biometrics

ARTIFICIAL INTELLIGENCE: Apple Creeps Your Face For Identity

Source: Techcrunch

Source: Techcrunch

Speaking of digital identity, forget the old world of pins and passwords. You are the best measure of your identity -- and you carry it with you just by being who you are. No technology or lookup is needed to map your genetic information to your fingerprint to your voice, these things are just you! And increasingly, financial, social and media graphs are an extension of ourselves in the real world. So shouldn't these things be tied to us like biology? That would be the real-world version of a cyborg, where information streams, from biological to technological, are merged into one.

Apple has been setting the bar in biometrics, whether financial services incumbents are ready or not. The thumb is as good for a credit card payment as a physical credit card; 60% of surveyed consumers are indifferent between them. The next step is facial recognition, which users will rely on to log into their phones in the next version of iOS. To do this, the phone projects 30,000 invisible dots onto your face, scans that information with an infrared camera to build a 3D mesh, and finally uses a locally-run neural network to identify you. No cloud, no upload, no blockchain -- all of this happens on the device.

This approach is part of a broader melding between the real and digital worlds. Scanning objects, knowing what they are, placing them in a rendered augmented reality -- these are things that in 5 years time will be the new normal. Are bankers ready to flush out decade-old password systems for artificial intelligence-driven authentication? Or is it another case of "wait and see"?

FINTECH: Biometrics as Trojan Horse

Source: FacePhi

Source: FacePhi

For this article quoting Autonomous NEXT, we considered the role of biometrics and wearables relative to financial services incumbents and high technology firms. While there has not been massive funding into the biometrics category (vs Insurtech or Regtech for example), the theme is growing consistently and there are many companies pursuing ideas onAngelList. Fintech startups like Eveverify and FacePhi are at the vanguard of these developments. Technologists have not quite gotten the form factors right for mass adoption, given that even the high profile Google Glass and Apple Watch struggled.

But now the banks are giving it a shot. Citigroup tested an eye-scanning ATM (eventually shelving the project), and is using voice, face and fingerprint recognition in its app. Wells Fargo, JP Morgan Chase and Bank of America are rolling out ATMs that connect to a customer's mobile device for authentication. Maybe the phone is just the new personal wallet, and it will hold all the tokens needed to authenticate into a bank account.

But this isn't without strategic danger, for who controls smartphones but Apple and Google! And if we broaden the theme to conversational interfaces, Amazon is there too. High tech companies have a vested interest in knowing their user from a financial perspective as well, primarily to increase personalization and service delivery in their own ecosystem. And as high-tech firm extend into financial features, can big banks rely on them as partners? So the question of who controls the industry, once we are all cyborgs, boils down to what ecosystem the consumer prefers. Hint -- it didn't work out so well for the Telecoms.

ARTIFICIAL INTELLIGENCE: Payments via Face Recognition

Source: MIT Tech Review, Face++

Source: MIT Tech Review, Face++

 Is it fair to say that privacy regulations are looser in China than in the West? That may be giving Chinese startups like Face++ very unique advantages. The company is a private unicorn whose technology maps human faces onto attributes that create an identifiable key. With video cameras everywhere, that key can be used for access across financial services, retail and transportation. Access a building, drive a vehicle, or pick up your tickets? Just show your face.

Of special interest is the use of this facial recognition to initiate payments through Alipay, a payments app with 120 million users. Yes, it's cool that an app can use machine learning to know users well enough for financial transactions. The bigger point is the synthesis of human intent (I want this) and technology (move your social capial from A to B). How soon will payments happen with the speed of a thought? Removing such friction implies increasing the velocity of money and consumption. Is it a surprise China leads the way?

ARTIFICIAL INTELLIGENCE: Uncanny Valley Realistic Chatbots

Source: MIT Tech Review, Soul Machines

Source: MIT Tech Review, Soul Machines

Why should we analyze virtual reality, the attention economy and preferences of the Millennials generations (including their addiction to video games) when thinking about financial services? Here is an example. A startup called Soul Machines has combined the functionality of chatbots, which includes natural language processing and machine vision, with hyper-realistic renderings of human faces. These virtual assistants can model facial expressions and read the facial expressions of their counterparties through machine learning image recognition techniques. We have covered before that speech generation (not using recorded voices but manufacturing human speech sounds) is around the corner as well.

To see the video in action click through the image below or on this link. Does it creep you out to have the bank teller, investment advisor, insurance agent or lawyer replaced by a 3D-rendered head connected to automated knowledge? If yes, welcome to the uncanny valley, a term for near life-like simulations that don't quite hit the mark. But note two things. First, Millennials have spent the last 20 years training how to interact with such characters in video games, such as Mass Effect. And second, virtual reality and augmented reality are scanning physical reality into 3D environments usingphotogrammetry, and we will soon forget the difference.