Algorithms can learn to collude.
http://www.businessinsider.com/robots-colluding-to-manipulate-markets-2015-4
http://www.businessinsider.com/robots-colluding-to-manipulate-markets-2015-4
WHEN ROBOTS COLLUDE: Computers are adopting a legally questionable means to crush the competition
- Thread starter Banjo
- Start date
Cool--- yins best learn how to program and leave the old fashion "chart" methods in the past or the future's not too bright...
Do you want to say that we entered in to a era where computers can think of a way which is illegal to all the other manual competitors because they are programmed to do so, or they are killing their competition by them selves with out manual help?
In either cases its a high alert to all the human who want a peaceful life.
In either cases its a high alert to all the human who want a peaceful life.
It's a zero sum game, once the bottom feeding quants begin to win too much the frequent human traders will simply throw in the towel; you already have this well underway with the idea of buy and hold ingrained into the minds of the public.
http://www.bloomberg.com/news/videos/2015-03-17/predicting-the-markets-how-ai-can-help
A future say a decade from now when Aidyia starts to run high level AGI, will likely be a future of Vanguard type funds which have 60-day repurchase time limits and which could remove themselves from the exchanges altogether, along with taking private all of their holdings ala Buffet.
Last edited:
This is simply alarmist at best. The computer does not learn to collude. It's either taught from the beginning how to do that or it does it purely based on coincidental factors. Until algorithms have the ability to learn (not statistically learn, but truly learn) and reimplement themselves (*at a core level*) on the fly I'm not too worried. In fact, algorithms gaming off the input of other algorithms can result in highly volatile feedback driven moves that one can take advantage of.
1.7M$ biology textbooks and 400$ Uber rides are good examples of that. The only problem is you can't "go long" on local Uber rates.
1.7M$ biology textbooks and 400$ Uber rides are good examples of that. The only problem is you can't "go long" on local Uber rates.
Very insightful, thank you for contributing. Are you Ben? surf
HONG KONG FIRM TAKES NEXT STEP TOWARD MACHINE DOMINATION OF WORLD FINANCIAL MARKETS
Published in News on December 16, 2014.
Aidyia Limited, a Hong Kong based asset manager, will use their uniquely advanced Artificial General Intelligence (AGI) technology to launch a US equity long/short fund in the first half of 2015. The launch results from a three-year effort, led by AGI pioneer Dr. Ben Goertzel, in adapting the technologies to the complexities of the financial market domain.
AGI is a branch of artificial intelligence aimed at machine learning mimicking the human brain’s breadth, depth and generality of understanding. Applied to financial markets, the result is financial prediction and trading systems with a human like ability to not only recognize mathematical patterns in market data, but to understand what these patterns mean in a broader context. Without the human limitations of memory, emotion, bias and mood, AGI systems can apply their financial insights in an objective way.
“One key advantage of our technology, over previous applications of AI
in the financial markets, is its unique ability to see the big picture and incorporate this in its analysis. This comes directly from the general scope of our system’s intelligence –- its ability to transfer what it’s learned from one domain to another”, says Dr. Goertzel, Chairman and Chief Scientist of Aidyia Limited. “Our use of AGI – Artificial General Intelligence – methods enables our systems to understand market data in ways that aren’t overfit to particular asset classes or market regimes. Our algorithms go far beyond just looking at technical patterns in price data –- they look at news in multiple languages, fundamental and economic data, and price and volume data from various markets. Our AGI system combines these various factors together in complex ways of their own devising, forming multiple predictive models giving information about a stock’s price movement in the next week or month.”
The company believes the technology will prove successful in other assets classes other than equities and could ultimately replace traditional fund management decisions currently made by humans. “Algorithmic trading already dominates high frequency trading decisions, leaving humans highly disadvantaged,” notes Ken Cooper,
the CEO of Aidyia. “Now is the time for longer-term decisions to be disrupted in the same way.”
Dr. Goertzel is a leader in the global AGI research community, and has been advancing AGI technology for over a decade with his research team, also employed by the company. Cassio Pennachin, the Chief Technology officer and long time colleague of Dr. Goertzel, has designed the portfolio and risk management infrastructure that utilizes the company’s AGI technology. Aidyia’s CEO, Ken Cooper, who has worked in Asian and US financial markets since 1995 with Lehman Brothers, Cargill, and BNP Paribas, will oversee the investment process, risk management as well as marketing of the fund.
The company has a base of operations in Hong Kong and plans to launch a series of funds. “Ultimately, we want to launch funds targeting all liquid asset classes, including stocks, bonds, currencies, commodities and derivatives markets”, said Mr. Cooper.
Founded in 2011, Aidyia Limited is a licensed asset management company by the Securities and Futures Commission of Hong Kong. The Company uses advanced artificial general intelligence technology to identify patterns and predict price movements in financial markets. A small group of private investors and the Hong Kong government’s
SERAP program provide funding for the company.
Post navigation
← ITC provides Aidyia with $4.7 million SERAP research funding
Published in News on December 16, 2014.
Aidyia Limited, a Hong Kong based asset manager, will use their uniquely advanced Artificial General Intelligence (AGI) technology to launch a US equity long/short fund in the first half of 2015. The launch results from a three-year effort, led by AGI pioneer Dr. Ben Goertzel, in adapting the technologies to the complexities of the financial market domain.
AGI is a branch of artificial intelligence aimed at machine learning mimicking the human brain’s breadth, depth and generality of understanding. Applied to financial markets, the result is financial prediction and trading systems with a human like ability to not only recognize mathematical patterns in market data, but to understand what these patterns mean in a broader context. Without the human limitations of memory, emotion, bias and mood, AGI systems can apply their financial insights in an objective way.
“One key advantage of our technology, over previous applications of AI
in the financial markets, is its unique ability to see the big picture and incorporate this in its analysis. This comes directly from the general scope of our system’s intelligence –- its ability to transfer what it’s learned from one domain to another”, says Dr. Goertzel, Chairman and Chief Scientist of Aidyia Limited. “Our use of AGI – Artificial General Intelligence – methods enables our systems to understand market data in ways that aren’t overfit to particular asset classes or market regimes. Our algorithms go far beyond just looking at technical patterns in price data –- they look at news in multiple languages, fundamental and economic data, and price and volume data from various markets. Our AGI system combines these various factors together in complex ways of their own devising, forming multiple predictive models giving information about a stock’s price movement in the next week or month.”
The company believes the technology will prove successful in other assets classes other than equities and could ultimately replace traditional fund management decisions currently made by humans. “Algorithmic trading already dominates high frequency trading decisions, leaving humans highly disadvantaged,” notes Ken Cooper,
the CEO of Aidyia. “Now is the time for longer-term decisions to be disrupted in the same way.”
Dr. Goertzel is a leader in the global AGI research community, and has been advancing AGI technology for over a decade with his research team, also employed by the company. Cassio Pennachin, the Chief Technology officer and long time colleague of Dr. Goertzel, has designed the portfolio and risk management infrastructure that utilizes the company’s AGI technology. Aidyia’s CEO, Ken Cooper, who has worked in Asian and US financial markets since 1995 with Lehman Brothers, Cargill, and BNP Paribas, will oversee the investment process, risk management as well as marketing of the fund.
The company has a base of operations in Hong Kong and plans to launch a series of funds. “Ultimately, we want to launch funds targeting all liquid asset classes, including stocks, bonds, currencies, commodities and derivatives markets”, said Mr. Cooper.
Founded in 2011, Aidyia Limited is a licensed asset management company by the Securities and Futures Commission of Hong Kong. The Company uses advanced artificial general intelligence technology to identify patterns and predict price movements in financial markets. A small group of private investors and the Hong Kong government’s
SERAP program provide funding for the company.
Post navigation
← ITC provides Aidyia with $4.7 million SERAP research funding
They are talking about computers selecting the optimal strategies based on some kind of game theory strategy analysis.
This isn't collusion but rather 100% rational behavior which creates the optimal outcome for the whole given the scenario (not necessarily in the best interest of all).
This isn't collusion but rather 100% rational behavior which creates the optimal outcome for the whole given the scenario (not necessarily in the best interest of all).
You are correct, we have AI and its limited to the programmers instructions. You are describing is what Ben Goertzel terms AGI (Artificial General Intelligence). AGI could be programmed to drive a car, but then by adapting those skills and learning it could ride a motorcycle. The distinction here is that AGI could solve a problem without being programmed for the task; one example is figuring out how to use a coffee grinder and make a pot of coffee.
Most of what I have read suggests in as little as 20 years (if resources are devoted) we could reach a singularity with AGI on par with human intelligence; however, whether thats true or not, just imagine in as little as 10 years what Ben could do to the stock market with a more basic AGI? We could be heading into a world were they simply have to shut the exchanges down.
Last edited:
Glad you found the video insightful; Ben is a leader in the AGI community so I have followed for a while, this new fund should be fascinating.
Last edited: