Google used your pictionary sketches to teach its AI to draw

Great stuff from the guys and dolls at GOOG – showing the power of AI and machine learning in every day situation, making math fun … doesn’t have to be cat videos all the time, right?

From the paper, titled: “A Neural Representation of Sketch Drawings“, we get this:

We present sketch-rnn, a recurrent neural network (RNN) able to construct stroke-based drawings of common objects. The model is trained on thousands of crude human-drawn images representing hundreds of classes. We outline a framework for conditional and unconditional sketch generation, and describe new robust training methods for generating coherent sketch drawings in a vector format.

And they conclude with:

In this work, we develop a methodology to model sketch drawings using recurrent neural networks. sketch-rnn is able to generate possible ways to finish an existing, but unfinished sketch drawing. Our model can also encode existing sketches into a latent vector, and generate similar looking sketches conditioned on the latent space. We demonstrate what it means to interpolate between two different sketches by interpolating between its latent space. After training on a particular class of image, our
model is able to correct features of an input it deems to be incorrect, and output a similar image with corrected properties. We also show that we can manipulate attributes of a sketch by augmenting the latent space, and demonstrate the importance of enforcing a prior distribution on the latent vector for coherent vector image generation.

Good stuff – and Happy Easter, so go on and draw some bunnies.

Source: Google used your pictionary sketches to teach its AI to draw

The AI In Fintech Market Map

OK – so how cool is this? James being included in the just released Top 100 AI fintech firms, as per the boys and gals at CB Insights. Thanks guys, so top 10 is my next target now. 

Funding to AI startups reached record highs in 2016 and applications for artificial intelligence technologies exist across nearly the entire spectrum of business. Highlighted here are the top 100 AI startups selected by CB Insights operating across numerous industry verticals.

Using the CB Insights database, we expanded upon our AI 100 analysis to identify companies that use AI in financial services and mapped them according to the areas where they’re operating. In broad terms, our analysis includes companies whose core offering includes the application of AI to serve the financial services industry, including commercial banking and credit offerings, insurance, asset management, accounting & personal finance, as well as regulatory & compliance services. In addition, many of these companies have additional use cases beyond financial services.

Source: The AI In Fintech Market Map

Data Dive: Alt-Lending, Auto Loans And HHGregg |

Ain’t seen nothing yet …


For those who are fans of big twists, stunning reversals and spectacular admissions of defeat, this week’s data dive should keep you vastly entertained.

Source: Data Dive: Alt-Lending, Auto Loans And HHGregg |

Machine Learning: The New Proving Ground for Competitive Advantage – MIT Technology Review

Leaving you in the dust, my friend. It’s my way, or the highway.

A recent survey conducted by MIT Technology Review Custom and Google Cloud reveals that while the majority of businesses are struggling to apply machine learning, others are hard at work developing strategies for the technology — and are already realizing genuine ROI.
The business world’s focus on machine learning (ML) may seem like an overnight development, but the buzz around this technology has been steadily growing since the early days of big data.

ML is beginning to deliver on the potential created by big data and analytics by turning raw data into useful, predictive tools for business. Innovation-minded business leaders are embracing ML as “the next big thing” and have already crafted ML strategies and initiatives that promise real benefits and return on investment (ROI).

The survey sought to reveal where organizations stand in terms of adopting ML strategies. Respondents included current ML strategists, representatives from companies planning to execute ML initiatives in the next months or years, and those with no ML plans for the foreseeable future.

Source: Machine Learning: The New Proving Ground for Competitive Advantage – MIT Technology Review

Credit Reports to Exclude Certain Negative Information, Boosting FICO Scores.

OK – not sure about this, but it has a distinct “deja-vu” feeling in my book – and if one has been in this business as long as I have, it’s clearly not pointing to anything good here.

The decision by the three major credit-reporting firms— Equifax Inc., Experian PLC and TransUnion—could help boost credit scores for millions of U.S. consumers, but could pose risks for lenders. The reports and scores often help decide how much consumers can borrow for a new house or car as well as determine their credit-card spending limit.

The unusual move by the influential firms comes partially in response to regulatory concerns. The three reporting bureaus rarely tinker with the information that goes on credit reports and that lenders consult to gauge consumers’ ability and willingness to pay back debts.

So let me get this straight. We’re going to make it easier for people to access credit (a good thing) by artificially boosting their credit rating (a bad thing)? Reminds me of the pre-housing bust period, where everybody and his dog got approved for NINJA loans, in the name of “democratization of access to housing”, and we’ve all seen what that has brought us. We’re still digging out of that one.

Maybe useful to read the latest Orchard stats on marketplace lending charge-offs by quarter as per the article here:

Source: Credit Reports to Exclude Certain Negative Information, Boosting FICO Scores – WSJ

Is it OK for lending algorithms to favor Ivy League schools?

This article goes deeper into the subject matter, and addresses a number of issues that are important indeed. In the long run though, it’s clear that this has only one way to go, and it’s a positive story. The human factor being what it is, there is always going to be a bias. Put more machines in charge, let them learn, and we’re off to a better credit world, no doubt. 

The main mantra used to be “software is eating the world”. Mine these days and going forward is “AI and Machine Learning is eating all the rest”. TGIF – but have a good read nevertheless.

“I think a baseline question is, how much disparate impact already exists in the system?” said Paul Gu, co-founder of the online consumer lender Upstart, which includes the potential borrower’s college in its underwriting criteria. “I think we would be kidding ourselves if we thought that the traditional way of underwriting was a completely unbiased way of underwriting. If you look at credit scores by any demographic, they’re extremely uneven. If you look at credit access in America, it’s extremely uneven.”

Source: Is it OK for lending algorithms to favor Ivy League schools?

Innovation in Credit Granting With Big Data – Ash Gupta

You want to hear from the big kahuna – you got it!

All you wanted to know about Big Data, AI and Machine Learning, applied also to the un- and underbanked part of the population, call it broadening of access to credit.

“Ash Gupta is an industry leader in machine learning and big data analytics. He promotes AXP wide innovation to drive revenue growth along with best-in-class Credit and Fraud results. His responsibilities extend across all AXP businesses and geographies. He is an executive officer and reports to the company’s Chairman & CEO. Mr. Gupta’s prior roles include company’s Chief Risk officer and CEO of US banking, along with broad leadership positions in Finance and Strategic Planning. Mr. Gupta earned an MBA from Columbia University and a bachelor’s in Engineering from Indian Institute of Technology (IIT), Delhi. Currently, he serves on the boards of Encore Capital Group (NASDAQ: ECPG), Big Brothers Big Sisters of New York, and South Asian Youth Action (Advisory Board).”


Unstoppable Trends in Online Lending – Noah Breslow

When Noah speaks – the room gets quiet.

An early player in the industry, he’s more than ever at the forefront of what’s happening out there, and his presentation did not disappoint. Equity investors have not been happy to say the least, and it remains to be seen where we go from here.

However, they are still going strong operationally, and one would expect that at some point, things might take a turn for the better. 

“Noah Breslow has served as OnDeck’s Chief Executive Officer and Chairman of our board of directors since June 2012 and as our Chief Operating Officer from October 2011 to June 2012. Previously, Noah held many executive roles at OnDeck including Chief Product Officer (2009-2011), Senior Vice President – Products and Technology (2007-2009) and Vice President, Products and Technology. Prior to joining OnDeck, Noah was Vice President of Marketing and Product Management for Tacit Networks, Inc., a provider of wide area network optimization solutions.

Mr. Breslow holds an S.B. in Computer Science and Engineering from the Massachusetts Institute of Technology and an M.B.A. with distinction from Harvard Business School.”


The Intersection of Technology & Consumer Credit – John Scully

John Sculley is the Vice Chairman to Lantern Credit, where he joined the Board of Managers in August 2015. Mr. Sculley served as chairman, CEO and CTO during a decade-long career at Apple Computer, Inc., following his tenure as the youngest president of the Pepsi-Cola Company. Most recently, he has founded several companies including Obi Worldphone, a Silicon Valley design-led company that markets high-quality smartphones at affordable price points. Additionally, Mr. Sculley is a founder of Zeta Interactive, one of the largest marketing cloud firms in consumer marketing-tech. He currently is an author, recognized expert and popular speaker about high-tech tools for tackling challenges such as corporate revitalization and the high cost of healthcare. Mr. Sculley received a bachelor’s degree from Brown University and an MBA from the Wharton School of the University of Pennsylvania.

Marcus – by Goldman Sachs

OK – so now that I’m back from the LendIt conference, just starting to realize that I attended very few presentations, as I had an otherwise jampacked one on one meeting schedule.

There, happy to point you to the different presentations as they were recorded by the awesome team at LendIt.

The Big Kahuna it is, Goldman Sachs, getting into the space with a splash. Marcus is going to move the needle, and you’ll understand why after watching this presentation.