Could AI / Machine learning come up with the method the way you have ?

[traderwald]

Hi guys,

As many say, you have to work many hours (and more) to come up with methods or system for trading. Is it correct to assume that ultimately the system or method involves a pattern or sequence or patterns that across various factors that are combined. (especially given that many are automated methods)

If so, would it be correct to say that any pattern that any trader has figured is something that AI / Machine learning system, created to identify such patterns and the possible profit percentage from such trades, would also have figured out.

If so:

1. What are the implications for trading in general ?

2. What would you comment on the efficiency of an arena where some people do hours and hours of manual work, while the AI system figures in seconds something that all the combined effort many successful traders takes hours of work, practice and automation to create ?

3. What do you think when you know that while you putting all the hours you put to create the system, the AI is doing it in seconds.

4. How to have an edge in an arena where AI capable machines are in the field.

 

[gaussian]

Oh look it's today's ML is eating the world post.

1. What are the implications for trading in general ?

Better fills. You aren't competing with MMs and desks using ML in any real capacity. The amount of data it takes to properly train an ML algorithm is out of reach of most retail traders.

2. What would you comment on the efficiency of an arena where some people do hours and hours of manual work, while the AI system figures in seconds something that all the combined effort many successful traders takes hours of work, practice and automation to create ?

ML models cannot be explained easily. Once you leave regressions and various types of hyperplane optimization (GLMs, SVMs) you get to "deep learning" which is basically impossible to explain. It's a blackbox you feed data to and hope it works right.

3. What do you think when you know that while you putting all the hours you put to create the system, the AI is doing it in seconds.

You don't know that. Additionally, timeframe wise, you probably aren't competing with ML anyway. Who cares.

4. How to have an edge in an arena where AI capable machines are in the field.

Don't suck. You should be more worried about the "monkeys with typewriters" problem places like Robinhood have introduced to the market.




As a side note can we please stop calling machine learning artificial intelligence. Machine learning is just mathematical optimization. Real AI, the kind people worry about, won't come from some probabilistic model.

 

[schizo]

Hi guys,

As many say, you have to work many hours (and more) to come up with methods or system for trading. Is it correct to assume that ultimately the system or method involves a pattern or sequence or patterns that across various factors that are combined. (especially given that many are automated methods)

If so, would it be correct to say that any pattern that any trader has figured is something that AI / Machine learning system, created to identify such patterns and the possible profit percentage from such trades, would also have figured out.

If so:

1. What are the implications for trading in general ?

2. What would you comment on the efficiency of an arena where some people do hours and hours of manual work, while the AI system figures in seconds something that all the combined effort many successful traders takes hours of work, practice and automation to create ?

3. What do you think when you know that while you putting all the hours you put to create the system, the AI is doing it in seconds.

4. How to have an edge in an arena where AI capable machines are in the field.

1. The implication is HUGE, even though it's in its infancy. However, once it reaches the level of Artificial General Intelligence, it is believed that the machine are able to "intellectualize" (eg. able to think and likely to code/program on its own).

2. As humans, there's no way you can compete against the machines. Well, unless you plug into Elon Musk's "Matrix". *Scarry thought*

3. Just as those who traded down in the Chicago futures pits are no longer around, almost all of today's traders who fail to adapt will perish as well.

4. When the day comes, it will not be AI vs humans, but AI vs AI. Hence, you need to have your own AI system. Start learning Python.

 

[guru]

Yeah, that's like a taxi driver asking whether self-driving cars will be a threat to them.
How can you be a professional at anything if you have to ask such questions?
The pros who have a real edge do not walk around asking whether machines will replace them. They focus on being the best at what they do, adapt to new situations when needed, come up with ideas to solve problems, and when all fails they understand why and move on.
Trading itself is very competitive and you cannot master it by being able to do only one thing and easily get eliminated. If you can't put up a good fight then you're not worth being a trader either. While if you're really good then you'll succeed in anything you do, whether trading manually, or creating technology for trading, or doing anything else.
I never seen any pros asking such questions, and for good reasons.

 

[Specterx]

AI as it currently exists isn't a threat to human traders - just another weapon in the ongoing HFT/quant arms race, in which human traders have never been able to compete directly. Present forms of AI are structurally incapable of matching the flexibility and adaptability of a skilled human trader, and it will remain so for the foreseeable future.

 

[speedo]

As long as there is a steady supply of natural unintelligence in the markets, I'm not worried about it.

 

[Laissez Faire]

If so, would it be correct to say that any pattern that any trader has figured is something that AI / Machine learning system, created to identify such patterns and the possible profit percentage from such trades, would also have figured out.

Not the stuff I'm using particularly since I'm using datasets (not raw data) that AI/ML Systems do not have access to in the first place.

I actually hired a programmer to run some ML algorithms on my stuff and he didn't come up with anything I could use and actually did not discover some of the stuff/relationships I already use.

Maybe some time in the future.

 

[trend2009]

Oh look it's today's ML is eating the world post.



Better fills. You aren't competing with MMs and desks using ML in any real capacity. The amount of data it takes to properly train an ML algorithm is out of reach of most retail traders.



ML models cannot be explained easily. Once you leave regressions and various types of hyperplane optimization (GLMs, SVMs) you get to "deep learning" which is basically impossible to explain. It's a blackbox you feed data to and hope it works right.



You don't know that. Additionally, timeframe wise, you probably aren't competing with ML anyway. Who cares.



Don't suck. You should be more worried about the "monkeys with typewriters" problem places like Robinhood have introduced to the market.




As a side note can we please stop calling machine learning artificial intelligence. Machine learning is just mathematical optimization. Real AI, the kind people worry about, won't come from some probabilistic model.

how did you find your strategy? machine helped or just from manual?

 

[trend2009]

Yeah, that's like a taxi driver asking whether self-driving cars will be a threat to them.
How can you be a professional at anything if you have to ask such questions?
The pros who have a real edge do not walk around asking whether machines will replace them. They focus on being the best at what they do, adapt to new situations when needed, come up with ideas to solve problems, and when all fails they understand why and move on.
Trading itself is very competitive and you cannot master it by being able to do only one thing and easily get eliminated. If you can't put up a good fight then you're not worth being a trader either. While if you're really good then you'll succeed in anything you do, whether trading manually, or creating technology for trading, or doing anything else.
I never seen any pros asking such questions, and for good reasons.

the question is legitimate since technology advances fast. even if AI has not totally replaced manual traders, it has already replaced lots of market makers or traders on sell side. some hedge funds have already employed AI to make money. if more hedge firms can make money using AI, more people on the other side will lose, and professional traders could be one of them. how many professional traders before internet era has faded out when online trading becomes the mainstream?

 

[murray t turtle]

As long as there is a steady supply of natural unintelligence in the markets, I'm not worried about it.

%%
Exactly.
That + a human mind creates machines;NOT vice versa.
And a billionaire trend folloWer noted what some call ''luck'' LOL, i call a small sample...................................................................................................................