Hacker Newsnew | past | comments | ask | show | jobs | submit | more fspear's commentslogin

would love to watch this, can you share the link?


>what if AI is totally possible, but our sacks of meat in skulls are not able to achieve it?

Won't be a problem if we use CRISPR to make ourselves smarter first!


Keep a portion of your eth and invest in promising ICOs.


napoleonx.ai


>That's actually a pretty strong compliment. One could reverse it and ask, why in America and Europe people are easy swayed by lies in "a catchy headline or a pretty image", instead of demanding actual information about the product?

Makes sense since Japanese culture is risk averse.


Do you have any examples of these ML models? I wonder if similar models can be applied to pick ICO's worthy of investment.


https://motherboard.vice.com/en_us/article/zmep4y/mit-resear...

https://qz.com/561206/algorithms-make-better-hiring-decision...

I don't think it would work for ICOs, assuming that they're open to the public. As we've seen, prices seem to be pretty irrational when everyone is trying to jump on the hype train.



Can you elaborate on the technical/scalability limitations you have seen in some of these coins and how to look out for these?


Any tips for noobs?


Get yourself a small linux box, install python, get some back data and start testing any trading idea you ca come across the web for free or from books form the library - all of them will fail, but it will give you other ideas to test and one day hopefully you'll find your edge... From there it's easy to automate...


And recognize you are up against the smartest brains in the world on discretionary and automated basis, some of whom can transact at sub-second frequency. Unless I see audited results from a reputable firm, everyone is full of BS. Don't believe everything you read.

~~a 25 year veteran of the markets, hedge funds, CTAs, equities, commodities, you name it.


EDIT: it’s not easy, but also it’s not easy to become a doctor, professional athlete or a rock star either


Any reason not to use a service like Quantopian?


Maybe I am paranoid due to my "real job" in systems/security, but here is why I would never use a service which will hold my trading strategy algo, especially in the cloud because:

1. Those are startups with limited funding and not too many employees, so they are the perfect hackers target for the perfect crime, the hacker don't have to steal anything, just copy the best strategy from time to time and use it until obsolete or just bet against you for example...

2. I don't see a 3rd party verification and audit on what exactly the employees of those companies have access to - if I am a rogue hedge fund, I have to compromise only one employee in such a company and ask him to send me all best trading strategies every month for example...

3. Hedge funds are major investors in such companies, the question is why and what information they are getting...


Can you ellaborate? Are there any good testing set up platforms or tutorials?


Are you able to share any resources that you'd think would be helpful to get someone started building stock trading robots and strategies?

Thanks in advance.


Just focus on the major stock exchanges NYSE, AMEX and NASDAQ, do not waste time with crypto, options and other distractions... good luck!


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: