A Machine Learning Perspective on Data About Me

A Machine Learning Perspective on Data About Me

Wednesday 10G

Convener: Adrian Gropper

Notes-taker(s): Scott Mace & Adrian Gropper

Tags for the session - technology discussed/ideas considered:

Machine learning, big data, privacy

Discussion notes, key understandings, outstanding questions, observations, and, if appropriate to this discussion: action items, next steps:

The ML perspective on data about me

Adrian Gropper.

Platforms are trying to add even more value. Look at your phone. Apple or Android. Microsoft couldn’t get into it. That’s an extreme case.

Now something is happening in Washington and with standards development. GAFA+M+Oracle+Salesforce+IBM.

As industries develop useful ontologies, APIs like Amazon no longer have to sell stupid bits, they look at the data and add value, based on inferences based on ML.

With ML inferences, advance science, manipulate purchases, safety.

Who should learn at my expense?

How open should the ML be?

To whose benefit?

Trade secrets are incompatible in science

Friction (good or bad)

“Good guy” data brokers

Cooperatives

Federated learning, a way of dealing with who should learn at my expense

Decision points, enforcement points (from policy)

Photo provided by Adrian Gropper: