tl;dr: Today, large data sets and powerful artificial intelligence capabilities are in the hands of corporations and governments. We all benefit from better, more personalized services. There are downsides, however, and they are significant.
Here’s the problem with closed Artificial Intelligence systems as they are evolving.
In this article from the Wall St. Journal, Cigna is profiled as having “proprietary algorithms” which can predict which of their customers are likely to overdose on opioids.
The good news is that it is working…for their customers.
Meanwhile, other people who happen to have insurance from other carriers, will potentially not be saved, because Cigna needs to maximize its profits.
In the future, this type of life-saving algorithm will be in the public domain. I wonder how many avoidable deaths it will take for people to demand?
This is why projects like SingularityNET, SEED, BigchainDB and Ocean Protocol are so important.
These are blockchain-inspired technologies that address different elements of the “AI stack.”
Each of them, like layers in a Big Mac play a role in delivering AI solutions.
[for a deep dive, here’s an article I wrote in VentureBeat on blockchain-based AI]
SingularityNet offers an open marketplace for AI agents… think “on-demand AI apps.”
As in, “I need an AI to optimize the seating chart for my son’s bar mitzvah based on who likes and doesn’t like each other so that there will be minimal fights at dinner.
BigChainDB offers a decentralized database and DeepBrainChain offers a decentralized computing platform, so you can run that query on computers not owned by one company–which means you know that no one is messing around with the computations. It would be like my aunt Eleanor changing the numbers on the place cards so she didn’t have to sit with my aunt Sylvia. Can’t have that happening.
OceanProtocol offers a substrate for data marketplaces where data providers can get paid for providing data, while actually keeping that data private.
At first glance this may look like a paradox!
But the trick is to bring AI compute to the data itself, which Ocean facilitates.
For example, let’s say my cousin Sol fills out a post-event survey from my son’s bar mitzvah that says he found cousin Kevin to be totally boring and Leonard to be scintillating. Sol keeps the survey results to himself, but on Ocean, the ability for others’ AIs to train against that data in their event planning.
So then, when my sister, Dina is planning her son’s bar mitzvah, the AI will ensure that Sol & Kevin aren’t next to each other, yet no one but Sol will know. How cool is that?”
Oh, bonus, both Sol and I will get a small payment for giving that data to Ocean Protocol when my sister Dina uses it.
And SEED is an open conversational AI, so that when my mom asks Alexa “how many people coming to the birthday party are gluten free?” she knows she is getting an accurate response.
Admittedly, this is a bit of a silly example (though it will save a ton of time for planning future events), but here’s the important thing.
None of these calculations happen behind close doors and they are available to anyone to use. The data and algorithms don’t belong to one person, enriching only her. It is available to everyone, making everyone’s life better.
Once a complex problem algorithm been created (what’s the best way to sit people at a bar mitzvah? how do we predict who is likely to overdose on opioids?), it does not need to get solved again.
Society can move forward more rapidly instead of wasting resources doing re-work.
Yes, you may need to pay to use it, but at least, you have the option. In today’s model, you don’t even have that choice.
Aetna, CareFirst, and all the others who could save the lives of people likely to overdose are going to have spend time and money figuring out what Cigna has already figured out. This is wasteful and counter-productive.
For bar mitzvahs, it’s not a game changer. Jews have overcome worse.
For preventing opioid addiction it’s a life-saver.
Originally Published at: https://www.neverstopmarketing.com/why-people-might-die-if-artificial-intelligence-and-data-doesnt-belong-to-everyone/
Why people might die if Artificial Intelligence and Data doesn’t belong to everyone… was originally published in Data Driven Investor on Medium, where people are continuing the conversation by highlighting and responding to this story.