Shared housing startups are taking off

When young adults leave the parental nest, they often follow a predictable pattern. First, move in with roommates. Then graduate to a single or couple’s pad. After that comes the big purchase of a single-family home. A lawnmower might be next.

Looking at the new home construction industry, one would have good reason to presume those norms were holding steady. About two-thirds of new homes being built in the U.S. this year are single-family dwellings, complete with tidy yards and plentiful parking.

In startup-land, however, the presumptions about where housing demand is going looks a bit different. Home sharing is on the rise, along with more temporary lease options, high-touch service and smaller spaces in sought-after urban locations.

Seeking roommates and venture capital

Crunchbase News analysis of residential-focused real estate startups uncovered a raft of companies with a shared and temporary housing focus that have raised funding in the past year or so.

This isn’t a U.S.-specific phenomenon. Funded shared and short-term housing startups are cropping up across the globe, from China to Europe to Southeast Asia. For this article, however, we’ll focus on U.S. startups. In the chart below, we feature several that have raised recent rounds.

Notice any commonalities? Yes, the startups listed are all based in either New York or the San Francisco Bay Area, two metropolises associated with scarce, pricey housing. But while these two metro areas offer the bulk of startups’ living spaces, they’re also operating in other cities, including Los Angeles, Seattle and Pittsburgh.

From white picket fences to high-rise partitions

The early developers of the U.S. suburban planned communities of the 1950s and 60s weren’t just selling houses. They were selling a vision of the American Dream, complete with quarter-acre lawns, dishwashers and spacious garages.

By the same token, today’s shared housing startups are selling another vision. It’s not just about renting a room; it’s also about being part of a community, making friends and exploring a new city.

One of the slogans for HubHaus is “rent one of our rooms and find your tribe.” Founded less than three years ago, the company now manages about 80 houses in Los Angeles and the San Francisco Bay Area, matching up roommates and planning group events.

Starcity pitches itself as an antidote to loneliness. “Social isolation is a growing epidemic—we solve this problem by bringing people together to create meaningful connections,” the company homepage states.

The San Francisco company also positions its model as a partial solution to housing shortages as it promotes high-density living. It claims to increase living capacity by three times the normal apartment building.

Costs and benefits

Shared housing startups are generally operating in the most expensive U.S. housing markets, so it’s difficult to categorize their offerings as cheap. That said, the cost is typically lower than a private apartment.

Mostly, the aim seems to be providing something affordable for working professionals willing to accept a smaller private living space in exchange for a choice location, easy move-in and a ready-made social network.

At Starcity, residents pay $2,000 to $2,300 a month, all expenses included, depending on length of stay. At HomeShare, which converts two-bedroom luxury flats to three-bedrooms with partitions, monthly rents start at about $1,000 and go up for larger spaces.

Shared and temporary housing startups also purport to offer some savings through flexible-term leases, typically with minimum stays of one to three months. Plus, they’re typically furnished, with no need to set up Wi-Fi or pay power bills.

Looking ahead

While it’s too soon to pick winners in the latest crop of shared and temporary housing startups, it’s not far-fetched to envision the broad market as one that could eventually attract much larger investment and valuations. After all, Airbnb has ascended to a $30 billion private market value for its marketplace of vacation and short-term rentals. And housing shortages in major cities indicate there’s plenty of demand for non-Airbnb options.

While we’re focusing here on residential-focused startups, it’s also worth noting that the trend toward temporary, flexible, high-service models has already gained a lot of traction for commercial spaces. Highly funded startups in this niche include Industrious, a provider of flexible-term, high-end office spaces, Knotel, a provider of customized workplaces, and Breather, which provides meeting and work rooms on demand. Collectively, those three companies have raised about $300 million to date.

At first glance, it may seem shared housing startups are scaling up at an off time. The millennial generation (born roughly 1980 to 1994) can no longer be stereotyped as a massive band of young folks new to “adulting.” The average member of the generation is 28, and older millennials are mid-to-late thirties. Many even own lawnmowers.

No worries. Gen Z, the group born after 1995, is another huge generation. So even if millennials age out of shared housing, demographic forecasts indicate there will plenty of twenty-somethings to rent those partitioned-off rooms.

For Apple, this year’s Global Accessibility Awareness Day is all about education

Following Apple’s education event in Chicago in March, I wrote about what the company’s announcements might mean for accessibility. After sitting in the audience covering the event, the big takeaway I had was Apple could “make serious inroads in furthering special education as well.” As I wrote, despite how well-designed the Classroom and Schoolwork apps seemingly are, Apple should do more to tailor their new tools to better serve students and educators in special education settings. After all, accessibility and special education are inextricably tied.

It turns out, Apple has, unsurprisingly, considered this.

“In many ways, education and accessibility beautifully overlap,” Sarah Herrlinger, Apple’s Senior Director of Global Accessibility Policy and Initiatives, said to me. “For us, the concept of differentiated learning and how the accessibility tools that we build in [to the products] help make that [learning] possible is really important to us.”

Apple’s philosophy toward accessibility and education isn’t about purposely targeting esoteric use cases such as IEP prep or specialized teaching methodologies.

In fact, Apple says there are many apps on the iOS App Store which do just that. The company instead believes special education students and teachers themselves should take the tools as they are and discover creative uses for them. Apple encourages those in schools to take the all-new, low-cost iPad and the new software and make them into the tools they need to teach and learn. It’s a sentiment that hearkens back how Steve Jobs pitched the original iPad: It’s a slab of metal and glass that can be whatever you wish it to be.

In other words, it’s Apple’s customers who put the ‘I’ in iPad.

In hindsight, Apple’s viewpoint for how they support special education makes total sense if you understand their ethos. Tim Cook often talks about building products that enrich people’s lives — in an education and accessibility context, this sentiment often becomes a literal truism. For many disabled people, iOS and the iPad is the conduit through which they access the world.

Apple ultimately owns the iPad and the message around it, but in actuality it’s the users who really transform it and give it its identity. This is ultimately what makes the tablet exceptional for learning. The device’s design is so inherently accessible that anyone, regardless of ability, can pick it up and go wild.

(Photo by Tomohiro Ohsumi/Getty Images)

Apple’s education team is special

At the March event, one of the onstage presenters was Kathleen Richardson, who works at Apple on their ConnectedED program. She is one of many who work on the company’s education team, whose group is tasked with working with schools and districts in evangelizing and integrating Apple products into their curricula.

I spoke with Meg Wilson, a former special education teacher who now works on education efforts inside Apple. A former Apple Distinguished Educator, Wilson is the resident “special education guru” who provides insight into how special education programs generally run. With that knowledge, she provides guidance on how Apple products can augment the process of individualizing and differentiating educational plans for special ed students.

A focus of our discussion was the Schoolwork app and how it could be used to suit the needs of teachers and support staff. One example Wilson cited was that of a speech therapy session, where a speech pathologist could use Schoolwork not necessarily for handouts, but for monitoring students’ progress toward IEP goals. Instead of the app showing a worksheet for the student to complete, it could show a data-tracking document for the therapist, who is recording info during lessons. “What we need in special ed is data — we need data,” Wilson said. She added Schoolwork can be used to “actually see the progress” students are making right from an iPad without mountains of paper. A key element to this, according to Wilson, is Schoolwork’s ability to modernize and streamline sharing. It makes conferring with other members of the IEP team a more continuous, dynamic endeavor. Rather than everyone convening once a year for an annual review of students’ progress, Wilson said, Schoolwork allows for “an amazing opportunity for collaboration amongst service providers.”

Wilson also emphasized the overarching theme of personalizing the iPad to suit the needs of teacher and student. “When you are creative with technology, you change people’s lives,” she said.

To her, the iPad and, especially, the new software scale for different learners and different environments really well. For special educators, for instance, Wilson said it’s easy to add one’s entire caseload to Schoolwork and have progress reports at the ready anytime. Likewise, the ability in Classroom to “lock” an entire class (or a single student) into an activity on an iPad, which takes its cues from iOS’s Guided Access feature, helps teachers ensure students stay engaged and on task during class. And for students, the intuitive nature of the iPad makes it so that students can instantly share their work with teachers.

But it isn’t only Apple who is changing education. Wilson made the case repeatedly that third-party developers are also making Apple’s solutions for education more compelling. She stressed there are many apps on the App Store that can help in special education settings (IEP prep, communication boards, etc.), and that Apple hears from developers who want to learn about accessibility and, crucially, how to make their apps accessible to all by supporting the discrete Accessibility features. Wilson shared an anecdote of an eye-opening experience for one developer, who expressed the idea of supporting accessibility “didn’t even occur to him,” but doing so made his app better.

One “big idea” that struck me from meeting with Wilson was how diverse Apple’s workforce truly is. Wilson is a former special education teacher. Apple’s health and fitness team reportedly is made up of such medical professionals as doctors and nurses. Apple’s education team is no different, as my conversation with Wilson attested. It’s notable how Apple brings together so many, from all walks of life, to help inform as they build these products. It really does intersect liberal arts with technology.

Apple makes learning code accessible to all

In early March, Lori Hawkins at the Austin American-Statesman reported on how Apple has made its Everyone Can Code program accessible to all. Hawkins wrote that representatives from Apple visited Austin’s Texas School for the Blind and Visually Impaired to teach students to fly drones with code written in the Swift Playgrounds app. As you’d expect, Swift Playgrounds is fully compatible with VoiceOver and even Switch Control. “When we said everyone should be able to code, we really meant everyone,” Herrlinger told the Statesman. “Hopefully these kids will leave this session and continue coding for a long time. Maybe it can inspire where their careers can go.” Herrlinger also appeared on a panel at the SXSW festival, where she and others discussed coding and accessibility pertaining to Everyone Can Code.

For Global Accessibility Awareness Day this year, Apple has announced that a slew of special education schools are adopting Everyone Can Code into their curricula. In a press release, the company says they “collaborated with engineers, educators, and programmers from various accessibility communities to make Everyone Can Code as accessible as possible.” They also note there are “additional tools and resources” which should aid non-visual learners to better understand coding environments.

In addition to the Texas School for the Blind and Visually Impaired in Austin, Apple says there are seven other institutions across the country that are implementing the Everyone Can Code curriculum. Among them are two Bay Area schools: the Northern California campuses of the California School for the Blind and the California School for the Deaf, both located in Fremont.

At a special kick-off event at CSD, students were visited by Apple employees — which included CEO Tim Cook — who came to the school to officially announce CSB and CSD’s participation in the Everyone Can Code program.

Students arrived at the school’s media lab for what they believed to be simply another day of coding. In reality, they were in for a  surprise as Tim Cook made his appearance. Members of Apple’s Accessibility team walked students through controlling drones and robots in Swift Playgrounds on an iPad. Cook — along with deaf activist and actor Nyle DiMarco — toured the room to visit with students and have them show off their work.

In an address to students, Cook said, “We are so happy to be here to kick off the Everyone Can Code curriculum with you. We believe accessibility is a fundamental human right and coding is part of that.”

In an interview Cook told me, “Accessibility has been a priority at Apple for a long time.” He continued: “We believe in focusing on ability rather than disability. We believe coding is a language — a language that should be accessible to everyone.” When I asked about any accessibility features he personally uses, Cook said due to hearing issues he likes to use closed-captioning whenever possible. And because he wears glasses, he likes to enlarge text on all of his devices, particularly the iPhone.

Accessibility-related Apple retail events

As in prior years, Apple is spending the month of May promoting accessibility and Global Accessibility Awareness Day by hosting numerous accessibility-centric events at its retail stores across the globe. (These are done throughout the year too.) These include workshops on the accessibility features across all Apple’s platforms, as well as talks and more. Apple says they have held “over 10,000 accessibility sessions” since 2017.

Today, on Global Accessibility Awareness Day 2018, Apple is holding accessibility-related events at several campuses worldwide, including its corporate headquarters in Cupertino, as well as at its satellite campuses in Austin, Cork and London.

Technology innovation on the second half of the chessboard

EarthNow recently announced a $1 billion investment, perhaps the largest-ever Series A financing round, to build a global constellation of satellites. Ant Financial announced plans to raise $9 billion at an expected $150 billion valuation, making it the most highly valued privately held company. Last year, SoftBank embarked on a $100 billion investment fund, 30 times larger than any prior venture fund.

The venture industry is scrambling to respond. Several established funds, including Sequoia, Khosla, Norwest and Battery, have recently announced by far their largest funds raised to date. Valuations and round sizes have doubled on average in the past five years.

The speed and magnitude with which technology innovation is moving is mind-boggling, even for those of us who have worked at the center of it for decades. Staid industries for which technology seemed irrelevant are transforming themselves or being disrupted by the Connected World, innovation made possible by the confluence of cloud, mobile, sensor and artificial intelligent technologies. McKinsey has noted that the internet-impacted industries represent 15 percent of our economy. The Internet of Things will impact the rest with a potential economic impact of $11 trillion by 2025.

Technology innovation is now a global village. China has moved from a technology laggard to fast follower to leader within the span of two decades. This year, venture investment in China is likely to surpass U.S. venture investment for the first time. Europe is producing cutting edge technology and companies; the Spotify IPO ago is just the latest example. Venture investors in Silicon Valley used to apply the bridge rule: If an investment involved crossing a bridge, then it was out of scope. Now many of us apply the two-flight rule: Any investment is fair game if it can be reached within two flights.

And yet we are left to ponder: Has the market run amok? Otherwise, what fundamentals are driving the longest bull run in venture history? Brynjolfsson and McAfee from MIT offer some perspective in “The Second Machine Age: Work, Progress, and Prosperity in a Time of Brilliant Technologies.”

First, they note that innovation is accelerating as we approach the “second half of the chessboard.” This analogy applies a parable to Moore’s Law. The game of chess originated during the sixth century in present day India during the Gupta Empire. As the story goes, the emperor was so impressed by the difficult, beautiful game that he invited the inventor to name his reward. The inventor said, “All I desire is some rice to feed my family,” and proposed to start with one grain of rice on the first square and double the grains of rice in each succeeding square.

Impressed with the inventor’s apparent modesty, the emperor replied, “make it so.” If the request were fully honored, the inventor would receive 1.8 x 10^19 grains of rice by the 64th square, more rice than has been produced in the history of the world. The midpoint of the board would receive 4 billion grains of rice, about one large field’s worth of rice. It was only as they headed into the second half of the chessboard that at least one of them got into trouble.

The range of possible innovations for aspiring entrepreneurs are broader than they have ever been.

Geoffrey Moore first proposed what has become Moore’s Law — the doubling of compute power every two years — in 1965. Moore initially indicated that he could foresee this pattern persisting at least 10 years.  Moore’s “Law” is merely a guideline, yet it has proven to be reliable over the past 50 years, and experts indicate it is likely to persist for another 10-15 years. If applied from the invention of semiconductors in 1958, then we are currently on the 30th square — rapidly approach the second half of the chessboard.

Until recently, the implications of Moore’s Law have been predictable. I first extrapolated Moore’s Law out 10-15 years starting in the 1990s. One could readily envision the miniaturization of computers, the rise of smart phones and Dick Tracy watches, the proliferation of sensors, higher processing speeds, storage capacity, compute power that would permit robotics, augmented intelligence and edge network computing. As we project forward, implanted devices, self-healing operations and autonomous vehicles seem imminent.

But as compute power far exceeds human capacity, it is increasingly difficult to apprehend the future implications of Moore’s Law. Much as with the emperor and inventor, the acceleration of innovations and magnitude of change puts us in promising but murkier territory as we enter the second half of the chessboard.

The second concept that Brynjolfsson and McAfee highlight is the delayed impact of fundamental innovation adoption. Pervasive utilization of the steam engine, internal combustion engine, electricity and indoor plumbing took decades, often 30-60 years. These innovations were often not adopted until new manufacturing facilities were built decades later.

We observe a similar trend in adoption of computer and internet technology. The publishing industry for books and newspapers was the most obvious application of the internet, yet it took well over a decade for our reading habits and the industry to adjust. Many would say this is still a work in progress. The financial industry is fundamentally a digital business, yet many practices remain entrenched: cash and credit card-based payments are but one example. The auto sector is just beginning to grapple with myriad new technologies. Surely the manufacturing and industrial sector will take longer still.

So two innovation trends are coinciding. Increases in compute power empower artificial intelligence, smart sensors and edge computing for the first time. Meanwhile, many industries are grappling to adopt technology available in the market for decades. The range of possible innovations for aspiring entrepreneurs are broader than they have ever been. The potential to transform industries has never been greater. More capital than ever before is available for good ideas. It is a great time to be an entrepreneur.

Food delivery’s untapped opportunity

Investors may have already placed their orders in the consumer food delivery space, but there’s still a missing recipe for solving the more than $250 billion business-to-business foodservice distribution problem that’s begging for venture firms to put more cooks in the kitchen. 

Stock prices for Sysco and US Foods, the two largest food distributors, are up by more than 20 percent since last summer, when Amazon bought Whole Foods. But, these companies haven’t made any material changes to their business model to counteract the threat of Amazon. I know a thing or two about the food services industry and the need for a B2B marketplace in an industry ripe with all of our favorite buzz words: fragmentation, last-mile logistics and a lack of pricing transparency.

The business-to-business food problem

Consumers have it good. Services such as Amazon and Instacart are pushing for our business and attention and thus making it great for the end users. By comparison, food and ingredient delivery for businesses is vastly underserved. The business of foodservice distribution hasn’t gotten nearly as much attention — or capital — as consumer delivery, and the industry is further behind when it comes to serving customers. Food-preparation facilities often face a number of difficulties getting the ingredients to cook the food we all enjoy.

Who are these food-preparation facilities? They range from your local restaurants, hotels, school and business cafeterias, catering companies, and many other facilities that supply to grocery markets, food trucks and so on. This market is gigantic. Ignoring all other facilities, just U.S. restaurants alone earn about $800 billion in annual sales. That’s based on research by the National Restaurant Association (the “other NRA”). Specific to foodservice distribution in the U.S., the estimated 2016 annual sales were a sizable $280 billion.

How it works today

Every one of these food-preparation facilities relies on a number of relationships with distributors (and sometimes, but rarely, directly from farms) to get their necessary ingredients. Some major national players, including Sysco and US Foods, mainly supply “dry goods.” For fresh meats, seafood and produce, plus other artisanal goods, these facilities rely on a large number of local wholesale distributors. A few examples of wholesalers and distributors near where I live in the San Francisco Bay Area are ABS Seafood, Golden Gate Meat Company, Green Leaf, Hodo Soy and VegiWorks.

Keep in mind that the vast majority of these food-prep businesses don’t shop for ingredients the way you and I may shop for ingredients from our local supermarkets or farmer markets. There’s too little margin in food and doing so would be too costly, as well as highly inefficient (e.g. having to pay to send staff out “grocery shopping”). A few small operators do buy ingredients from wholesale chains such as Costco or Restaurant Depot. But in general, it’s way more efficient to place an order with a distributor and get the goods delivered directly to your food-prep facility.

But that’s where the problems lie. These distributors are completely fragmented, and the quality of fresh ingredients varies meaningfully from one distributor to the next. Prices fluctuate constantly, typically on a weekly basis. What’s worse is delivery timeliness, or rather the lack thereof. These distributors each employs their own delivery staff and refrigerated trucks. There is a limited number of 6 am deliveries they can make for a given delivery fleet.

As a food business operator, you may be ordering quality ingredients at the right price, but if the delivery doesn’t show up on time, you’re outta luck. You won’t be able to prepare the food in time, all the while paying for staff who are sitting around waiting for ingredients to arrive.

As a result, you keep getting seemingly random offline pitches with promotions and price breaks from these distributors. But there’s no way to ensure timely delivery. Everybody makes verbal promises and it’s all based on who you know. Things may work for a week or two until you get “deprioritized” by one of the distributors and you have to start the process of finding the next one.

You intentionally rotate among the different distributors, just to keep them “on their toes.”

The opportunity for a food distribution platform

What’s missing is a platform that hosts a catalog of products from these distributors, with updatable availability, pricing and inventory. On it, food businesses could browse for products and place orders. Fulfillment can be done by the distributors at the beginning, but ultimately that operation may need to be done by the platform to maintain consistent quality of service. Reliable fulfillment may end up being the biggest differentiator for such a platform.

I’m aware of startups that have tried to become the dominant B2B platform for food service distribution. But it takes meaningful resources to get to critical mass, and these startups tend to flame out before reaching that point. It’s not necessarily their fault for not being effective.

This industry has low margins, is slow to adopt new technologies and has many incumbent players. But the opportunity to design and execute on this platform is significant, with clear ROI as a reward and a built-in moat once it reaches critical mass.

Food-prep businesses are hungry for a better solution. And as any food entrepreneur knows, hungry customers are the best kind.

Society needs the Artificial Intelligence Data Protection Act now

On December 31, 2015, I published my original call to arms for society’s rational regulation of artificial intelligence before it is too late. I explained certain reasons why someone who is against solving problems through regulation would propose precisely that mechanism to help hedge the threats created by AI, and announced my proposed legislation: The Artificial Intelligence Data Protection Act (AIDPA).

Since 2015, we have witnessed AI’s rapidly evolving national and international growth and adoption that will soon impact every phase of mankind’s life, from birth to death, sex to religion, politics to war, education to emotion, jobs to unemployment.

Three of many recent developments confirm why now is the time for the AIDPA: (1) a McKinsey study from late 2017 determined that up to 800 million workers worldwide may lose their jobs to AI by 2030, half of contemporary work functions could be automated by 2055 and other recent studies suggest as many as 47 percent of U.S. jobs could be threatened by automation or AI over the next few decades; (2) AI has now created IP with little or no human involvement and continues to be programmed, tested and used to do so; see my Twitter for a library of media reports on AI-created IP; (3) tech giants and regulators are starting to acknowledge that industries that create and use AI should be at least partially responsible for minimizing the impact of AI-displaced workers.

Now – and not later — society must address AI’s legal, economic and social implications with regard to IP and employment. Current legislation does not adequately account for the new challenges, threats and needs presented by the impact of AI. The question is not “if” but “when” society will regulate AI. Rather than leave the job solely to politicians, industry should lead the way through the AIDPA. The urgency to finalize and enact the AIDPA cannot be understated.

This article addresses the AIDPA’s twin focuses (AI’s threats to intellectual property rights and the labor force) and presents a proposed framework to address them. The AIDPA is intended to provide industry with a voice in regulating AI while promoting its safe, secure and ethical use. The United States must lead the way in regulating AI, and leaders in industry, technology and ethics should join together to finalize and enact the AIDPA — the first and most important legislation of its kind.

Intellectual property considerations

The AIDPA’s focuses on ownership of IP and the security risks resulting from machine learning that exceeds its initial programming and/or that by virtue of its programming becomes capable of autonomous human-like reasoning. For a host of legal and technical reasons, current IP laws cannot adequately account for IP created by AI working independent of human involvement or oversight (music, art, medical techniques, processes to communicate, processes to kill, etc.) or that exceeds its initial programming. AI also will acquire vast amounts of confidential information through its ability to collect, process, analyze and utilize mass amounts of data.

Chief AI officer

The AIDPA will require covered entities (see below) to employ a “chief AI officer,” who, among other things, is responsible for monitoring AI within the workplace, creating company-wide plans for AI-impacted employment, implementing the AIDPA regulations, enacting company-wide safeguards that monitor for and respond to malicious AI activity and accounting for AI-created IP.

Governing body

The AIDPA will also establish a governing body (the “AI Board”), staffed with industry, technical, ethical and legal experts, designed to bring specialized expertise and consistency to regulating AI in industry, encourage industry participation, promulgate safety and ethical regulations and adjudicate AI-related IP disputes. The AIB will also ensure that covered entities, through their CIAO, determine if and when certain AI should be outlawed, constrained in specific ways, and/or “terminated” and, where necessary, will enforce the AIDPA’s mandates by making these ultimate determinations.

Industry also will have annual AI-related worker displacement reporting requirements and the AIB will be responsible for analyzing and reporting on AI’s displacement impact on the labor market. Finally, the AIB will administrate and adjudicate disputes related to the AI Worker Realignment Program, which will be funded under the AIDPA.

Ownership, infringement and misappropriation

With regard to AI-created IP, there are many questions of ownership and liability for infringement and misappropriation. Under current IP laws, ownership (and standing to sue) are generally restricted to humans. The AIDPA will allow, under certain circumstances, for IP to be owned by the AI which created it (and in certain circumstances the entity or individual who “owns” the AI machine) in the context of addressing and defining IP rights for non-human created works, set the parameters for human ownership of AI-created IP and, as noted above, determine what AI is off-limits and when AI ownership and even the AI itself must be restrained or terminated.

With regard to infringement and misappropriation, existing law provides that a person or entity is generally liable for infringement regardless of their knowledge of the infringement. The AIDPA will limit the liability of corporations and humans for infringement to cases where there is knowledge of and/or active participation in the infringement.

Employment considerations

The AIDPA currently defines covered entities as government contractors and organizations with 300 or more employees or annual revenue in excess of $30 million that utilize AI or develop or deploy AI-created IP in a manner that results in (i) layoffs of at least 75 workers during a 30-day period on account of implementation and/or use of AI; or (ii) an AI facility opening defined as a covered employer establishing a new facility (brick and mortar), an operation (i.e. a new logistics hub with autonomous trucks and no human drivers) and/or a line of business (i.e. a call center staffed solely with AI-machines) that utilizes AI machines to perform job functions in lieu of what historically was performed by 40 or more humans or (iii) an AI Readjustment, defined as 30 or more workers who experience a reduction of 50 percent or more in their working hours or the loss of more than 75 percent of their job functions, either of which negatively alters the amount of their compensable time.

In the event of a triggering event, the AIDPA provides for certain notice requirements. For example, in the case of layoffs, the AIDPA requires covered entities to provide at least 60 days notice to the impacted workers, which period shall be extended to 180 days for employees who enter and continue approved educational and/or employment retraining through the AIDPA’s Worker Realignment Program. Impacted workers also will be eligible for certain supplemental payments funded through the AIDPA for specified periods. The AIDPA also requires covered entities to submit annual reports on the use of AI and its statistical impact on the labor market.

The dirty “T” word

Like it or not, the undeniable scope and societal impact of AI-caused worker displacement, coupled with the massive reduction in payroll expense for covered entities and the resulting loss in government revenue, mandates that covered entities play a substantial role in funding society’s efforts to respond to and retrain displaced workers.

If it is to be assumed that mass worker displacement left unchecked has the potential to cause serious societal disruption and that AI taxation by politicians is inevitable, then this is not a provocative proposition. It is simply society being intellectually honest with itself. In 2017, Bill Gates proposed a tax on companies using AI which could be used to finance programs for the elderly and others with unmet needs. That same year, San Francisco Supervisor Jane Kim created a task force to explore an AI tax to fund education. And in Europe, Mady Delvaux, a member of the European Parliament, proposed a similar framework as part of an unsuccessful effort to enact AI legislation.

The question for industry is simple: Should the AI taxation framework be left solely to politicians, or should industry that will create and deploy AI play an important role in its formulation. The AIDPA answers that question by including a taxation component designed to secure the necessary funds for society to adjust to AI’s impact.

While still being studied and finalized, the AIDPA favors a tripartite approach for covered entities that is calculated based on (i) a minimum AI “flat” tax; plus a percentage of (ii) human labor cost savings; and (ii) profits generated by AI. The AIDPA provides that the revenue generated from the AI tax shall be used solely for two purposes: (i) retraining workers displaced by AI through the Work Realignment Program and (ii) basic supplemental income payments for AI-displaced workers for a set period.

Questions remain regarding how AI in the workplace should be regulated, but now is the time for lawyers, industry, academia, regulators and politicians to come together to finalize and enact the AIDPA.

Broadening education investments to full-stack solutions

As an education investor, one of my favorite sayings is that education is the next industry to be disrupted by technology, and has been for the past twenty years.

When I started my career at Warburg Pincus, I inherited a portfolio of technology companies that senior partners naively believed would solve major problems in our education system.

It would have worked out fine, of course, except for all the people. Teachers weren’t always interested in changing the way they taught. IT staff weren’t always capable of implementing new technologies. And schools weren’t always 100% rational in their purchasing decisions. And so while, given the size of the market, projections inexorably led to $100M companies, sales cycles stretched asymptotically and deals never seemed to close, particularly in K-12 education.

My current firm, University Ventures, began life in 2011 with the goal of funding the next wave of innovation in higher education. Much of our early work did revolve around technology, such as backing companies that helped universities develop and deploy online degree programs. But it turned out that in making traditional degree programs more accessible, we weren’t addressing the fundamental problem.

At the time, America was in the process of recovering from the Great Recession, and it was clear that students were facing twin crises of college affordability and post-college employability. The fundamental problem we needed to solve was to help individuals traverse from point A to point B, where point B is a good first job – or a better job – in a growing sector of the economy.

Once we embarked on this journey, we figured out that the education-to-employment missing link was in the “last mile” and conceptualized “last-mile training” as the logical bridge over the skills gap. Last-mile training has two distinct elements.

The first is training on the digital skills that traditional postsecondary institutions aren’t addressing, and that are increasingly listed in job descriptions across all sectors of the economy (and particularly for entry-level jobs). This digital training can be as extensive as coding, or as minimal as becoming proficient on a SaaS platform utilized for a horizontal function (e.g., Salesforce CRM) or for a particular role in an industry vertical. The second is reducing friction on both sides of the human capital equation: friction that might impede candidates from getting the requisite last-mile training (education friction), and friction on the employer side that reduces the likelihood of hire (hiring friction). Successful last-mile models absorb education and hiring friction away from candidates and employers, eliminating tuition and guaranteeing employment outcomes for candidates, while typically providing employers with the opportunity to evaluate candidates’ work before making hiring decisions. Today we have eight portfolio companies that take on risk themselves in order to reduce friction for candidates and employers.

The first clearly viable last-mile training model is the combination with staffing. Staffing companies are a promising investment target for our broadened focus because they have their finger on the pulse of the talent needs of their clients. Moreover, staffing in the U.S. is a $150B industry consisting of profitable companies looking to move up the value chain with higher margin, differentiated products.

Because fill rates on job reqs can be as low as 20% in some skill gap areas of technology and health care, there is no question that differentiation is required; many companies view staffing vendors as commodities because they continue to fish in the same small pool of talent, often serving up the exact same talent as competitors in response to reqs.

Adding last-mile training to staffing not only frees the supply of talent by providing purpose-trained, job-ready, inexpensive talent at scale, but also increases margins and accelerates growth. It is this potential that has prompted staffing market leader Adecco (market cap ~$12B) to acquire coding bootcamp leader General Assembly for $412.5M. The acquisition launches Adecco down a promising new growth vector combining last-mile training and staffing.

We believe that staffing is only the most obvious last-mile training model. Witness the rise of pathways to employment like Education at Work. Owned by the not-for-profit Strada Education Network, Education at Work operates call centers on the campuses of universities like University of Utah and Arizona State for the express purpose of providing last-mile training to students in sales and customer support roles. Clients can then hire proven talent once students graduate. Education at Work has hired over 2,000 students into its call centers since its inception in 2012.

Education at Work is the earliest example of what we call outsourced apprenticeships. For years policy makers have taken expensive junkets to Germany and Switzerland to view their vaunted apprenticeship models – ones we’ll never be able to replicate here for about a hundred different reasons. This week, Ivanka Trump’s Task Force on Apprenticeship Expansion submitted a report to the President with a “roadmap… for a new and more flexible apprenticeship model,” but no clear or compelling vision for scaling apprenticeships in America.

Outsourced apprenticeships are a uniquely American model for apprenticeships, where service providers like call centers, marketing firms, software development shops and others decide to differentiate not only based on services, but also based on provision of purpose-trained entry-level talent. Unlike traditional apprenticeship models, employers don’t need to worry about bringing apprentices on-site and managing them; in these models, apprentices sit at the service provider doing client work, proving their ability to do the job, reducing hiring friction with every passing day until they’re hired by clients.

America leads the world in many areas and outsourcing is one of them. Outsourced apprenticeships are an opportunity for America to leapfrog into leadership in alternative pathways to good jobs. All it will take is service providers to recognize that clients will welcome and pay for the additional value of talent provision. We foresee such models emerging across a range of industries and intend to invest in companies ideally positioned to launch them.

All of these next generation last-mile training businesses will deliver education and training – predominantly technical/digital training as well as soft-skills where employers also see a major gap. They’ll also be highly driven by technology; technology will be utilized to source, assess and screen talent – increasingly via methods that resemble science fiction more than traditional HR practices – as well as to match talent to employers and positions. But they’re not EdTech businesses as much as they are full-stack solutions for both candidates and employers: candidates receive guaranteed pathways to employment that are not only free – they’re paid to do it; and employers are able to ascertain talent and fit before hiring.

While last-mile solutions can help alleviate the student loan debt and underemployment plaguing Millennials (and which put Gen Z in similar peril), they also have the potential to serve two other important social purposes. The first is diversity.

Just as last-mile providers have their finger on the pulse of the skill needs of their clients, they can do the same for other needs, like diversity. Last-mile providers are sourcing and launching cohorts that directly address skill needs, as well as diversity needs.

The second is retraining and reskilling of older, displaced workers. For generations, college classrooms were the sole option provided to such workers. But we’re unlikely to engage those workers in greatest need of reskilling if college classrooms – environments where they were previously unsuccessful – are the sole, or even initial modality. As last-mile training models are in simulated or actual workplaces, they are much more accessible to displaced workers.

Finally, the emergence of last-mile full-stack solutions like outsourced apprenticeships raises the question of whether enterprises might not only seek to outsource entry-level hiring, but all hiring. Why even hire an experienced worker from outside the company if there’s an intermediary willing to source, assess and screen, upskill, match, and provide workers on a no-risk trial basis? As sourcing, screening, skill-building, and matching technologies become more advanced, why not offload the risk of a bad hire to an outsourced talent partner? Most employers would willingly pay a premium to reduce the risk of bad hires, or even mediocre hires. If the market does evolve in this direction, education investors with a full-stack focus have the potential to create value in every sector of the economy, making traditional investment categories of “edtech” seem not only naïve, but also quaint.

 

Looking for a better exit? Get out of the game early

VC investing is a game of putting money into a company and hopefully getting more out. In an ideal world, the value placed on a company at acquisition or initial public offering would be some large multiple of the amount of money its investors committed.

As it happens, that multiple on invested capital (MOIC) makes for a fairly decent heuristic for measuring company and investor performance. Most critically, it provides a handy metric to use for answering these questions: For US-based companies, have exit multiples changed in a meaningful way over time? And, if so, does this suggest something about the investment landscape overall?

Coming up with an answer to this question required a specific subset of funding and exit data from Crunchbase. If you’re interested in the how and why behind the data, check out the Data and Methodology section at the very end of the article. If not, we’ll cut right to the chase.

Exit multiples may be on the rise

A rather conservative analysis of Crunchbase data suggests that, over the past decade or so, exit multiples were on the rise before leveling off somewhat.

Below, you can see a chart depicting median MOIC for a set of U.S.-based companies with complete (as best we can tell) equity funding histories stretching back to Series A or earlier, which also have a known valuation at time of exit. That valuation is either the price paid by an acquirer or the value of the company at the time it went public. In an effort to reduce the impact of outliers, we only used companies with two or more recorded funding rounds. With that throat-clearing out of the way, here’s median MOIC over time:

It should be noted for the record that the shape of the above chart somewhat changes depending on the data is filtered. Including exit multiples for companies with only one reported funding round resulted in slightly higher median figures for each year and a more steady linear climb upward. But that’s probably due to the number of comically-high multiples some companies with single small rounds and large exit values produced. There are surely examples of companies that raised $1 million once and later sold for $100 million, but those are fewer and further between than companies with missing data from later rounds.

What might be driving the rise in exit multiples?

The rise in exit multiples may have to do with the fact that more companies are getting acquired at earlier stages.

Crunchbase data suggests that startups earlier in the funding cycle tend to deliver better exit multiples. In an effort to denoise the data a bit, we took the Crunchbase exit dataset and filtered out the companies that raised only one round. (Companies that raised only one round produced a lot of crazy outlier data points that skewed final results.)

This suggests that, in general, the earlier a startup is acquired in the funding cycle the more likely it is to deliver larger multiples on invested capital.

Now, granted, we’re working off of small sample sets with a fair amount of variability here, particularly for startups on opposite ends of the funding lifecycle. There is going to be some sampling bias here. Founders and investors are less likely to self-report disappointing numbers; therefore, these findings aren’t ironclad from the perspective of statistical significance.

But it’s a finding that nonetheless echoes a prior Crunchbase News analysis, which found a slight but statistically significant inverse relationship between the amount of capital a startup raises and the multiple its exit delivers to investors and other stakeholders. In other words, startups that raise less money (such as those at seed and early-stage) tend to deliver better multiples on invested capital.

The changing population of companies finding exits

So does the tendency for earlier-stage companies to deliver better investment multiples have anything to do with upward movement in MOIC ratios? It could, particularly if more seed and early-stage startups are headed to the exit these days. And as it turns out, our data suggest that’s happening.

The chart below shows the breakdown of exits for venture-backed companies based on the last stage of funding the company raised prior to being acquired or going public. We show a decade’s worth of funding data, this time including all exits from U.S. companies with known venture funding histories since seed or early stage—some 5,275 liquidity events in all. For 2018, we also include stats for exits through the beginning of May. Given reporting delays and the fact that there are still eight months left in the year, this is certainly subject to change.

Now, to be clear, over the past decade, there has been some notable growth in the overall number of exits for U.S.-based venture-backed companies across all stages.

But, in some ways, the raw number of deals doesn’t much matter. After all, figures from several years ago aren’t really actionable to founders and investors looking for an exit sometime this year. What matters, then, is what the mix of exits looks like, and at least for the set of companies we analyzed here, the past ten years brought an ultimately small but nonetheless notable shift in the mix of companies that get acquired or go public.

Seed and early-stage companies now make up a larger proportion of the population of exited companies now than in the past. And since companies at that stage tend to deliver higher multiples, it is likely responsible for part of the increase over time.

There are certainly other factors besides the influx of seed and early-stage ventures into the mix of exits, but sussing those out will require further investigation.

It should go without saying that any venture-backed company that gets acquired or goes public is a success, at least of some sort. After all, a tiny fraction of new businesses secure outside funding from angel investors or venture capitalists, and only a small proportion of those get acquired.

Any exit is better than none.

Data and methodology

Let’s start by saying that there is probably no canonically correct way to do this sort of analysis and that since Crunchbase News is working off of private company data, what hasn’t been aggregated programmatically is subject to self-reporting bias. Founders and investors are more likely to disclose exit valuations that make them look good, so this may skew our findings higher.

Definition of funding stages

Here, we use the same funding stage definitions as Crunchbase News does in its quarterly reporting.

  • Seed/Angel-stage deals include financings that are classified as a seed or angel, including accelerator fundings and equity crowdfunding below $5 million.
  • Early stage venture include financings that are classified as a Series A or B, venture rounds without a designated series that are below $15M, and equity crowdfunding above $5 million.
  • Late stage venture include financings that are classified as a Series C+ and venture rounds greater than $15M.
  • Technology Growth include private equity investments in companies that have previously raised venture capital rounds.

Building the base dataset

Here are the basic process we used:

  1. We started by aggregating pre-IPO venture funding raised by U.S.-based companies. We focused only on equity funding only (angel, seed, convertible notes, equity crowdfunding, Series A, Series B, etc.), and did not include debt financing, grants, product crowdfunding, or other non-equity funding events. We did include private equity rounds, if and only if PE was the terminal round and the company had raised a seed, angel, or VC round prior to raising PE.
  2. For each company, we recorded the stage of its first and last known funding rounds.
  3. We excluded any company whose first round was Series B or later.
  4. We excluded companies that were missing dollar amounts for any of their equity funding rounds.
  5. We then retrieved the valuations at acquisition or IPO for each of the companies, again excluding any companies for which terminal private market valuation was not known.
  6. Finally, for each company, we divided valuation at exit by the amount of known venture funding, resulting in the multiple on invested capital from equity financing events.

In conjunction with choosing to start with Series A and earlier funding events, we believe this produced a set of companies with reasonably complete funding histories. Granted, there are “unknown unknowns,” like later rounds that weren’t captured in Crunchbase, but there is no good way to control for those.

The UK and USA need to extend their “special relationship” to technology development

The UK and the USA have always had an enduring bond, with diplomatic, cultural and economic ties that have remained firm for centuries.

We live in an era of profound change, and are living with technologies set to change things ever faster. If Britain and America work together to develop these technologies for the good of mankind, in a way that is open and free, yet also safe and good for our citizens, we can maintain the global lead our nations have enjoyed in the fields of innovation.

Over past months we have seen some very significant strides forward in this business relationship. All of the biggest US companies have made decisions to invest in the UK. Apple is developing a new HQ in the iconic Battersea Power Station, close to the new US embassy, while Google is building a billion dollar new HQ in the increasingly fashionable King’s Cross. Facebook, Amazon, IBM and Microsoft are all extending their operations, and a multitude of smaller US firms are basing their international headquarters in London.

They are all coming here because as we prepare to leave the EU we are building a forward looking Britain that is open to the wider world, and tech is at the heart of this.

Similarly, there have been major expansions or new investment from British firms into the US. Jaguar Land Rover, the UK’s largest automotive manufacturer, supports more than 9,000 jobs in the USA and have recently opened their new multimillion-dollar corporate North America HQ in New Jersey.  iProov, a leading British provider of biometric facial verification technology, became the first international company to be awarded a contract from the US Department of Homeland Security Science & Technology Directorate’s Silicon Valley Innovation Program last month.

We want to work with our global partners – to share expertise, and encourage investment – as we harness technology for the wider good. And that of course includes our old friend and closest ally, the USA.

We have a great deal to offer.

The UK was recently ranked the most AI ready nation among all the OECD countries. In the past three years, new AI start-ups have been created in the UK on an almost weekly basis.

Recently, UK government and industry together committed over $1 billion to support our AI sector, much of which will go towards entrepreneurs. Funding has been set aside to create a nationwide network of tech incubators, that we’re calling “Tech Nation”, which will support new AI businesses as they get off the ground.

We are also excited by — and I am a firm advocate for — the development of blockchain and similar technologies. The UK is leading the way in many areas where blockchain has the potential to be used, such as Fintech. There are now more people working in UK Fintech than in New York or in Singapore, Hong Kong and Australia combined.

And we are eminent in the development of immersive technologies, like Augmented and Virtual Reality, which look set to radically improve many areas of life in coming years, with applications as varied as flight simulation and surgical training techniques.

There is so much to be gained from close collaboration between our two countries on these new technologies and from sharing our expertise.

Together, we can reap the economic benefits of stealing an early lead in their development. We estimate that AI, for example, if widely adopted, could add $33 billion to the UK economy. But, perhaps most importantly, we can also work together to build a strong regulatory and ethical frameworks for their wider application.

It is the role of governments across the world, the UK and US included, to set frameworks for these decentralised, cross border systems so we can manage their use in a safe and effective way.

Our aim should be to harness the power and capability of technology but always for the benefit of, and in service to the populace.

We in the UK are avowedly pro-tech, always seeking to put its power in the hands of our citizens.

We have all learned valuable lessons from the recent scandals regarding data use, most recently around Facebook’s use of data.

We want to build a system that protects and cherishes the freedom of the Internet while protecting the rights of individuals, and their property, including intellectual property.

We want to see freedom in a framework; where our tech entrepreneurs have the space to innovate, knowing they do so with full public trust. Trust underpins a strong economy, and trust in data underpins a strong digital economy.

So in the UK we are developing a Digital Charter, to agree norms and rules for the online world and put them into practice. Our starting point is that what is unacceptable offline should not be tolerated in the online world. That includes how tech companies treat private citizens and use their data, as well as how people treat each other online.

Important changes like these cannot be agreed by one country alone. It is more important than ever that we work together and find common ground so we can make sure that tech continues to change the world for the better. Based on our mutual love of freedom and individual rights Britain and America have through history risen to challenges together. I firmly believe working together we can build that brighter future.