data for feeding AI to predict price?

[zartblu]

Are there any companies specializing on providing data you can use for AI to predict priceses? I mean more than 'historical stock pricings' including soft factors such as how often it BTC got searched in Google and how the ecosystem in general develops?

Would anybody fund an ICO doing this?

[ragnar0k]

I'm using barchart for NASDAQ and feed that to AI.
I know barchart now is offering shitcoins and BC data as well, so I suggest checking them, price is affordable if you are willing to pay (in alternative, you can use APIs from coinmarketcap)
Unfortunately my experience with AI was a bit mixed as market data is really noisy, if you have to pay commissions on top its hard to make a living

[digaran]

Are there any companies specializing on providing data you can use for AI to predict priceses? I mean more than 'historical stock pricings' including soft factors such as how often it BTC got searched in Google and how the ecosystem in general develops?

Would anybody fund an ICO doing this?

Yes Wings Bancor, Legs, Arms and some other already sold in ICO platforms are doing this, however there are no actual predictions but another coin to trade on exchanges giving the chance to devs to earn more money. I had a failed attempt at becoming the next Satoshi myself a few months back, I was really close to convince people to list my "activity" token on their exchanges. are we going to hear from you in the coming days? if so then let me beg you for some bounties and specially for English translation stop lol full stop.

[CoinsCoinsEverywhere]

What's the nature of your AI program?  Is it something you've designed, something someone has made for you, a commercial product you've seen, ...?  I haven't worked with AI myself, but from what I know of it, it's very hard to design good predictive algorithms, especially for financial markets because there's quite a bit of chaos that simply cannot be predicted.  It's also hard to deal with changing market personality over time.  An algorithm that works great this year may fail miserably next year.  I watched an AI program that someone else made for trading futures markets.  Overall, it was quite profitable for about a year and a half, after which it quickly lost most of the money that people invested with it.

[edmond_dantes]

Thats a funny joke not gonna work with something as crazy as bitcoin

[CoinsCoinsEverywhere]

Thats a funny joke not gonna work with something as crazy as bitcoin

Yes, but I don't think it has much to do with bitcoin being crazy, which I assume is a reference to its volatility.  In some ways, bitcoin price might actually be easier to predict than other financial markets.  From what I've heard, as well as some of my own limited observations, bitcoin tends to follow predictable technical patterns (double top/bottom, head and shoulders, reversals on divergence, etc.) fairly often, which is not as often the case with other financial markets.  The difficulty comes from the enormous complexity of trying to model a financial market.  Your model has to be pretty accurate if you're going to remain profitable over the long term.  Trying to figure out all the major factors that go into determining price is a very difficult task.