashes happen because a human wasn't perfect in programming that algorithm, is it of any value to blame the machines? Should I blame my computer for being vulnerable to hackers, or should I blame the programmers for leaving/exploiting holes in my system, or better yet should I blame myself for investing in a flawed device to begin with? When it comes to the markets, it's not the machines that devalue your IRA-- it's the banks that you have given your money to that use flawed systems to manage your retirement. Perhaps the loss of wealth in our country hasn't been because of technology, but rather the collective failure of an entire industry to better understand technology's implications in our modern world? "We use mathematics to make (our complex world) simpler" says Mr. Gourley. Except mathematics make up our complex world to begin with. It's a universal truth. It's up to us to evolve our mathematical understanding in order to make our ever changing world simpler, not the other way around. Never before has the market been so efficient, so fair, so liquid, so accessible, and so affordable for the little guy, and that's all thanks to algo's and high frequency trading. Stop being afraid of what these fear mongers say and take control of your own wealth and financial future. Or at the very least, don't expect someone who only pretends to know more than you to honestly care about your finances any more than you would.

[@badboy7946]

Well said brother. The older I get the more I see people who have access to address a crowd on a public platform like a stage, live event, tv show and even the internet try to get an emotional rise out of the audience by coming up with a bunch of baloney just so the average folk who has never looked into the subject can relate and say "oh yeah off course now I know why I don't have enough money to retire the HFT's are to be blamed, I'm getting fat and have all kinds of illnesses it must be because the government is injecting stuff in my food for it to grow faster" & its just funny how the cycle works the people buying into this sort of non sense get to be lazy and blame others for their shortcomings and lack of taking responsibility and the person presenting well they get to sell something at the end that promises a solution to the problem (presented in a manipulated way by themselves in the first place) & when it doesn't work the cycle repeats itself and goes something like " I got ripped off its all his fault I did nothing wrong" instead of " oh I think I should do some research and see what I should eat and how much rather than buying pills, and look into how the markets work rather than investing with some fund whose sales rep promised a 6000% gain in 99 days. But anyways thanks for reading my post, Good Day.

[@rhinothegod]

Well said brother!

[@arBmind]

Is this our world anymore or is it the world of machines and algorithms?

[@forfitlife]

Andreas Reischuck it's SKYNET :o

[@kendalmeldrum3029]

Andreas Reischuck great video telling the truth, I did not need to buy the program to know it is a scam?, i use ANDY LANK TOOL for daily income.

[@johngury]

"speed of light takes 65 MIliseconds to travel from london to NY" Uhhhh, right there Dr. Sean

[@nathanma4273]

7:13 GE stock traded @$25, now it is $8.79. How is the algorithm perform on GE?

[@x10mark24]

probably fine, they only hold the stock for seconds most of the time and profit from tiny movements in the price not long term fundamental trends.

[@MazeFrame]

When machines generate money from nothing, the average person is in trouble...

[@Evan-McM]

+MatzeGamer ...how....

[@padkirsch]

+MatzeGamer haha fiat currency. just like now!! :)

[@Sanathvarma]

MazeFrame but who employs these machines? Retail traders? Hedge funds? Government?

[@drgutai1]

urm, yes.

[@coenraadloubser5768]

And now they can even hedge against you pulling the plug or blowing them up, with cryptocurrency. Aaaargggh!!!

[@basiceconomyorg]

Great video ... added to my website to interesting links ... let me know what you think about my 10 pages publicaion. IMO, HFT would end if every order would be taxed with small Tobin tax ...

[@TheBryanmauro]

The yuman brine is fascinating.

[@hillbillymagick7964]

Insert AMC & GME. Thank you HFT & A.I. for the opportunity of a lifetime!

[@johngury]

Distance from London to NY (5 567.11 kilometers) / the speed of light = 18.5698801 milliseconds I'm sure he never did the calculation. 4:48

[@MrJerryHsu]

john gury THIS

[@fredsalfa]

A human still has to write the algorithm. Doesnt mean the algorithm is always successful. Its only as good as the human is as a trader.

[@nirajgupta9982]

When algorithms can learn it becomes a little more ambiguous.

[@macmos1]

that's why machine learning exists.

[@anonimu0000]

Alfa Won they call it AI and ML

[@fredsalfa]

Name one person or AI or ML that has made alot of money. I bet you no one can. And when I say alot of money Im talking about George Sorros or William Buffett type of money

[@victos-vertex]

That's simply false. We already have AI that is better in specialized fields than it's creator. it's just a matter of time the same will happen with trading. We already have cracked the game of "Go", we have AI that wins tournaments of Texas Hold'em, Deep Blue kills chess. There are several ML paths to take and deep learning is just one of them. Algorithms get better every day and no - they do not depend on the knowledge of the creator in the respective field the AI is used in. If the AI was only as good as the creator then none of those advances would've been possible. How would you beat the world's best GO player with an AI if it wasn't made by the world's best GO player then? I can throw an AI at an atari game without knowing anything about it and this AI will be better than any human being in no time. The AI doesn't know it's playing a game, neither does it know how it works. It simply tries to optimize a function (in that case winning the game, by throwing outputs at it) based on the given inputs (in that case the raw pixel data). At no time at all has the creator of such an AI anything to do with the games themselves. In fact considering deep neural nets we don't even know exactly what's going on within hidden layers of the net. We have a vague understanding but we don't know exactly. Taking computer vision as an example: throw a picture through a bunch of convolution layers, throw some normal invisible layers behind that, and an output layer. What does it do? it spits out numbers based on the input. Nice now throw backwards propagation at it. what you now have is a basic convolutional neural network that doesn't know anything about it's task. now throw millions of pictures at it and tell it to maximize the functions to output "cat" and every time there is no cat in the picture you simply tell it "wrong". this net - that doesn't know jack about cats or pictures or what ever - will eventually be able to identify whether or not a cat is in a given picture with quite a remarkable accuracy. that without the creator ever describing what the hell a cat even is and what the algorithm should look for. You then could take that very same algorithm and do the same with everything else that is a classification problem. And now guess what trading - buy vs sell in simple terms - is. Now just like we haven't solved computer vision we haven't solved trading and many other classification problems - yet.

[@carlosfierro3302]

So, do stock prices ever depend on profit, growth, efficiency?

[@Evan-McM]

+Carlos Fierro In the long run, yes.

[@thrillamoe50]

but these are futures, (indicative pricing/quotations) look at the dow jones and compare with wall st 30 they are different, its an over the counter derivative.

[@mannyzagri7451]

The King is not in danger as both pieces are of the same color

[@currentcomentor1026]

It takes atleast 2 pieces to create a checkmate

[@string22]

Not in the hft sense that you're thinking of, banks are just brokers in the hft world, they just allow hedge funds etc to trade in their name or they'll execute trades for them. banks have tiny tiny prop hft units, if that, and their being phased out.

[@1schwererziehbar1]

i for one welcome our new algorithmic overlords.

[@momo2fry]

This youtube account is not mine would you be more comfortable discussing this via email?

[@Raison_d-etre]

HFTs can be and often are more malicious than are apparent in this video.

[@SOSTacoJohnson]

the big banks are some of the largest hft firms in the world.

[@scottab140]

Anyone one else figure out that the chess game was actually over and not a thinking process of if check-mate was taken place? Hint- There is no other king on the board so the game is won.

[@MrJerryHsu]

scottab140 no the game was not over, there are numerous squares to where the king can escape to, the game is considered a draw because a lone knight will NEVER mate a lone king. It was to check if any existing danger was on the board, then he goes on to explain the time it takes for a non-expert chess player to recognize if they are in DANGER or not. 600ms right?

[@filipion-h3k]

What if the algortithms find this video and decide to shut it down?

[@momo2fry]

how can one learn about algorithm based trading?

[@stankwho]

As long as humans create algorithms they arent going to take over anything. Algorithms still based on fundimentals and technicals. Nothing new.

[@shibrik]

Which programming language is used for execution in HFT ?

[@Jansonsrob]

Kakao shares are cheap we need to buy them now

[@SomeInfo-ib3wz]

So wrong. Zero plus scalping and latency arbitrage just two examples of how many high frequency strategies are based off of technological and queue priority advantages rather than fundamental "value" or technicals like price movement which are both not empirically verifiable unlike HFT strategies. That's why Rennaissance, Shaw, and the other legends can charge fees like 40% or 50%, their strategies almost never lose with technological advantages using special order types and strategies.

[@eloforfree]

i didnt quite got the idea what augmented intelligence exactly is. Or in other words, how exactly does he imagine such thing?

[@BYSMA4th]

Think about equipping a cpu onto your brain to "enhance" how much you can think in 1 second. Basically, you are still in control of your thoughts, but thanks to the machine you are allowed to think of it faster.

[@onee]

eloforfree There are robots out there which are programmed to collaborate with humans. Apparently they're more productive than robots who work on their own or humans who work on their own.

[@harendrasingh_22]

8:17 like the audience ! Hahaha. Atleast they understand the stuff.

[@Denver_Risley]

It's STILL our world. We're not helpless damsels in distress unless you just feel better about complaining.

[@nachoiyaam9986]

06:00 - Statistical Arbitrage

[@goldennuggetz5312]

But only newbie traders trade the short time frame charts. Average Joe the pro usually hangs out on the 120 240 and day charts and these HFT's provide major liquidity for those. These flash crashes are most likely big bank manipuations of the markets, not some fancey hocus pocus algo glitch. And let's dont forget that most of all the crap that runs the markets is build on XP and even older. There is the real danger.

[@IDNeon357]

Ulfric ArrowKnee You cannot possibly know or understand a complex system that is 55% HFT-Algorithmic trades. All you know are political talking points. The fact is we have no real knowledge about what's really going on in a partly automated market and we sure as hell don't know how it affects real people in that market.

[@goldennuggetz5312]

Uhm, no. All the big banks took major fines in 2014 and 2015 for manipulating the forex market, which I retail trade. There were quite a few system crashes on large indexes across Europe including DAX this year. This is because all the crap is build and updated to old windows systems.

[@CzechRiot]

Honestly I think some people paint this holocaust apocalypse vaporizing of people's money thing just for sensationalism, like global warming flooding whole continents and mutant viruses dissipating humankind (which is more likely though...). If algorithms go berzerk and do way too crazy things, the regulators will just nullify everything that happened and set things back to how it was before the "catastrophe". Although random individual casualties are mostly irrelevant (as there are always homeless shelters available)

[@marklopresti1630]

Actually his presentation is not true. The algorithmic benefits of High Frequency Trading HAS NOT result in smaller spreads. False. Look at the research from Nanex LLC in Chicago. nanex.net

[@marvinratchford7913]

I'd like algorithmic technology to perform detection, elimination and forensic tracking of ALL malicious viruses....

[@coenraadloubser5768]

Sounds like the pretext of a popular movie...

[@felixg4785]

This comment sections is full of trading bots.

[@cours458]

what does he mean algorythm that we don't understand? algorythm are written by people, of course they understand the algorythm they write.

[@coenraadloubser5768]

No they don.t because the algorithm goes like this: 1) Look for random correlations between data points 2) Test it on other data points 3) Repeat until it works out 4) Now try it on real world data

[@ChesterDoraemon]

this guy actually has no idea what he is talking about. anyone who cherrypicks examples deserves immediate skepticism and condemnation, as most losing traders are correct on average about 50% of the time and can easily find many examples where there thesis works, yet their account is always an embarrassment.

[@dotrading3394]

Ok, trading with the coming of HFT has changed, but it does not mean that it is impossible for the human race through a discretionary trading ...

[@Zander101084]

what do you mean original question? you only had one question, and if you were talking about the one about the chess board, I already told you: "i don't think that was really the point".

[@momo2fry]

well unfortunately not my world, so would you have any experience in the field or anything dealing with programming at least? We have a team that is willing to pay for your services. Or anyone you could refer to? thanks

[@kennedysan1045]

Is it worth putting ridiculous buy and sell orders on shares you want or own.

[@Who-yo3xh]

SEC not doing its job

[@mdogzino]

This stuff is way too advanced for any regulatory body like SEC to understand.

[@Triple5Sellout]

Alea iacta est 12/3/14

[@username42]

is it possible in crypto binance ?

[@lerch25]

Does this all hit anyone else as more than a little predatory and slimy?

[@BYSMA4th]

I certainly think so, not only do these algorithms compete with each other, but they almost make mincemeat of the human trader with small investments. If an event as devastating as 9/11 happened today, computers would sell faster than a human would receive the news of the event.

[@timoj1044]

BYSMA4th That's exactly my feelings on the markets, and that's why I'm a programmer.

[@brandonzimmer4286]

***** Amen brother

[@CzechRiot]

+BYSMA4th That's why I stopped trading.

[@Evan-McM]

+BYSMA4th And so what? What's wrong with that? Valuations of a financial market should not be based on morality or emotion, they should be based purely on a supply-demand basis.

[@timvincent8649]

If we all understand the laws and rules of this universe, then we don't need artificial intelligence.

[@charless8707]

He was saying that algos are taking over markets and we are no longer living in the same world (the nanoseconds world). Then to take on that world we need some super human type of exoskeleton.... in the form of artificial intelligence (which is another form of algorithm) and still humans are limited to 650-900 milliseconds. Does he really think any human can follow what that artificial intelligence is doing at nanoseconds?In his Augmented Intelligence explanation on how to do this, one can only analyze the data AFTER THE FACT through visualizations or math. Which basically what he has done when he is showing charts from NANEX data visualizations. He is not solving anything. I am not defending any of the trading algorithms but I can tell you this, with a rule based trading algorithm, you can more readily understand the rules and ascertain if the rules of those algorithms caused the crash ( which some had proven it was, such as BATS crash). You can then formulate mitigating rules to prevent it. With AI based algorithm it is a black box unless you build into that AI that it can explain its actions. Let's say if it was built upon neural network, you cannot really understand what its rules are and what it is doing, it is even worse. Just because AI is improving a lot does not mean it is the solution for everything. He had said it himself, why should an algorithm which lived in nanosecond world decides our world, yet his solution is to put more algorithm in. The solution is simple, put a speed limit, something that IEX is doing though I think IEX can and should slow down the speed further.

[@alexusanin1632]

TO give you an example of an exoskeleton in this case: instead of having algo trade right away, you have an algo that scans the news and tracks the markets. It foresees the news, foresees the upcoming crash from all other algos that trade, and then notifies you. Yo have 7-10 minutes before market goes back to normal, so you buy low, and sell as soon as price approaches the normal again. If there were microcrashes recorded, they have something in common most likely. So here Algo is not taking decision to trade, but is notifying you that something is coming.

[@EnginAtik]

If we add tax to trade, it acts as resistance and the system becomes Lyapunov stable.

[@benedictamstel3809]

Nobody is a looser only when you click on the right button

[@anonimu0000]

benedict amstel i'm a loser if you click faster than me

[@susancorgi]

The rise of the machine is real! Where is john corner?

[@RaptureReady2025]

He hasn’t been born yet. But I’m dating Sarah Connor right now.

[@mikestein2643]

At ~16 :36 what does he say? ..... it turns out that a human? machine will beat any machine in the world?passive?

[@Kane-ib5sn]

he might be wrong about the 'augmented intelligence theory'. that being, a machine + human is greater than a machine (w/ highest AI). because, if the human isn't at a high enough level conceptually, he/she/they will mess up the optimized algos that are functioning, for a 'subjective' possibility - we call this 'a miracle'...the machines will evaluate it as a probability failure; and defeat you... that's been my experience - trying playing chess with the assistance of a machine, against that same machine which has no impedance...you will lose, unless you're a Super Grand Master... and even then, that's not playing against the highest level AI - namely AlphaZero, or whatever the name is now...

[@sirAlfred1888]

he have elvish ears

[@MrEdwardGotham]

The Geth are coming...

[@jjreddick377]

Skynet will emerge from HFT.

[@mdogzino]

There's other industries that use way more advanced machine learning & algorithms. Like Uber & Facebook.

[@abinashkumar8162]

2024 ai 📈

[@jerryrobinson958]

The volume on this video sucks

[@Zander101084]

i don't think that was really the point.

[@deplorablekulak6823]

Ah disagree with that sublimation

[@verderriscursey]

what does he mean we dont know what happened? Somebody crashed it on purpose and got rich!

[@Raison_d-etre]

This guy hasn't read The Flash Boys.

[@CzechRiot]

And what would he conclude if he did?

[@Raison_d-etre]

+CzechRiot He'd have been a bit more critical of HFTs.

[@CzechRiot]

Reason Apparently they have a limited time to speak and his focus was not in these specifics you're referencing.

[@Raison_d-etre]

+CzechRiot Even so.

[@CzechRiot]

What would you want him to address?

[@fburton8]

And the *purpose* of all this lightning computation? Pffft!

[@CzechRiot]

more booze and hoes.

[@Evan-McM]

+CzechRiot Literally none of the people working at these firms are motivated by these things.

[@CzechRiot]

Evan McMiller maybe at your firm...

[@harendrasingh_22]

10:53 hahahaha.

[@whiskymarko]

HFT is a scma and those people belong in jail

[@ReneeBryant-o1e]

Young Daniel Jackson Paul Wilson William

[@igorschmidlapp6987]

A.I. will make big money to finance their robot construction costs to build the army to dominate mankind... ;-P

[@YoungMerle-f6v]

White William Hall Jose Lopez Sarah

[@wordlesssong]

This is a terrible presentation, completely plagarized from "Financial black swans driven by ultrafast machine ecology" by Johnson et al in 2012.

[@sourabhdash3208]

bull shit.. all those fundamental valuations of companies goes out of the window.. this is just math.. no need for any financial wizards.. wall st at its best.. bull shit