EXACTLY to your last sentence. In the late 80's I worked on a AI manufacturing scheduling system. And it was basically where we did knowledge acquisition of a guy who had scheduled the plant for 20+ years and then developed heuristics and rules to emulate his thinking. Same thing here ... someone has to code the algorithms/rules. And a good trader can still make $$$. Back in the mid 90's when I started trading full time I paid $25 or more each side of a trade. Now commissions are dirt cheap and even with penny spreads I find it easier to make money now.
Trading will be dead in next 10-20 years
- Thread starter trade5656
- Start date
EXACTLY to your last sentence. In the late 80's I worked on a AI manufacturing scheduling system. And it was basically where we did knowledge acquisition of a guy who had scheduled the plant for 20+ years and then developed heuristics and rules to emulate his thinking. Same thing here ... someone has to code the algorithms/rules. And a good trader can still make $$$. Back in the mid 90's when I started trading full time I paid $25 or more each side of a trade. Now commissions are dirt cheap and even with penny spreads I find it easier to make money now.
Agree, but AI will analyze historical data in nano seconds. Decisions will be made at the same speed if not faster (no data/memory required). A human to determine an action may take minutes if not more to achieve a result. I can see of form of AI assisting traders now.
Last edited:
No argument on the speed aspect to analyze and make a decision. I trade mostly on feel, trading the same stocks over and over. I tend to know the trading range so I don't have a lot of data to analyze, mainly gut feel and current market psychology.
Yes, some traders use that method. But if you sat down with a good systems analyst/ programmer, they can decompose your "gut feel" into methods, functions, dependencies, and programs. Reproduce your results and make it more reliable and show you where it may go against you.
first of all many people here don't understand AI... AI is not same as algo/HFT ...
AI is different...
Algo just follows a set of instruction and directly managed by human.
In AI the human codes it to learn and then the AI based on his learning creates his own trading strategies.Human has no control what the AI will come up. Human has control on how the AI will start learning and creates its intelligence.
We have 6 billion human beings and each human is unique. So each AI developed by diff humans will be unique will disagree on pricing and continue to produce a healthy market.
Also there will be many AI failures just like human traders.
to be honest though very early to speculate how it will go down.
AI is different...
Algo just follows a set of instruction and directly managed by human.
In AI the human codes it to learn and then the AI based on his learning creates his own trading strategies.Human has no control what the AI will come up. Human has control on how the AI will start learning and creates its intelligence.
We have 6 billion human beings and each human is unique. So each AI developed by diff humans will be unique will disagree on pricing and continue to produce a healthy market.
Also there will be many AI failures just like human traders.
to be honest though very early to speculate how it will go down.
Doesn't how much processing power is available, it will not solve the problem. Garbage in, garbage out. As long as there are humans, we're going to have a lot of garbage...
It seems that many traders fear AI by thinking, AI will be to humans what humans are to animals. Well, even with all our smarts today we still find it difficult to swat a fly, wipe out cockroaches, or capture a wild monkey in the jungle. So something tells me that we'll survive!
How do mere mortals learn? AI is only as good as it's teacher. Think about it ...