The most obvious are Machine Learning based crypto hedge funds. Which already exist.

Then there's projects like SONM, Golem and Elastic. Which could have very powerful implications for ML and DL researchers and engineers, since computing power is hard to come by if you're not part of a massive tech company.

There's also Presearch, which seeks to disrupt web searches by paying out the ad revenue to the users of its search engine. And search engines themselves are heavily built upon Machine Learning these days.

Perhaps the blockchain could be used to record self-driving cars as well, to create an immutable database that could be used for legal purposes in the event of a traffic accident.


You could probably take just about any random thing and find a way for it to benefit from either the blockchain or AI or both, even if it might not always be obvious.

Well a dataset which can't be manipulated to fit your opinion will do wonders for the world, we still have people who think earth is flat, imagine verifying every single step through block chain. There will be no more doubting science, things will become transparent in my opinion. Imagine an AI running on a block chain, its code can't be corrupted by a single evil person this will make public more comfortable imo. Every decision the AI takes can be recorded on to the blockchain etc the possibilities are endless with AI and blockchain working together.

I think Golem is one of use case as graphic processing power is important for deep learning AI training these days but not every AI researcher or startup has the resource to afford these heavy demand of graphic cards , this is where Golem comes to fill the gap by taking computing and graphic power from the people and rent to these AI startup at a relatively low price

Advertising, Content Creation, Social Media, I see blockchain and AI merging on several levels, it's interesting to see the amount of new projects coming out every day.

It takes time. AI has long left its "experimental stage" though. The only thing missing is even more computing power.

Google recently released info about their new AI that can program and improve itself... Things are starting to get crazy.

Well the datasets users would have to feed in or the parameters for data collection by an AI would still be exposed to initial bias, though of course, if it were all transparent, others would be able to rank objectiveness and propose add ons or adjustments to improve that.

I see verification of identity, as well as compliance, that already use big data, machine learning and deep learning, as potential candidates for blockchain innovation. In the sense that immutability... proof of work or other cryptographic algorithm can help secure data against intrusion and alteration.

There already is an organisation that is dedicated to discussing, considering, monitoring and building frameworks for ethical design in AI. The Institute of Electrical and Electronics Engineers (IEEE) published at the end of last year the first draft of this a "common standards" for so called "ethically-aligned design" (EAD): http://standards.ieee.org/develop/indconn/ec/ead_v1.pdf

At the time of receiving feedback, until May this year I believe, I think there was one suggestion to branch out the study to consider blockchain use in some of the solutions within, although it was pointed out that some kill-data mechanism (to prevent AI from overstepping) would be against the immutable design of blockchain.

Possibly not sure how it would work but it could be something very interesting if it did happen.

Very interesting, haven't seen these yet. Thanks for the links!