John Melas-Kyriazi

A couple of months ago, I decided to take a hiatus from Tumblr as I continue focusing my time on long-form writing.

I may decide to revive my Tumblr page in the future. In the meantime, please follow me on Twitter and Medium, and follow Spark on Medium.

There has recently been an explosion of cynicism in the Twitterverse (and generally in the tech community) about machine learning and AI. This was taken up a notch after Facebook’s F8 conference earlier this week and the splashy launch of their bot platform on Messenger.

I understand the cynicism. When you dig in, many of the companies claiming to be leveraging machine learning are doing nothing of the sort. It’s easy to find companies that are building weak ML-driven user experiences, turning off early adopters. Others are building products centered around gimmicky ML features that won’t create any sort of long term barrier to competitors, frustrating would-be investors.

I think we’re probably on our way up to the peak of the AI hype cycle right now. A lot of money is going into projects that will not end up being valuable to society. It's normal at this stage, and it’s just fine.

Let’s get past the hype. Bots are an interesting use case for machine learning, but it goes much deeper than that. There is fundamental innovation happening around us. 

I spent most of the day today at MLconf in New York. It was a packed house at 230 5th Ave — a melting pot of engineers, scientists, and entrepreneurs from every corner of the globe. As a physical scientist / engineer, I usually feel like a techie. Not today.

Consumer-facing bots came up a few times during the conference, but they weren't the focus. People were talking about new, advanced statistical techniques that I had never seen before in my research on the space. Software to allow data scientists to easily scale machine learning algorithms to parallel computing using only Python. Neural networks for advanced speech recognition, trained off of unimaginably large data sets. Models to predict medical issues from streams of live patient data from hospitals. I got a demo that blew my mind... but I promised not to talk about it publicly, so I won’t.

Here’s the simple truth: many of the smartest people in the world are throwing every ounce of their brain power into creating the building blocks for a machine learning software ecosystem. There have been some success stories so far, as well as some failures. While it’s still early, I don’t think it’s ever a winning proposition to bet against humanity’s ability to come together to solve big, technical problems.

I’m sitting in our San Francisco office. With a few taps on my smartphone, I’m on Facetime with colleagues in Lagos, Nigeria, getting a live update on a new hire for one of our portfolio companies. This is as close as we can get to teleportation in 2016, and it allows for a mode of “work” that did not exist even a few decades ago.

As technology continues to shrink the effective distance between every human on the planet, it’s melting away geographic barriers between co-workers and ushering in a new type of globally distributed workforce. For startups and larger companies alike, embracing and innovating around this new model of work will be critical to competing in a dynamic global market.

Why Distributed?

It’s no secret that great local talent is expensive and difficult to hire across a number of industries. Meanwhile, smart and ambitious folks around the world are looking for the best opportunities, which may not be available in their local economy. (As my friend Niko Bonatsos likes to say: talent is universal, while opportunity is not.) This is not a new revelation -- companies have been looking across geographic barriers to find talent for decades.

In the 1980’s, a critical mass of western companies started moving specific business functions to specialized contractors overseas. These companies were able to lower their operating expenses by taking advantage of cost-effective talent, increasing their profitability while fueling the growth of a massive outsourcing industry in countries like India and Bangladesh. Cognizant, one of many large IT outsourcing companies based in India, generated a whopping $12B in revenue over the last twelve months alone.

Historically, though, there’s been a big gap between the individuals on either side of this business relationship due to communication difficulties, cultural differences, and educational limitations. Distributed contractors traditionally aren’t considered "co-workers”, but rather part of an outsourced team hired to handle a business process (e.g. a call center) or a discrete project (think: the overseas engineering team building a website for a U.S. company). Some firms also offer “outstaffing” services, where their employees act as remote full-time employee equivalents for their clients; however, this has traditionally been relegated to roles like software QA versus higher-level, strategic work.

The Case for the Virtual Employee

While outsourcing and entry-level outstaffing will always exist and will likely continue to grow in the future, we’re now seeing a new type of model emerging: the distributed organization, where employees at every level of the company are geographically distributed. Automattic (the creator of Wordpress, valued at over $1B), for instance, is a fully distributed company, with remote workers in 43 countries, while Stack Exchange and Upworthy give all employees the option to work from home if desired.

A map of Automattic’s global, distributed team (www.automattic.com/about)

At its core, like everything in our capitalist system, this trend is driven by companies’ fiduciary responsibility to maximize returns to shareholders.

Research (e.g. from Jerry Porras at Stanford) has shown that superior long-term business performance comes from tightly integrated, values-driven teams. At the same time, it’s often substantially less expensive to build distributed teams than local ones. The way to optimize on both fronts, then, is to deeply integrate talented remote workers into a company’s culture, values, and organizational structure.

This has become a possibility only recently with advances in collaboration technology. Just as phones revolutionized the way business was done in the 20th century, collaborative SaaS products like GitHub, Jira, and Asana, team messaging applications like Slack, and a variety of video conferencing platforms are altering the modern workplace. “Working” with someone across the world today could mean a video conference paired with a screen share and exchanging links in a Slack thread. It won’t stop here -- just think about the potential impact of virtual and augmented reality on distributed work.

Compare all of this to a landline phone call and a fax machine 30 years ago: it’s a different ball game.

Cultural Integration: The Glue of a Distributed Future

Despite rapid advancements in collaboration technology, it’s still nearly impossible to find a true substitute for “real life” when it comes to building trust and establishing company culture. Technology is critical to bridging the gap, but there’s a broader challenge that revolves around training and education.

Startups are now rising to meet this challenge. Andela, a company we backed at Spark, is an interesting case study in this category. Andela helps companies scale by deeply integrating remote software developers, based in Africa, into their tech teams. They focus on the selection and training of software engineers, helping them not only master coding but also culturally integrate into teams thousands of miles away. Andela developers are close with their clients, posting on each other’s Facebook walls, joking on Twitter, and sharing funny GIFs. When they video conference into an all-hands meeting, Andela developers don’t just sit back and listen; they actively give feedback on engineering milestones and product roadmaps.

It’s a bit ironic: as technology to empower remote communication improves, the challenge for startups increasingly lies in perfecting the human component.

What we’re witnessing today with distributed teams emerging is just the beginning of a broader movement. The real opportunity lies in taking this model of a distributed team and applying it to the global workforce. By harnessing the massive, untapped talent pool that exists in every corner of our planet, we will be able to maximize the potential of companies and ideas -- regardless of where they originate.

Growing up, I’d spend entire weekends plowing through stacks of books, frustrating my younger brother to no end when he wanted to hang out. (He still teases me about this.) I particularly loved science fiction, and my favorite author in high school was Neal Stephenson; I devoured books like Snow Crash, The Diamond Age, and Zodiac.

For some reason, reading wasn’t a big part of my life during college, graduate school, and in my first job. I was focused on other things — initially immersing myself in a new field (physics/materials science), and then helping to grow a fledgling startup (StartX).

When I joined Spark a couple years ago, two things changed. I started traveling for work and spending a lot of time in the air with weak/no WiFi. I also became surrounded by partners (Andrew, Bijan, Ryan, and more) that deeply enjoy reading and talking about books.

I took the opportunity to start reading again, and I’m right back where I started. Not only do I enjoy reading immensely, but I think reading novels about the future also makes me a better investor. Science fiction writers are some of the most creative minds in the world, and I love experiencing what they are dreaming. My colleague Ryan at Spark has written a fantastic exploratory piece on the mental benefits of reading — I highly encourage everyone to read it here. 

Next on my reading list...

I have to admit that for long-form content, I’ve never made the transition to e-books. Maybe I’m a dinosaur, but I like the way a book feels in my hand.

Like many investors, I’m curious about the way technology will change the way we read and interact with written content. Scribd has a Netflix model for books. Goodreads built a reading tool and social network for readers (although I unfortunately feel like it was acquired before it realized its vision). Wattpad is a platform for emerging writers that also allows readers to deeply interact with content. Medium is a publishing and reading network built for short/medium-form content. Spark backed Tumblr back in the day, a product I use constantly (and which you’re using right now).

I think there is more innovation to come, and I hope to meet entrepreneurs with big ideas in this space. And let me know if you have any great science fiction book recommendations.

I just got back last night from a machine learning conference hosted by the University of Toronto Creative Destruction Lab. It was a well-organized conference that had tremendous energy. There is clearly something very interesting happening in machine learning right now.

One of the most important themes of the conference was the rapid acceleration of commercial opportunities for deep learning. There were a number of big developments in this area shortly before the conference, including Google open-sourcing its TensorFlow deep learning library. Geoff Hinton (legendary professor at University of Toronto, now employed by Google) gave a well-received talk on deep neural networks. Many of companies that presented (such as Atomwise and Clarifai) have deep learning at the core of their business. I believe that innovation in this area will only continue to accelerate as technology improves and businesses build crucial enabling infrastructure.

Geoff Hinton schooling the crowd on neural networks

After the conference, I hosted a dinner with my friend Blair Silverberg from DFJ for a small group of academics, journalists, entrepreneurs, and business people focused on machine learning. I made some new friends and left inspired by the market applications for machine learning, from real-time transportation optimization to deep learning-driven molecular diagnostics.

Not only was the conference a success, but I have to say that I was impressed by the city of Toronto. I had fun while I was there and ate well — which, if you know me, is always a priority. I also had a chance to visit with folks from the University of Toronto Creative Destruction Lab and the MaRS incubator. There’s an awful lot of talent in Toronto directing intellectual and creative energy toward startups.

Entrepreneurs get most of the glory in the tech community, and they truly do deserve our praise: they take massive personal risks and make a huge positive impact on the world with the technology products they build. Investors also get a lot of credit for technology successes — likely more than we deserve, although it sure feels good for us when we back a winner and get to be a part of their story.

Today, I want to focus on another another group of people who make an enormous contribution to our society through technology but are rarely given the spotlight. These are the folks who freely contribute to the Internet's open and community-driven information platforms. 

I'm talking about the professor who adds information to a Wikipedia article in her area of expertise and corrects a mistake she finds; the data scientist who contributes to an open source machine learning library that earns him no money, but supports thousands of developers in their work; the engineer who thoughtfully engages with peers on Stack Overflow and helps to curate the community's growing knowledge base; the concerned citizen who shares information about an unfolding news story on Twitter.

I was a daily active user of Wikipedia in college and in graduate school. It was an incredible resource for me as a student and remains an important part of my life to this day. I still remember pouring over the Wikipedia article on the Schrödinger equation my senior year, trying to make heads or tails of a completely different way of thinking about the world. When I last checked over the weekend, 930 users have contributed 2114 times to get this article to where it is now.

The quality and scale of this distributed collaboration -- without any financial incentive -- absolutely blows me away. 

These collaborators show us one of the best aspects of humanity: a love for knowledge and a desire to share it with others. The Internet unlocks this human urge on a global scale. This is one of the many reasons that I fundamentally believe technology to be a force of good in our society.

We were thrilled to announce our investment in Andela earlier this summer. Andela is building a marketplace to connect the most talented developers in Africa, whom they train, to technology companies seeking talent.

I worked closely on this deal and had the opportunity to travel to Lagos with my partner Alex Finkelstein as part of our diligence process. I’m excited to continue working with Jeremy and his incredibly talented team as they grow Andela into a global leader in education.

Alex and me visiting with Andela at their Amity campus in Lagos, Nigeria

I previously wrote a blog post discussing three ingredients for why VCs invest, so I thought I would quickly shine some light on our investment thesis for Andela under that framework.

1. Could this have a huge impact?

We think Andela has the potential to become an enormous company.

Andela’s goal is to train 100,000 software developers in Africa over the next 10 years. Each developer generates substantial revenue as a remote employee. The company pays its developers generous salaries, invests in ongoing education, training, and infrastructure, and keeps the rest. By training and employing tens of thousands of developers, Andela has the opportunity to create a powerful global workforce tuned to the technology economy and potentially billions of dollars in enterprise value. Realizing this big vision will require skillful execution around operations and geographical expansion (see #3 below); however, after spending time with the founders, team, and developers in Lagos, we became believers.

Not only is Andela targeting a large business opportunity, but it also has the potential to improve people's’ lives in regions that suffer from persistent youth un/underemployment. When we visited the Lagos campus, we saw first-hand how local businesses that serve Andela’s software developers are flourishing. The developers are spending their salaries on food, clothing, entertainment, and more in their local communities. Additionally, they’re saving money and getting access to health insurance, which are critical pieces for a stable middle class. By creating a highly educated, well-trained, and well-compensated community of technology professionals, Andela is helping to build a true middle class in Lagos and soon in cities all over Africa. These are also just the sorts of entrepreneurial folks that will become job creators in the future, which is one of the many reasons why local governments are so supportive of what the company is doing.

2. Could this company have a sustainable competitive advantage at scale?

Andela is building a two-sided marketplace, where the value received by agents on one side of the marketplace increases with the number of agents participating on other side. As Andela grows and improves its pool of developers (supply), it can more easily acquire, service, and retain top clients that are looking for high-quality developers and specific skill-sets. As its client-base (demand) increases in size and variety, Andela is able to more effectively recruit new developers and retain existing ones by paying the most competitive salaries, investing in more specialized educational training, and offering the most prestigious program (see #3). This network effect (supply drives demand; demand drives supply) creates a long-term structural advantage for Andela as the company achieves scale. This advantage is strongest on a local basis, so while they are growing their business in Lagos, the company just opened a campus in Nairobi and is now in the process of aggressively expanding its geographical footprint to other major cities in Africa.

Andela also benefits from local economies of scale as its locations grow in size. Many of the expenses on the ground (e.g. real estate, labor) turn out to scale sub-linearly with the number of software developers employed at that location. As each location grows larger, the company benefits from these economies of scale and improved purchasing power for goods like computer equipment and furniture, allowing it to operate at a higher margin or offer more aggressive pricing than smaller competitors.

3. Does this business have short-term barriers to entry that will allow it to grow to the point where it builds a sustainable competitive advantage? 

The operational complexity of this business is a substantial barrier. There are few teams that have the ability to build high-level relationships with US software companies while running coding bootcamps and building engineering campuses on the ground in Africa; Andela happens to be one of them. While having a US team is necessary for building the demand side of the marketplace, i.e. acquiring clients, building this business remotely from the US without staff on the ground in Africa would be nearly impossible. We saw first-hand how many resources we take for granted in the US — accessible real estate, power, efficient transportation, and internet access — are not at all a given where Andela operates. Not only does Andela need to manage these issues for its own team, but also for its growing community of developers, nearly all of whom currently train, live, and work on-premises.

Christina and I got married in SF a few weeks ago, and we just got back from a wonderful two-week honeymoon on the Big Island of Hawaii and in Tokyo/Kyoto in Japan. It was a fantastic trip and a long-anticipated opportunity for us to decompress post-wedding, unplug from email, and explore new parts of the world.

However, we did not unplug entirely from the internet on our honeymoon. Yes, we posted a few honeymoon Instagrams. More importantly, whenever we had WiFi (or sometimes even when we didn’t...) we pulled up Foursquare to help us figure out what to do and where to eat while traveling. Suggestions from friends and locals who knew those places well were important, but Foursquare ended up being the invaluable tool for us on-the-go and completely replaced a guidebook. (Disclaimer: Spark is an investor, but this is not a paid piece of content marketing, I promise.)

Aloha, Hawaii!

On the Big Island in Hawaii, Foursquare was a constant companion to help us decide where to eat. We gravitated away from anything in the “6-7” range and looked to things in the “8-9” range - places like Merriman’s (8.9) and Ulu (9.0). One place we found on Foursquare and initially discounted was Sushi Rock, an incredible sushi joint about 45 minutes away from where we were staying on the west part of the island. It’s a hole in the wall in a small town, and we would have never considered going without its whopping 9.1 rating on Foursquare. When a local also highly recommended the place, we decided to make the trip up to Hawi to eat there. It was one of the best and most inventive sushi meals we’ve had in the US. 

Dinner at Ulu Grill on the Big Island

Hawaii was easy to navigate, however, compared to Japan, and that’s where Foursquare shone for us. We used Foursquare as a map and as a navigational tool, pulling up restaurants, museums, and temples, and showing the Japanese kanji characters to taxi drivers. When we got to Kyoto, we looked at the highest rated temples on Foursquare and used that as our guidebook, visiting places like the Kinkaku-ji (golden pavilion), Ginkaku-ji (silver pavilion), Ryoan-ji (zen rock garden), and Kiyomizu-dera (pure water temple). We also had so much fun discovering local spots on Foursquare, some of which didn’t even have English names - like this incredible sake bar with a 9.1 rating, where we were the only foreigners in a smoky bar full of professionals drinking, eating sushi and dumplings, and chain-smoking at a long standing bar. 

Kinkaku-ji in Kyoto

This was our first time using Foursquare while traveling outside the continental US, and it was a fantastic experience. We felt like we had the wisdom of the crowd with us at all times, even in places where we could barely communicate.

VC firms like Spark see thousands of deals a year and only invest in 10-15. What makes us pull the trigger? 

Every investor is different, and people can get excited about wildly different companies. However, I believe you can boil down almost any venture investment thesis into three basic ingredients. This isn’t rocket science, but hopefully provides a simple window into how we think about things on this side of the table. 

1. Could this have a huge impact?

Imagination around what could be “huge” is a VC’s secret sauce. Spark happens to be a firm that’s very product and team-focused: we invest in high-performing teams building beautiful products we believe will capture the imagination of the world. Most VC firms are highly focused on business model and market, carefully analyzing a company's unit economics and the total addressable market for a new product or service before committing capital. We do this too, especially at the growth stage, but we also invest in companies creating completely new product categories, like Oculus. When Spark invested in Twitter, we had no idea how big the “market” was, but we believed that the product and team had the potential to change the way that people communicate globally. (To be clear — I had nothing to do with either of these investments and hadn’t yet joined the firm.) 

When we make an investment, we’re about to get married to a founding team for what could be a decade. We take this commitment seriously and think critically about the team’s ability to shoot for the stars and push through the hardest times because of the passion they have for their product. That's why we have backed some founders in our portfolio multiple times: we trust them and know that they are thinking big. 

From a financial perspective, discipline around only investing in companies that could be huge is probably the most important feature of a venture firm. For a $450M fund like ours to meet LP expectations, it needs to contain multiple exits in the $B range, or at least one big multi-$B outcome. (1) 

2. Could this company have a sustainable competitive advantage at scale?

Most businesses don’t fall into this camp, but venture investments need to. Businesses that could be huge in theory but have no competitive advantage at scale are difficult to invest in as a venture firm, as they are easily displaced over the 10-year investment horizon of our deals. The ideal is to invest in a company that could be winner-take-all in its category. 

Oftentimes, for internet companies this advantage has to do with network effects — for example, the “strong” network effects of a communication platform (e.g. Kik, Whatsapp, Snapchat), where users pull other users onto a platform because it’s necessary for their product experience, or the “weak” network effects of a marketplace (e.g. Kitchensurfing, Uber), where increased supply drives increased demand and vice-versa. For enterprise software businesses, the moat is often based on platform / data lock-in (e.g. SurveyMonkey, eShares). Sometimes, you get all three (e.g. Slack: a communication tool, app marketplace, and system of record for collaboration). (2) 

3. Does this business have substantial barriers to entry?

Operationally complex businesses or ones that rely on IP have the advantage of keeping out hordes of copycats and newcomers. This isn’t a necessary condition, and only provides a short-to-medium-term advantage. However, barriers help us to believe that we’re picking the right horse in the race in a potential winner-take-all category (see #2) where a long-term sustainable competitive advantage can be built. Postmates is an example of a logistically complicated business where only a few competitors have achieved any scale in the US. Freight Farms has substantial IP around its vertical farm technology and produce growing techniques, which gives it a leg up on the competition in building a distributed food production network. Biotechnology firms invest in companies that go through the FDA process, a harrowing ordeal and huge barrier to entry. (3) Oftentimes, an unusually talented technical team or one with a substantial “unfair advantage” (relationships, etc.) in their given industry creates an effective barrier and discourages would-be competitors. Palantir is a success story in this category.

Putting these three ingredients together, you have the recipe for an enormous outcome. With a huge business opportunity, the potential for a long-term competitive advantage at scale, and barriers for newcomers, there’s the chance for a truly defensible category-leader to break out. Companies like this are career-defining investments and are the ones that drive outsized returns in venture capital.

Ultimately VCs are human, and venture capital is a human business. We care about these three ingredients, but we also invest with our heart. When we can really see a founder in his or her product and know that the passion runs deep, that’s what helps us to become believers. Out of the thousands of teams we encounter each year at Spark, the 10-15 we back are the ones that win both our minds and our hearts.

Thanks to Andrew Parker, Ryan Shmeizer, Danya Cheskis-Gold, and Christina Phillips for helping me to edit and think through this piece.

--

(1) Some quick VC math on this, which is applicable to funds of our size/stage at Spark: Limited partners are looking for 3x+ returns over 10 years when they invest in a VC firm. VCs who lead early-stage deals like Spark own roughly 15% of a company when it exits. Therefore, each $1B exit returns ~$150M to a VC fund. Most venture investments fail or merely return invested capital, so only the companies that exit for many times our investment valuation drive performance. You can do the math on what the aggregate exits need to be for a fund like ours to return 3x+. It’s a lot. 

(2)  I couldn’t be more excited about our recent investment in Slack, both as an avid user and for these more intellectual reasons.

In the 20th century, we perfected the art of centralized industrial production. We built big farms, factories, and power plants. This was a critical chapter in human history, allowing us to scale quickly as a species and develop modern cities, transportation technologies, large-scale IT infrastructure, and high-throughput agriculture. 

We can now produce things cheaply in high volumes. However, in a world where we expect immediacy and quality, the slow movement of goods oftentimes translates into frustrating inefficiencies and suboptimal product experiences. Transportation hasn’t changed substantially in decades; neither has our electrical grid. I can send bits via WhatsApp to my friend in China instantly via mobile data networks, but it’s still difficult to move atoms. As a doctor, I might urgently need a custom medical device to be fabricated while I’m in the operating room; it’s not helpful to have one shipped a week later from an overseas manufacturing facility. As a consumer, I would rather eat a fresh tomato plucked from a vine nearby over one that’s traveled three thousand miles on a flatbed truck. And as a home owner in a developing country, I would prefer to produce my own power locally than rely on power from a plant delivered via an expensive and unreliable grid.

I see this product experience and efficiency gap as a huge opportunity for distributed production systems — ones where goods are produced where they are consumed. Technology entrepreneurs, driven by increasingly demanding consumers, will realize these opportunities as computers become cheaper and more powerful according to Moore’s Law, internet access improves, and advanced industrial technologies become more widely available and physically smaller.

Residential solar installation from SolarCity

Distributed energy is an interesting case study. Energy transmission is an inherently wasteful and inefficient process, with around 8% of power lost in transmission in the US and substantial costs incurred in coordinating energy allocation from massive power plants. Over the last 10 years, solar has gone from an environmentalist dream to a large global industry ($50B+) powering a growing distributed energy network. In the US alone, over 500,000 homes are now powered by solar. This growth has been driven by decreasing solar panel costs and improved energy efficiency, consumer environmental awareness, government subsidies, and the rise of solar finance platforms like SolarCity, Vivint, and Sunrun. This movement has only just begun in emerging markets, where there’s a big opportunity for distributed solar and energy storage to empower consumers and businesses to achieve energy independence and circumvent unreliable national grid systems.

Freight Farms - Leafy Green Machine

Agriculture is now undergoing a similar revolution. With the water crisis in California, we’re even more aware than ever of the environmental toll (water, fuel, spoilage) from growing and transporting crops thousands of miles from big farms. For this reason, and because fresh local food actually tastes better, many consumers are willing to pay to support local farms and access better produce. Widespread internet access is a key enabler of this local food trend, with marketplaces like Good Eggs helping to connect local farmers and consumers. Distributed farming technology is also rapidly improving and helping to drive down local food production costs. We invested in a company called Freight Farms, which sells a cloud-connected indoor farm that can operate 365 days per year in any environment. These farms, which leverage low-cost LEDs and small computers, pump out fresh produce while using 90% less water than a traditional farm. They fit nicely in parking lots and are beginning to dot major cities across the US, right where the produce is consumed.

Reblog from Spark's Tumblr:

-

At Spark, we like entrepreneurs and companies with causes who use technology to make a difference. Not only do they make the world a better place, but they also hire and retain better and are generally more persistent in achieving their goals.

There are few greater opportunities for positive impact than in food. With our environment straining to meet existing agricultural demands, a growing global middle class, and population on track to reach nine billion people by 2050, improving the way food is grown, distributed, and prepared is certainly a worthy cause.

Last week, we partnered with Freight Farms and Silicon Valley Bank to host an intimate dinner for food/tech innovators. Cafe ArtScience was the perfect venue for this discussion; we partner with entrepreneurs working at the intersection of art/design and technology/science, and Café ArtScience is a restaurant, bar, and café where culinary art, science and design meet the sustainable future of food. The dinner brought together a group of roughly twenty entrepreneurs, academics, investors, and food industry professionals. The goal of the evening was to gather the ingredients for a great discussion – delicious food, innovative people, and an excuse to get together – and see what interesting themes emerged.

Automation

Automation, an area where we see huge opportunity at Spark, was one major theme we explored that evening. With ever-improving robotics technology and the ability to efficiently collect, process, store, and learn from massive amounts of data, we believe that automation is going to be a key driver of innovation along the food value chain. We spoke with Harvest Automation, which uses agricultural robots to improve the efficiency of various types of farms. We learned about Cabbige’s business intelligence solution to help local farmers access better data to improve their pricing on a daily basis, allowing them to earn more on their yield. And, course, our colleagues at Freight Farms shared their exciting vision around a cloud-connected, automated urban farm.

Getting Meals on Plates

We also heard from entrepreneurs looking to put meals on consumers’ plates as they demand access to better and healthier food, both at home and in the workplace. With investments in Postmates and Kitchensurfing, we’re putting our money where our mouth is in this category as well. We enjoyed talking with Just Add Cooking, which is doing direct-to-consumer meal delivery (similar to Blue Apron and Plated) using locally sourced ingredients, and LeanBox, which brings healthy meals to medium-sized businesses typically underserved by catering and food services businesses. Six Foods told us about the “Chirp” — a new type of high-protein corn chip powered by cricket protein. And those of us who like to tinker couldn’t hear enough about SproutsIO, which is building an automated indoor home hydroponic garden.

We’d like to thank these entrepreneurs and the other great folks from the Boston food industry who spent time with us last week. We’re excited to continue bringing together folks in the food/tech ecosystem and playing our part in supporting this movement. If you’re building a startup that’s pursuing a worthy cause in food/tech, please reach out and tell us about it.

I recently participated in an interview for Requests for Startups, a new startup newsletter. I had fun with this. Interview linked here and copied below:

What startup verticals interest you most right now?

I’m highly interested in food and agriculture technology. From farm to table, our food ecosystem is rapidly changing. Startups have a huge opportunity to disrupt the food and agriculture supply chain by leveraging new technologies: think low-cost LEDs for indoor farming, drones for imaging, bioengineered and alternative foods, ubiquitous smartphones, etc.

At Spark, we’re an investor in several companies that are focused on improving the consumer dining experience, including Postmates, Kitchensurfing, and Cover. We’re also interested in the infrastructure that will enable our food system to scale sustainably. My first investment at Spark (sponsored by my partner Todd Dagres) was in a company called Freight Farms, which is powering water-efficient urban agriculture through modular hydroponic farms.

What are your biggest predictions for the year ahead?

2014 was the year when Uber became a household name across the US and much of the world. When I talk to Uber drivers (and Postmates and Lyft drivers), the #1 reason they like their job is flexibility. The ability to “plug in and plug out” of the workforce when it’s convenient is incredibly empowering. It enables folks to realize their full potential: to pursue side projects and other part time jobs, manage their family life, and go to school.

In 2015, I think tech-enabled labor marketplaces like Uber will continue to scale quickly and touch more and more parts of our workforce. Companies like Work Market (one of ours at Spark) that help corporations to manage contingent workforces will benefit from this trend, as well as companies building services for the independent contractor, such as Stride Health and Zen99.

Are there any specific company ideas that you really want someone to build and would potentially fund?

I would love to fund a company that has made a fundamental breakthrough in producing nutritious, delicious, and cheap food with a low environmental footprint. Along these lines, one trend that’s fascinating to me is the use of insects to make high-protein food that’s actually quite tasty and requires a fraction of the resources (feed, water) of traditional animal meat. The problem right now is that it’s still much more expensive to grow insects than chicken.

If you could wave a magic wand and instantly have any imaginable solution to a problem you're facing (personally or at work), what problem would you solve?

While flights are expensive, I think that the UX for commercial air travel is fundamentally broken. The airline business is a tough one — it's highly regulated, capital intensive, and sensitive to macroeconomic issues and weather. Flights are often delayed or canceled at the last minute. Going through security is a pain. Getting onto the plane is difficult and there’s never enough room for luggage (cue the dreaded “sir, the bins are full, and we’re going to gate check your bag…"). Seats are cramped and airplanes are always too hot or too cold.