Do we need Artifical Intelligence in trading? | Page 2 | Crypto Craft

Options

Go to
Permalinks
Print Thread

Similar Threads

Pipimint Fresh Chart Sharing-Where Feedback Flows & Intelligence Grows 34 replies

Artifical Intelligence EA 28 replies

Artificial Intelligence EA 16 replies

Attachments: Do we need Artifical Intelligence in trading?
Exit Attachments

Do we need Artifical Intelligence in trading?

  • #21
  • Sep 20, 2017 7:46am Sep 20, 2017 7:46am
Quoting jdieieh
{quote} Thats is right. I will take at least couple of years training for a person to paint like this robot. And this robot paints with 5 brushes at once, pretty good portraits for my eyes! So... if trading is art, where this leaves us? So... if trading is math, where this leaves us? So... if trading is analytical skills, where this leaves us? And... what if machines change trading that is not art, not math, not analysis any more? Who of us will understand advanced AI Goldman Sachs will use. It will not be publish in the books. And even if you could...
As I mentioned in my first comment, there are some attempts to use AI in the trading world, but not to make the trades itself, but more like being a follower. Maybe Goldman Sachs is one of them. Another one that I know is SixCap ...
Or, have it ever crossed your mind that "AI" is just a marketing jargon to lure customers in? ...
Market is not random but unpredictable
  • #22
  • Sep 20, 2017 7:46am Sep 20, 2017 7:46am
Quoting The-Flipper
{quote} Did you? There is nothing new under the sun. Trading was, is and will always be the same. Unless you have access to a crystal ball to foretell the future.

Garry Kasparov - No. 1 rated player in the world from 1986 to 2005 - that is 19 years.

He said: "You can’t overestimate the importance of psychology in chess, and as much as some players try to downplay it, I believe that winning requires a constant and strong psychology not just at the board but in every aspect of your life."

Deep Blue did NOT have any psychology and beat Kasparov many times. Which means psychology in chess was not actually important at all, as the master though. Which means the master was actually wrong on what really matters in chess, despite his experience and undisputed authority at the time.

Just drawing analogies... analogies, because I cannot tell you exactly how it will look. But I tell, most likely, it will look completely different to what masters of today tell us. Perhaps for the same reason why Turtles method does not work that well any more
  • #23
  • Sep 20, 2017 7:52am Sep 20, 2017 7:52am
Quoting 4for4
{quote} Or, have it ever crossed your mind that "AI" is just a marketing jargon to lure customers in? ...
Have you seen work of Boston Dynamics and alike? Have it ever crossed your mind that actually it might work like the Terminator and it wont be just a sci-fi movie. As Jules Verne's submarine fantasy is actually reality today. Most of those concepts are already alive and working on battle fields.

How much battle field has changed from 100 year ago, when a peasant took his rifle and went shooting. Today is not even heavily armored soldier with self aiming rifle that matters. But a swarm of armed drones able to cooperate and ambush. War was always leading in technology.

The same process will be happening here. No tomorrow, not next year, but it will...

Dont think in terms of today. As I said, today is AI embryo. Wait till it grows.
  • #24
  • Sep 20, 2017 7:56am Sep 20, 2017 7:56am
Quoting jdieieh
{quote} Have you seen work of Boston Dynamics and alike? Have it ever crossed your mind that actually it might work like the Terminator and it wont be just a sci-fi movie. As Jules Verne's submarine fantasy is actually reality today. Most of those concepts are already alive and working on battle fields. How much battle field has changed from 100 year ago, when a peasant took his rifle and went shooting. Today is not even heavily armored soldier with self aiming rifle. But swarm of armored drones able to cooperate and ambush. The same process will...
We're talking AI for trading, aren't we? ...
Market is not random but unpredictable
  • #25
  • Sep 20, 2017 8:06am Sep 20, 2017 8:06am
Quoting 4for4
{quote} We're talking AI for trading, aren't we? ...
We understand that progress is introduced at all fields at once, with war being leading edge, right?

First functional airplanes were used to spy on enemies positions, shortly after followed to bombard them and kill. Only when war finished, RICH people could used them to travel much faster then even before. Today a backpacker with two bucks in the pocket can do the same.

We can draw analogies and project, right? Is isn't how technical trend trending supposed to work, projecting?
  • #26
  • Sep 20, 2017 8:26am Sep 20, 2017 8:26am
Quoting jdieieh
{quote} Is isn't how technical trend trending supposed to work, projecting?
Nope. If technical trend analysis works as it's supposed to be, everybody will become a successful trader ...
Referring to the title of this thread "Do we need AI in trading?", my definite answer is "NO" ... Those who are already able to trade profitably don't need any artificial intelligence as their instincts and intelligence have been "honed" to a point that they know when to wait and when to act ...
Market is not random but unpredictable
  • #27
  • Sep 20, 2017 8:37am Sep 20, 2017 8:37am
Quoting 4for4
{quote} Nope. If technical trend analysis works as it's supposed to be, everybody will become a successful trader ... Referring to the title of this thread "Do we need AI in trading?", my definite answer is "NO" ... Those who are already able to trade profitably don't need any artificial intelligence as their instincts and intelligence have been "honed" to a point that they know when to wait and when to act ...
The original question was not the most fortunate. More appropriate one is: "Will we need AI in trading?" and the answer is: "Definitely YES". In relatively short future. Surely before most of reader will grow a single gray hair.

Those who speculate today and will not adapt, will not be able to speculate anymore successfully. And you will not know when to wait and when to act, because rules will be changing fluidly based feedback of masses behaviour at the time being, which means most likely you.

For the same reason as you, being an excellent marathon runner cannot possibly beat a 200-pound couch-potato in a car on 100 km run. Those who have imagination, understand. Those who dont, are relatively safe, for now...

It always has been like that and always will be. That one thing never changes
  • #28
  • Edited 2:24pm Sep 20, 2017 1:59pm | Edited 2:24pm
I look at facts, every time I put my hands on the computer I lose.
  • #29
  • Oct 5, 2017 5:50am Oct 5, 2017 5:50am
Although I am a beginner in AI, but as far as I know about AI it has shown positive impacts on every industry. If we are going to use AI (which many big firms have already started) in trading, it certainly can improve the results by learning from the trends, inputs, previous experience.
  • #30
  • Oct 5, 2017 8:28am Oct 5, 2017 8:28am
Quoting deeps001
Although I am a beginner in AI, but as far as I know about AI it has shown positive impacts on every industry. If we are going to use AI (which many big firms have already started) in trading, it certainly can improve the results by learning from the trends, inputs, previous experience.
Oh, finally some discussion is back. I started to think that nobody in this forum is willing to learn about future competitors

However I do not agree on "positive" impact on every industry. It is actually quite opposite. Let me give you couple examples to illustrates:

1. About 10-20 years ago, your typical admin/clear workflow was replaced by semi-automated processes. In many cases it meant admin team had to be cut in half or quarter. Some of those people could re skill and get other, but lower paid jobs somewhere else. Lower paid, because the same process was happening elsewhere as well. Some of them skilled up and got new job, but...

2. 5 years ago that higher level of jobs have been replaced with automation as well. Let me give you example: advance accounting skills has been replaced with software like Xero as well and automated taxation office processes and integration. Lawyers has been replaced in basic research and advisory jobs by cleaver IT expert systems. Some move up for more specialize projects, but there is only so many specialized projects available. General practitioner doctors has been replaced by internet knowledge, so they dont have nice steady flow of patience anymore, unless they specialize (and pay hundred of thousands dollars and slave off as interims for years) . In general medium level professional and management jobs are being replaced with embryo-AI.

3. In next 5 years driver, shop assistants, brick layers, IT support guy, para-legal, medium management ect ect ect will be gone. Most of government clerks will be gone as well. They will be replaced by cleaver new IT system provided by external contractors.

In my opinion social impact of semi-advanced AI will be very negative. This type of AI will good enough to take out job and mean of living, but will no be able to provide us cheap essential stuff we need to live. Bounty will be taken by few on the top (as today, but even worse) and the rest will living in semi-poverty.

Obviously there will be opportunities for well skilled, educated and creative people, but it will not provide opportunities for all those who loss their way of earning, only for some (and in general for lower benefit). There will be very narrow class of super rich beneficiaries (1%), supporting class of professionals and knowledge workers (10%) and every one else being very poor (89%).

Job will come back to America (or to you country), but it will be done by robots. This trend happens everywhere, pushing people into so called "shared-economy and gig-economy", which is really euphemism for low paid, casual, temporary jobs. No steady income, few good opportunities available and those available mostly already exploited for peanuts with huge social and economics consequences.

In a sense, change from the current information age to new parading will be similar to change from the the industrialist to information age. Let me tell you, that was not peaceful change. It was paid by sweat and blood. And two global wars, the last one with potential to annihilate the whole civilization (and counting)

No, AI advancement most likely will not be positive. in short and medium term...
  • #31
  • Oct 22, 2017 1:09am Oct 22, 2017 1:09am
...AlphaGo Zero, the latest evolution of AlphaGo, the first computer program to defeat a world champion at the ancient Chinese game of Go. Zero is even more powerful and is arguably the strongest Go player in history.

Previous versions of AlphaGo initially trained on thousands of human amateur and professional games to learn how to play Go. AlphaGo Zero skips this step and learns to play simply by playing games against itself, starting from completely random play. In doing so, it quickly surpassed human level of play and defeated the previously published champion-defeating version of AlphaGo by 100 games to 0.

https://deepmind.com/blog/alphago-ze...rning-scratch/
  • #32
  • Oct 23, 2017 10:22am Oct 23, 2017 10:22am
Quoting Ulule
Hi, I don't want to start a thread on specific trade system. I only wonder if we - human traders - really have something to do in biz yet. I don't even think about HF systems because I don't think they are AI. But there is no doubt that there is a real threat that humans will be replaced with algorithms in next few years in many even in hand-managed branches as mutual funds. How to prepare to this new conditions when human will be competing with more and more sophisticated algorithms? Would like to see your opinions. Are we threatened by new tech...
I somehow don't think AI would ultimately replace certain aspects of trading. It will certainly help a lot, but it would be hard to completely replace us humans because certain human discretions and gut feels you simply can't replicate in a machine. If the entire wall street has a doom and gloom feeling around it, you'd feel that but not the machines. I even heard they're trying to get AI to start being more human with sentiment analysis and less robot haha.
  • #33
  • Oct 24, 2017 4:16am Oct 24, 2017 4:16am
Quoting ForexLLC
{quote}...(AI) would be hard to completely replace us humans because certain human discretions and gut feels you simply can't replicate in a machine. If the entire wall street has a doom and gloom feeling around it, you'd feel that but not the machines.
You feel "the entire wall street has a doom and gloom feeling around it"??? How do you "feel" it? By some kind of cosmic, communal, Avatar-like connection?

Or perhaps by social, digital and indeed market clues. You open a twitter on your mobile and read a comment or two. You talk with you friend next trading desk and ask him opinion. Then you looks a price and make testing order. Then you manager goes to lunch, hear some more gossips and ask you to to put more order. The more news are coming in and more people are reacting to it. And you are observing that and reacting to it.

That is how you "feel" the entire wall street has a doom and gloom feeling around it, right? By absorbing information and observing action.

Why exactly AI could not do it? What stops AI read the same news and tweets you do? It has to 1000 times faster and in 1000 larger volume than you. What stops your friend at trading desk to input his opinion into the system? What stops your manager to input even more information into the system? AI can observes all those multitude inputs and process almost instantly.

Can you?

I will repeat words by Lee Sedol after loosing against AlphaGo, who in his field in much much higher experienced and ranked than you in trading world: " the moves the computer played would never happen in a human-to-human match."

Think about this, it was just a Go AI embryo and beat the human champion, the one mastering the game for 20+ years, without any problem. That human, Lee Sedol, is ranked 3500 ELO now and so other human champs. And that AI embryo has been just beaten by a new one, AlphaGo Zero, 100:0! 100:0!!! The new AlphaGo Zero has ELO rating 5000+, all by playing with itself and learning. No human help. From scratch. It took it 2 days to master what a human took 20+ years.

Think again...
  • #34
  • Oct 27, 2017 5:04am Oct 27, 2017 5:04am
... you say: chess or go are easy fields for AI to beat human... lets try something radically different..

Valentino Rossi - one of the most successful motorcycle racers of all time, with 9 Grand Prix World Championships to his name – seven of which are in the premier class.

Yahama - (among others) a well known motorcycle manufacturer

"Yamaha has revealed its new humanoid motorcycle-riding robot at the Tokyo Motor Show. Its goal? To beat MotoGP world champion Valentino Rossi. Motobot is designed to approach riding "from a human rider's standpoint -- with no modifications made to the motorcycle itself".

When that Motobot beats 9 times human champion (and it will sooner or later), in racing, on real track, on real sport motorcycle... will you start believe that in few years your well-honed and sweat-obtained speculating skills in forex will become simply obsolete? Or at least consider your plan for that outcome? Or are you going to wait till AI will shake you out of your capital?

Think again...

http://www.wired.co.uk/article/yamah...alentino-rossi
  • #35
  • Nov 1, 2017 4:28am Nov 1, 2017 4:28am
Quoting Ulule
Hi, I don't want to start a thread on specific trade system. I only wonder if we - human traders - really have something to do in biz yet. I don't even think about HF systems because I don't think they are AI. But there is no doubt that there is a real threat that humans will be replaced with algorithms in next few years in many even in hand-managed branches as mutual funds. How to prepare to this new conditions when human will be competing with more and more sophisticated algorithms? Would like to see your opinions. Are we threatened by new tech...
The thing about artificial intelligence is that it cannot 'create or imagine' anything new on its own.Only a human brain can do that.Markets as they say are dynamic.and therefore markets are govern by some unknown hidden laws of nature sure it can help us in speeding up the calculations,faster trading or high speed trading,FASTER calculations of riskreward and probability of an event.but it cannot help us to predict the outcomes of any particular event in the stock markets or forex by anything more by chance.Simply because the markets we are trading are based on a natural laws.and artificial intelligence is not intelligent enough to find out those laws which governs the markets.SO there is nothing to be threatened about them.Period.
No one ever Discovered anything new by colouring inside the lines.
  • #36
  • Nov 1, 2017 5:04am Nov 1, 2017 5:04am
Quoting Hdhorda4
{quote} The thing about artificial intelligence is that it cannot 'create or imagine' anything new on its own... ...SO there is nothing to be threatened about them.Period.
You are dead wrong on your very first sentence

"DeepMind researchers give a short review of “a new family of approaches for imagination-based planning.” The so-called Imagination-Augmented Agents, or I2As, use an internal ‘imagination encoder’ that helps the AI decide what are and what aren’t useful predictions about its environment.

The researchers argue that giving AI imagination is crucial for dealing with real-world environments, where it’s helpful to test a few possible outcomes of actions ‘in your head’ to predict which one is best."

https://thenextweb.com/artificial-in...magine-future/

I would also comment on the rest of your premise, but that would distract and derail the core argument of your belief.

So you can either continue believe in your subjective opinion about creativity and imagination and keep sticking to an imaginary model of the reality, or you can refer to an existing, practical example of AI imagination (and further theoretical ground work) to reassess your belief
  • #37
  • Nov 3, 2017 5:20am Nov 3, 2017 5:20am
Something hard to do for a machine is the "single experiment learning". When a baby ape sees his mom break a nut with a stone for the first time he can now do it by himself. Not only he can repeat the operation but he also learnt that he can break any nut with a stone. Learning algorithms average huge amounts of data to achieve the same level of generalization. Google used hundreds of billions of images. AlphaGo Zero had to play against itself to generate datasets. For trading single experiment learning is an important capability because we lack data. Given the very high interconnection between all the markets we can argue that the correlation is so high that we have one single dataset. Especially when you only focus on FX.
No greed. No fear. Just maths.
  • #38
  • Dec 2, 2017 4:35pm Dec 2, 2017 4:35pm
How many of you can do standing back flip? You have been using your skeleton and muscles for 20+, 30+ or more years. That is enough practice. Can you do it right now, like this robot? Perfect back-flip, from a standing position, on a pedestal, without fear of falling down?

Inserted Video
  • #39
  • Dec 2, 2017 6:39pm Dec 2, 2017 6:39pm
Quoting jdieieh
{quote} The original question was not the most fortunate. More appropriate one is: "Will we need AI in trading?" and the answer is: "Definitely YES". In relatively short future. Surely before most of reader will grow a single gray hair. Those who speculate today and will not adapt, will not be able to speculate anymore successfully. And you will not know when to wait and when to act, because rules will be changing fluidly based feedback of masses behaviour at the time being, which means most likely you. For the same reason as you, being an excellent...
AIs are already impacting society and trading. Today's markets are already very different from the ones we've seen 20 years ago, thanks to algos, and many jobs have already been replaced by automation. Once general-like AIs will become common the world will change drastically, and I'm frankly afraid of the future we're about to experience because probably this won't turn out very well for the average joe.
  • #40
  • Dec 2, 2017 11:16pm Dec 2, 2017 11:16pm
Quoting R3ta
{quote} AIs are already impacting society and trading. Today's markets are already very different from the ones we've seen 20 years ago, thanks to algos, and many jobs have already been replaced by automation. Once general-like AIs will become common the world will change drastically, and I'm frankly afraid of the future we're about to experience because probably this won't turn out very well for the average joe.
The good thing is if such big moves does move. It will carry our trade forward. The bad part is does it stay there or will it just spike. If AI does intervene. Spikes will happen and then flat line. Retail traders have no pattern nor form to trade.

Having said that. Patterns are what AI are program to look for and intervene.

Tip if spike does happen. And a high wick appears. Most probably price will return back to that area to retest. At least this is what I see many times.
You're going to pay a price for every bloody thing you do and don't do
1 trader viewing now