Definitely

I'm surprised such discussions still hold place it is old news that trading algorithms are replacing human traders and will continue to replace as technology develops

Of course, it is already being used in some hedge funds and investment banks.

Ha ha ha Leave me think of it, and there I warn you .. When it's already broken ... Lols..

Why "Vs.", why not "with"?!

The best results are usually achieved when human and manchine complement, not substitute each other.

No, because AI and data science will not be able to converge to an efficient financial market model.

I agree with Vasily Nekrasov : today, when algorithmic trading is pretty primitive , all automated systems can thrive as parasites...and sometimes, the same as real parasites, they can have a really bad impact on whatever they parasite on...

Agree with Shalom. It's a non discussion . AI is on track to disrupt every conceivable area or industry that today has human intervention. The irony to this is , as humans seek to outsmart each other by making redundant Human inefficiencies , we ultimately make any need for human intervention redundant. What are the career paths of the future for our kids? Programming? #tippingpoint

signal to noise ratio and non stationary means are always going to be a problem.

Yes, if hedge fund have evolved to be such in name only, with numerical results similar to other "actively" managed funds. If they cannot even outperform passive funds, they are under an existential threat, with or without AI.

Human behavior is repetitive and so are our mistakes, so (no doubt) we need the machines to get rid of our emotional approach to markets. However, in the past, we still needed the right (human) input for the machine outcome to be successful. Today´s AI will rapidly change the way HF approach the markets.

Thank you everyone for your thoughtful feedback. Indeed, the progress in AI and machine learning technology have advanced leaps and bounds in the recent years due in part by the positive network effects. However, we still haven't had a major breakthrough in the multi-dimensional AI. David Siegel of Two Sigma said last year that AI still lacks common sense and it may be a while until we have an AI running a pure autonomous trading. I can certainly see that more managers will incorporate AI in their trading program, but still have human oversight to avoid systematic crash. Our objective for this panel is for our managers to see that the disruption is on the horizon and they should be prepared to adapt and evolve. Btw, we have yet to see a classic value investor who wants to share an opposing view of AI on this thread....but that's probably because we are in a quant channel! I will summarize our findings on this topic after our conference next week.

I think that it already has. Consider all the traditional active management funds that have shut down over the past few years.

We have been working with machine learning for more than 10 years. Using algorithms and advanced probability strategies. The data and knowledge we gathered is useful for progressing into deep learning and eventually into AI. The road has been long and not without problems. What I have seen in the last three months is mind blowing....the development is rapidly speeding up. The proprietary AI strategies will never be discussed and shared with the public for obvious reasons, however I think we will soon see products that behave very much like traditional funds, smart beta ETFs with fairly simple but effective strategies that are inexpensive.

I swear, I saw somewhere non-learning robot (R) with embedded quant's analytics in historical non-stopped tests cumulatively covered over 400 YEARS at different Financial Markets (FM). Ir produced hundred and thousands net gains per cents. It makes direct relations to this discussion of hedge Funds (HF) surviving. Do not be too modest: human's innovation is and will be much higher than machine due to understanding the causation process in great favor of human innovation. AI speed is a different story, and it has nothing to do with embedded into R analytic quality. The zest of HF performance is in the predictive FM analytic. In other words HF needs tiny-tiny, little-little Albert Einsteins. HFs competitions will be at the quantitative levels of predictive analytics Overall HF will approach to the pure Quant Fund. More interesting things follow from here like Sharpe Ratio value...

I saw a show the other night that progressed the machines to a point that they could "think", which included an understanding of human emotions and more. They learned to a point beyond that which humans could think. Science fiction or not? Much of what GH Wells foresaw has become true. They may go beyond our imagination? Thanks, David