AI

For a R(AI)ny Day

A penny saved via algorithm is a penny earned. Now try 500 billion pennies—i.e., $5 billion. That’s the amount AI-assisted savings app Digit has socked away for its users. 

• How it works: Using supervised machine learning, Digit’s model analyzes a user’s spending patterns, then aims to tuck away unnoticeable amounts. 

Hayden chatted with CEO Ethan Bloch for an inside look at Digit’s algorithm. 

Mixed signals 

Signing up for the app allows it to analyze months’ to years’ worth of your transactional data—and to make judgment calls on your behalf. Like many personal finance apps, Digit shares data and analytics with some third parties and can use aggregated personal info to recommend products or display content. 

Four main signals are behind every savings decision: 

• Your checking account balance (and whether that’s high or low for you) 
• When you’ll likely be paid next (and how much) 
• Which bills may be coming due next (and what they’ll cost) 
• How you’ve spent over the past week (and how that compares to your norm) 

Since Digit’s model uses supervised ML, it’s in constant competition with itself to improve. Each time a new version of the model wins out, an engineer decides whether to greenlight it. 

Getting schooled 

It took the Digit team between six and nine months to develop V1 of the algorithm. The process was especially slow-going because—like any K-12 kid—the model only had five days a week to learn. (Money transfers + weekends go together like peanut butter and mayonnaise.) 

• Another hang-up: For every user, Digit’s algorithm makes a savings decision once daily—meaning it initially took the team a full day to gauge the effects of any changes. 

The fix: Engineers built a basic savings simulator that allows for up to 15 simulations per day, providing a boost to learning speed. 

It’s complicated 

As Digit adds more customization—and, soon, investing and retirement features—it’s had to tweak the algorithm in ways it “wasn’t designed for in the beginning,” says Bloch. In effect, he means AI’s greatest nemesis: common sense. 

Scenario: Let’s say it’s safe to move $40 out of your account today, but you’ve got three goals: building up an emergency fund, paying down credit card debt, and saving for a vacation. How should your $40 be divided? 

• Even a human financial advisor might need to consult with you here. For ML, which lacks context and nuance, it’s a much bigger hurdle. 

“We still have a lot of work to do, I think, to make this more intelligent,” says Bloch. “We’re just at the beginning.”

VOICE

Google Assistant, Alexa, and Siri Walk into a Bar

Nobody really talked to them when they entered the bar a few years ago. Now they’re smarter. Their squads (supported device ecosystems) are larger. Will anyone offer to buy them a drink now?

Google Assistant: Google is expanding Assistant’s repertoire at a rapid clip. It recently (and quietly) added smart home scheduling capabilities. Ask “Run the coffee maker in 15 minutes” or “Turn on the disco ball when the clock strikes nine,” and with the right devices and any luck, ye shall receive. 

Alexa: Amazon is improving Alexa while competing with itself in the “What new inanimate objects can we make smart?” game. This week, Amazon added Alexa fitness-tracking integrations to its Echo Buds earbuds, and last week, it opened the sale of $250 Echo Frames (Alexa glasses sans smart display) to the general public. 

Siri: Apple is...not sure. It recently said Siri has an installed base of 1 billion devices and 25 billion total interactions per month, which implies low overall engagement. 

• “Compared to its rivals, Siri seems to still be in middle school,” VC/writer Om Malik noted.

SPONSORED BY NASDAQ

Your Board and Cybersecurity

The threat of cybersecurity incidents is certainly not going away. And as boards and company leadership teams look to mitigate these threats, it may be difficult to find experienced—and diverse—directors. 

To help, Nasdaq has put together a guide, Harness Your Board’s Diversity for Strong Cybersecurity, which highlights diversity’s potential impact on cybersecurity oversight (p.s. It’s based on a Nasdaq study that found organizations with diverse boards may have stronger foundations for more successful cybersecurity strategies). 

And by making age diversity a focus of your efforts, you can bring in people with a focus in areas such as technology and cybersecurity. 

To learn more about how a diverse board can help improve your company’s cybersecurity, read Nasdaq’s guide here.

CONNECTIVITY

A Different Kind of UBI

Not universal basic income, but usage-based insurance. 

Yesterday, pay-per-mile underwriter Metromile said it would go public via SPAC at a valuation of $1.3 billion. The insurtech startup puts connected devices in vehicles and uses the data to generate personalized quotes. Allstate, Progressive, and USAA have implemented similar UBI programs.

• Why UBI is consumer-friendly: You pay as you go, based on mileage and anti-road rage willpower. Traditionally, you’re charged a flat fee determined by demographic data and vehicle make/model. 
• Why it’s not: Privacy tradeoffs. You’re handing over info about your driving behaviors—and potentially your location—in exchange for better rates. 
• Why UBI is insurer-friendly: It improves risk-scoring and underwriting models. Plus, it’s a carrot to incentivize safe driving. 

Looking ahead: UBI will almost inevitably gain market share as connected cars also do across the U.S. Beyond auto—in home, health, and life—insurgent insurtechs and incumbents are using more AI to automate underwriting and claims processing.

BITS & BYTES

Stat: GM is offering Cadillac dealerships $500,000 to abandon ship if they won’t embrace EVs, per Automotive News. 

Quote: “It is very fluent. It is very articulate. It is very good at producing reasonable-sounding text. What it does not do, however, is think in advance.” —Mark Riedl, a professor and researcher at Georgia Tech, on GPT-3.  

Read: MIT Tech Review has “the Zoom-fatigued person’s guide to connecting virtually on Thanksgiving.” The title says it all. 

SPONSORED BY GOODR

Gamers go outside, too. And when they emerge from their digital caves, they’ll be donning goodr sunglasses: the fun, functional, and affordable shades now come in an 8-Bit line of retro gaming-inspired sun shielders. With polarized lenses, no-slip and no-bounce frames, and comin’ in at just $25, goodrs are a no-brainer—even if you spend most of your time splattering alien brains across computer screens. Look cool, play goodr.

WHAT ELSE IS BREWING

• Ready Player Two will hold a virtual treasure hunt in Roblox on Dec. 1.
• Apple has lobbied against a bill aimed at stopping Uighur forced labor in China, WaPo reports.
• India banned 43 more Chinese apps, including AliExpress.
• Microsoft sells gaming consoles at a loss, Xbox chief Phil Spencer confirmed to The Verge.
• California will let robotaxi services charge for rides (with regulatory approval).
• China is imposing identity requirements on live-streamers. 

TRIVIA

We know, we know: Last week’s trivia was SpaceX-themed, and this week’s is all about Tesla. But with the company’s brand-new $500 billion market cap—and its CEO’s new status as the world’s second-richest man—we decided to lean in. 

Take the quiz here.

TECH THINGAMABOBS

For the edge of glory: An OpenAI research scientist posted this masterpiece of a Twitter thread: Lady Gaga as diagrams of AI systems. 

For awkward moments: Drift AI created Uncle Larry, a Thanksgiving dinner chatbot, to give you the perfect response to everything from “Who’d you vote for?” to “Why are you still single?” 

ICYMI

Catch up on the top Emerging Tech Brew stories from the past few editions: 

• Roblox’s vision of the metaverse 
• Amazon lays off dozens from Prime Air
• Reports: Uber in talks to sell self-driving division