FLIOT: How AI & IoT will change privacy and augmented UX forever

[spoiler: it's not looking like a dystopia]

Regarding AI & IoT - from the consumer to enterprise level - let's get the elephant in the room out of the way first.

Problems: Data privacy, ownership, bias, trust & delivery, and here's the thing, the cat is out of the bag. From facial recognition we've seen security and UX enhancements at near-frictionless point of sale, experience tailoring or augmentation, to exponentially more accurate and predictive advertising - consumers. People feel mounting discomfort with the uncanny accuracy and magic of data-augmented shopping (not to mention the biases baked into the homogeneously lopsided training data that we've seen). Even forensic data from financial behavior, location, gait, and tone recognition can and are being deployed - so the data treatment seems to be the crux upon which to dedicate efforts. With Distributed Ledger Technology, however, privacy is back within reach - and possibly moreso than ever before.

Enter Federated Learning protocols - and the new wave of encryption methodologies.

*"Federated Learning enables mobile phones to collaboratively learn a shared prediction model while keeping all the training data on device, decoupling the ability to do machine learning from the need to store the data in the cloud. This goes beyond the use of local models that make predictions on mobile devices (like the Mobile Vision API and On-Device Smart Reply) by bringing model training to the device as well."

Federated Learning and the advantages of hardware like neuromorphic chips, have set the table for machine learning to work with data from disparate sources, creating insights about consumer behavior, for example, but without the risk incumbent in actually sharing raw data.

Federated Learning introduces the opportunity to have your cake and to eat it too here. From opt-in systems of engagement from the keyholders of the systems of record - on-device (especially secure leveraging tech like neuromorphic chips), and as many distributed nodes as there are people for bad actors to try to pressure into forfeiting data. Retailers and advertisers utilizing data via Federated Learning also eliminate the cost and liability inherent in keeping aggregate honeypots of personal user data up & protected - as the data stays with users, on-device.

The distributed tokenization of data, further, is capable of supporting a new infrastructure for trustless data as nonfungible assets, which also allow for individualized licensing for an earned microtransactional UBI from health, behavioral and consumptive data. Tokenization to also address automation threats to ceteris paribus livelihoods, more data is better - especially in leapfrogging emerging markets

TL;DR: When it comes to data security, privacy, and liability balanced against its unbelievable & largely dormant utility - not only is trustless tech > pinky promises, but we are also likely on the cusp of a new era for consumer, vendor, and advertiser positive sum markets. A win, win, win.

*Google's Federated Learning research: https://ai.googleblog.com/2017/04/federated-learning-collaborative.html

**Snips.ai , Private by Design, Decentralized Voice Assistant was built on Federated Learning software, acquired by Sonos in 2019