With “machine learning” excitement in full swing, Boston is buzzing as always with it’s own batch of startups, accelerators, and VCs, and I decided to do a deep dive into the local artificial intelligence ecosystem.
While Boston often doesn’t attract the same press attention as Silicon Valley, it certainly hasn’t stopped being a hotbed for innovation – and is argued to have it’s advantages (in addition to it’s disadvantages) over the Bay Area for both science and startups.
After my speech at the AI Applications Summit (based on our recent research on Machine Learning in Healthcare) in Boston on December 2017, I decided to stay for another entire week, doing my best to interview as many professionals in the AI community (investors, founders, researchers, non-profits) as possible, in-person.
Many of the recordings of those interviews are up on our artificial intelligence podcast now (with some still to be published), but I got in a lot of lengthy one-to-one chats outside of the podcast format – focusing entirely on the pros, cons, and trends in the AI ecosystem in Boston.
My goal was to create the most in-depth assessment conducted to date on where AI in Boston is headed, and what it means for researchers, startup founders, and existing enterprises who might be considering Boston to be their home.
I’ve broken down the article below into the following sections, feel free to skip ahead to what might be relevant to you:
We’ll start with the beginning, 60 long year ago:
Massachusetts Route 128 probably doesn’t ring a bell.
It’s a semi-circle bit of highway circling around Boston.
It’s the equivalent of “Silicon Valley”, but in New England. In the 1950’s, Route 128 gained prominence as a center of technology innovation and new businesses. Over time, the cache of the area faded. Innovation in Boston didn’t stop, but its baton was in large part passed along to the San Francisco Bay Area.
While we might think of Silicon Valley when we think about AI (Siri, Facebook, Google), the pre AI-winter days of AI were tied closely to the Northeast.
The term “artificial intelligence” is purported to have been created at the 1956 Dartmouth Conference, organized by the late John McCarthy and the late Marvin Minsky, among others.
It was Minsky and Seymour Papert of the MIT AI Laboratory who garnered millions from the Advanced Research Projects Agency (later known as DARPA) in the 60’s, and helped to make Boston a hotbed for AI innovation in the decades that followed.
While Misky’s optimism (for example, that “general intelligence” might be reached by the 1970’s) didn’t pan out, and AI funding dried up in both the 70’s and the 80’s, MIT remained a center of gravity for AI around the world.
Today, BU, UMass, MIT, Harvard, and others have thriving artificial intelligence programs for both grad students and undergrads, and many Boston locals believe that this strong academic base (Boston has about 35 colleges in the area) – combined with industry expertise – will make Boston uniquely poised to produce scientific and commercial successes in the years to come.
Like any other major tech city, Boston has seen the hype wave of artificial intelligence building over the last 3 years, and (not surprisingly, given Boston’s academic talent pool) they’ve spun out their own ecosystem of AI-focused startups, AI-focused VCs, and AI-focused consulting firms.
But has this influx of AI attention been helpful – or hurtful?
Thought over a dozen interviews I found some pretty consistent patterns about how the AI spotlight has impacted founders and businesspeople, and it’s a trend that was somewhat similar to what we say in our Montreal AI ecosystem feature.
We begin with the upsides:
It seems that the upside of AI hype has been around customer conversations. The statement that most sums up the sentiment I heard from Boston founders came from Brennan White co-founder and CEO of Cortex:
“Pros are all on the customer side… The first few months were “why would we want AI, what’s AI”… and just in the last 18 months that’s completely flipped. People are Google AI in marketing, they have the concepts in their head, they understand how powerful it is.
You talk to people who have researched and talked about this internally, so it’s more than just the visionary-types who are interested in an AI product.”
I heard lots of similar statements in my 10-day Boston tour. Customers no longer need a lengthy emplaination of what machine learning is – and even better – many are actively interested in the technology for a variety of reasons, including:
- Envy of the latest technology, and a yearning to “do AI” to appear or be innovative
- An understanding of the unique problems that AI and machine learning and best suited to solve
- A curiosity spurned on by the hype, and an interest in learning more
- A fear of being left behind by innovative startups and competitors who might be adopting machine learning already
While some of these conversations are more productive than others, it seems clear that there has been an aggregate “win” for founders and execs of AI companies.
Rana el Kaliouby is co-founder and CEO at Affectiva:
“I would say overall it’s been really exciting… In the automotive sector, we’ve realized that the AI hype has really helped.”
The fear of missing out and the vision of future possibilities is enough to leave many industry leaders eager to determine where AI fits into their business.
The same eagerness and curiosity that grew Emerj’s traffic by 10x in 2017 (mostly from a huge increase in organic search volume, for basic terms like “what is machine learning” to more sector specific searches about ML in healthcare) is the motive that spurs customer conversations for startups like Rana’s.
She mentioned that business leaders realize they need to “wake up” the the technology, and that on the aggregate that has led to a lot more inbound prospects. In addition, prospects often come to the conversation with a tertiary understanding of the technology to start with, and seem to require less hand-holding understanding on the basics of neural nets and AI in general.
While raising money is never easy, I got the impression that investors are more open to and interested in AI-based businesses. In addition, VC firms like Glasswing and Hyperplane have sprung up to invest in AI specifically. It seems that angel investment money might be easier to come by for seed-stage AI businesses than it was 5-6 years ago – which may change as the hype dies down or machine learning tools becomes more commoditized and accessible.
The upsides of customer interest and customer understanding seem to be offset by other developments that have washed ashore with the AI hype wave.
The popularity and curiosity around AI is no secret, and that’s brought about a wave of startups pitching “artificial intelligence” along with every other AI-related buzzword under the sun. Communicating one’s expertise to customers and (especially) investors now seems to be a more of a battle. In Rana’s words:
“I will say that as a company we need to educate the market because everyone is ‘AI’ now… things like explaining the definition of AI and ML, and why is that real as opposed to someone who isn’t doing this at all… explaining the value of data for neural nets… we’ve found that with all the hype we’ve had to do a lot more evangelism.”
Her sentiments were shared by almost all of the companies I spoke with, and the trend is probably even more obvious from the perspective of investors. Brendan Kohler is co-founder and CTO Sentenai and also co-founder at Hyperplane Venture Capital:
“Over the past 3 years as part of a machine intelligence focused venture fund we’ve seen a lot of companies mention AI…. but underneath there isn’t much going on.”
We’ve seen plenty of the same here at Emerj, and we spend a good deal of our time trying to either ignore or dispel the false advertising of AI.
In my recent “Enterprise AI Adoption” article I mentioned that out of an 100 so-called “AI” startups in any given sector, we find that only about 33% of them have any AI expertise at all (i.e. employees or executives with robust AI or ML experience in academia or previous jobs). This certainly makes our job challenging here when we play the role of industry analyst, and it makes for unique challenges for companies who need to convey they’re differentiators.
It was noted on a few occasions that while machine learning is a useful technology, it is often not necessary to deliver business results – and merely serves the purpose of marketing. Marshall Moutenot is co-founder at Upstream:
“Machine learning is a hammer. When there’s a shiny new hammer, people run around looking for things to hit with than hammer… whether or not the old hammer would do just fine.
…Sometimes it works. Some investors want to see that ML or AI. Certain customers get excited about ML or AI… but a lot of the time you can just use statistics.”
The uptick in customer interest (read: more demand) has come higher prices of AI-related talent (read: limited supply). Most of the Boston firms I spoke with mentioned that talent has gotten more expensive.
While some companies (including Sentenai) mentioned that they had grabbed ML researchers from firms in California (with a little help from Boston’s better cost of living, and an interesting project to work on), it’s likely to suppose that Massachusetts in general loses more talent than it would like to big West Coast firms who can pay astronomic salaries and come along with more cache, cash, and clout than most any newborn Boston startup can offer.
Some of our interviewees doubt whether thousands of founders riding the AI hype-wave (by faking the use of AI, or using the term merely as a marketing term), and the same may be the case in academia – where the influx of AI-related research may be watering down the new insights in the field at large.
Marco Lagi is a PhD who spent years applying AI at various organizations, an is now Senior Software Engineer – Machine Learning – at Hubspot in Boston. When pressed about the pros and cons of the hype around machine learning, he went immediately to the impact on research:
“The downside is that suppose I read an academic paper and spend a couple of months implementing that algorithm, but it doesn’t work since the paper has not considered some relevant factor,and has not been peer-reviewed; this increases my distrust into other papers I’m going to be skeptical about trying new stuff.
…this could also cause some of these cutting edge labs in ML to dwindle in funding”
One of the more humorous statements I heard about AI hype was during my interview with Massimilano Versace, CEO at Boston-based Neurala:
“AI is the ‘organic; for the tech industry. A few years ago you go to a grocery store you had one organic vegetable, now everything is organic.
Similarly everything is ‘AI’ today. It is sprinkled everywhere.”
The “organic” analogy is an apt one, and sometimes worth laughing about.
Based on my interviews it seems that the spotlight on AI has had a beneficial impact on AI-focused startups in the Boston area. Time will tell whether or not the increase in customer conversations will be enough to pay for increasingly pricey data science talent.
Ask a startup founder why they started their business in Boston, you have a 75 percent likelihood of hearing:
“Well, I wend to school here.”
Big surprise there. Boston attracts scientists and students, and many of them want to create companies.
So we know how they got here – but what are the advantages that Boston has to offer to AI startups?
Dominance in Robotics and Life Sciences
Boston surely doesn’t have a Facebook or Google equivalent, but it has the industries where it has world class strengths.
Rana of Affectiva framed this well:
“The ‘social wave’ that happened in SF didn’t happen here and Boston missed out on that, but I feel that it didn’t fit-in culturally speaking. But AI makes absolute technology you have deep technology as a focus, you have a lot of talent and a lot of businesses, you have the right industries [robotics, life sciences and medical,]… so I just feel like it makes a ton of sense.”
The “fertile crescent” of Kendall Square is home to a tremendous amount of biotech innovation, fed in large part by the academic powerhouses around it.
Brendan Kohler himself mentioned both drug discovery and robotics again:
“As an ecosystem Boston is really unique in that there has always been lot if talent in AI because the systems around robotic and drug manufacture have existed for a long time. In the machine intelligence space this is Boston’s domain expertise, the marrying of research and commercial experience with data to produce predictive systems.”
Vlad Sejnoha is CTO at Nuance Communications, one of the Boston area’s best-known AI companies (known to most consumers via their Dragon Dictation product). Headquartered in Burlington, MA, Nuance has an advanced AI lab in Sunnyvale, California (where I’ve interviewed their AI Director, Dr. Charles Ortiz), but Vlad mentions healthcare’s footprint as one of the good reasons to continue growing and hiring in Boston:
A lot of the OEMs now have design centers in the valley so it’s good to be in proximity to them there, but Boston has other advantages like a great healthcare industry for example.”
Robotics is another major area of focus for Boston-area tech. According to the Massachusetts Technology Collaborative:
- Massachusetts has 122 commercial companies in the robotics cluster. Between 2011 and 2015, 33 new robotics businesses were created in Mass., up 57% from five years earlier
- Roughly 90% of the mobile ground robotics supplied to the U.S. military were developed in Massachusetts
- Mass.-based robotics companies received over $190 million in private investments in 2015, equal to 23% of total U.S. funding and second highest of any state, only behind California.
Dr. Rodney Brooks is founder of ReThink Robotics, and is one of Boston’s most prominent robotics experts:
“We have a really strong ecosystem of universities, MIT has robot labs all over the institute… Harvard, NorthEastern, Brandeis, WPI has an undergrad major in robotics, my company covers a lot of people from WPI.
Amazon has 600 employees here in Boston last I checked [after buying Kiva Systems]… they stayed here dues to the massive number of people coming out of universities.”
The difference, suspects Rodney, isn’t just in academic talent, but in venture money with a focus:
“Robotics tends to be a longer play than many of the things we see on the West Coast…. this requires VCs and private equity firms that are in for a long term to see as their mission to influence certain industries…. there are a lot of VCs in Boston that have experience (both positive and negative) in robotics… and who perhaps are a bit more sophisticated in determining what’s a good invest and what’s a flash in the pan.”
This leads us into the next common theme when it comes to advantages of the Boston AI ecosystem:
Hard Problems and Hard Science
I heard a broad smattering terms like this during my interviews in Boston:
- “Boston likes more challenging scientific problems”
- “We like hard problems in Boston”
- “Boston focused on more long-term projects”
To some degree I chalked this up to some kind of excuse for not having a Google or a Facebook. Sort of like “well you can win at soccer but we’re better at baseball” kind of way of getting out of direct competition with Silicon Valley when it comes to startups.
While this supposition is likely right in part, it’s also a bit unfair of a judgement, and there’s likely a good deal of truth in the “hard problems” predilections of Boston’s founders. I suspect this has to do with Boston’s bent towards research and science.
In San Francisco, being a scientist isn’t cool. You’re a founder or a VC or nobody cares.
In the Boston startup ecosystem you can be a lawyer or a doctor… or a scientist… and not feel bad about yourself. This is a bit of my own perspective (having lived for years in both ecosystems), but it seems to ring true with what I heard chatting with Boston founders.
If we just reference the quotes above in the “Dominance in Robotics and Life Sciences” section above, we can see a number of hints at this bent towards hard, long-term technical problem solving.
- Rana mentioned the the social web never made it big in Boston due to cultural fit, but she feels AI is doing just fine because it involves “deep technology as a focus”
- Brendan calls “the marrying of research and commercial experience with data” Boston’s expertise
- Rodney said outright “Robotics tends to be a longer play than many of the things we see on the West Coast”
The good news about Boston creating lots of scientists (and not so many jobs) is that talent is more accessible. Mike Lacobucci is CEO of Interactions:
“In Boston, and to an extent in New Jersey and New York, we have a concentration of some of the most well known data scientists in this field, some of whom were part of inventing this technology which made hiring a lot of talent in one place easier.”
This may have to do with a difference in culture, which we’ll get into in greater depth in the “Boston vs San Francisco” section further below in this article.
The Universities, Obviously
MIT, Harvard, Tufts, BU, BC… and on and on.
The fresh infusions of young, bright, ambitious young people to Boston schools is a massive advantage to the Boston ecosystem, and more or less every single person I spoke with echoed the same.
With a little poking and prying my interviewees, I coaxed out what I think is the differentiator from Boston and other college-heavy areas (Chicago, Philadelphia, etc) is the following:
- Reputation. “Harvard” and “MIT” are downright eminent the world over. Gates and Zuckerberg went to Harvard (both famously dropping out), but 7 Presidents of the United States didn’t drop out, and got their degrees at Harvard (including George W. Bush and Barak Obama). Exceptionally bright and exceptionally wealthy people have been filling their brains at Boston schools for a long, long time, and the luster and austerity of the area keeps it’s magnetic pull strong.
- Proximity. The overlap and clustering of so many colleges and universities in Boston, MA and Cambridge, MA is said to have less of a wide sprawl (like that of the Universities in Atlanta, for example) than other college cities.
These obviously aren’t claims against other metro areas, they’re simple the stated “differentiators” of the Boston area, from the lips of Boston founders.
Certainly NYC, the San Francisco Bay Area, Chicago, and other cities have world-renown universities. At least to those of a technical or entrepreneurial bent, Stanford might be stealing much of the limelight. In the US, university “reputation” might have it’s oldest roots in Boston (Harvard was established in 1636 for crying out loud), but other cities have their strengths, and Boston had better not rest on its laurels.
As for college “proximity”, I did a bit of homework:
Apparently Los Angeles isn’t so far behind in terms of total university students in the greater metro area.
That being said, the Boston Metro Area spans a rather large stretch of land, down near Providence, RI and deep into New Hampshire:
The founders I spoke with were likely talking about that little tight circle, where indeed there are a great many elite universities. For all I know, a similar “tight circle” exists in NYC or LA as well, I haven’t been able to find any closer density numbers.
As with any city, Boston has both its strengths and weaknesses. Below I highlight two of the more commonly stated disadvantages of Boston for young companies.
Limited VC Funding
Andrew Antos and Nischal Nadhamuni from Boston-based Klarity Law are still early in their venture, and considering whether to make Boston their HQ or not. Nischal told me:
“When [Boston] AI companies get to the point where they need to access more talent at bigger scale, that’s when they will have to make the decision of whether to move [to the Bay area] or to open an office; in a typical case, around the time of Series A funding.”
Over and over again this idea is echoed. Every ecosystem outside of the Bay Area says the same thing (with the occasional exception of NYC). Namely: When you need the real money, you need to go west.
Andrew from Klarity mirrors his co-founder’s perspective:
“After closing our seed rounds our considerations are shifting, we are thinking about hiring and getting more customers – a lot whom are in the Bay Area, so having a presence there is important to us to some degree.”
Post seed stage, it’s time to go to the Pacific.
Sentenai’s Rohit Gupta spent years on the West Coast, and while he sees improvements in Boston’s funding resources, he – like most – sees room for improvement:
“The [Boston] investing community for a while really lagged, but is starting to catch up. The ecosystem is also about the corporates – like GE opening offices here… I think Boston is still playing catch-up when it comes to the full ecosystem, though it’s getting better.”
Some may frown on Silicon Valley’s emphasis on “going big or going home”, and the huge, expensive VC “dice rolls” of the Bay Area. Though it may be reeking havoc on the SF housing market, it seems to encourage an attitude that nonetheless is probably a plus when it comes time to grow something big:
The ability to stomach risk.
As mentioned, in SF you’re either a founder or nothing. I’m not judging this as either inherently good or bad, it just happens to be the case much more so than Boston. In Boston, you can be a doctor and not feel bad about yourself. You can be a $100MM company and not feel bad about yourself.
Not the case with SF, where that same kind of conservatism – for better or for worse – just doesn’t exist.
Byron Galbraith, Chief Scientist at Talla put it well:
“I find that Boston is incredibly conservative in terms of business and they tend to be very big serious businesses like biotechnology or hardware, things that have lots of capital expense, things that need a lot of people… you can’t get a couple of grad students and start a business, these companies tend to go to San Francisco”
In this statement Byron touches on the conservatism of Boston (sticking to what they know well), but he also touches on the “hard problems” bent mentioned by so many other Boston AI folks. Byron continues:
“There are reasons why the Bay Area is preferable to Boston; in terms of the tech scene I like Boston better, but access to capital and enthusiasm around new technologies is higher out there [Bay Area] in the general population.”
Byron and I jested about how far this “enthusiasm” goes. Apps like Yo seemed pretty ridiculous to the non-SF world. I told him that there is some kind of app to help people find places for their dogs to relieve themselves – and this company advertises by putting vinyl on the outside of busses (where I have seen them on a number of occasions in Hayes Valley, where I am).
Surely there is an absurdity to the yearning for the new and different – but there’s also no doubt that it’s paid dividends for Bay Area, and has built its repute as the world’s technology hotbed. Pros and cons both, for sure.
It may be this same conservatism and willingness to stick to the laboratory that holds it back from mingling with other communities as well. For all it’s focus on academia, Cortex’s Brennan White sees Boston as somewhat more isolated in terms of its academic focus, Boston founders and talent might not cross-pollinate as much as other ecosystems do:
“Having traveled to Toronto and San Francisco, there are three centers Stanford, MIT and University of Toronto, not necessarily in that order but Boston is a little bit more insular than the other cities, they seem to collaborate a lot more; from my personal experience, Boston people tend to stick with Boston people.”
I didn’t hear any founders other than Brennan mention this particular dynamic about Boston, but it seems to mesh well with an analysis on Boston.com (using data from Hired.com) that showed Boston to be “importing” less of its talent (as a percentage of the total talent pool) than other major tech hubs like Seattle and SF.
There seems to be ample evidence that Boston “brain drain” loses a lot more talent to the workforces of other cities – and tends to create more degrees than it does jobs (source: ComputerWorld). It could be the weather, but it likely says something about the
As mentioned is the “Disadvantages” section, Boston leaves much to be improved upon in the eyes of its local entrepreneurs. The image is more nuanced than simply saying it’s worse than SF, and I aimed to get a more robust picture of the current state of investment funding in the Boston area.
A Growing Focus on AI
Two or three years ago I first started chatting with the team at BootstrapLabs here in SF (including covering and speaking at their annual Applied AI Conference), as the first had shifted their focus entirely to focus on artificial intelligence. I’m sure they’re not the only SF firm to do so – and Boston has firms doing the same.
“The biggest gap here is that the funding ecosystem has moved out West, but that’s VC forms like Glasswing and Hyperplane are starting to turn the tide, so we are seeing a lot more companies stay here.”
I don’t have any data on the number of Boston companies leaving for West Coast funding and whether than number has increased or decreased in the last 5 years, so I’m unable to back up that perspective. Brendan has some interesting insight as to why Boston is behind SF in overall VC presence:
“If you look at it from an ecosystem perspective, most of the long standing Boston venture funds made all their money in the 70s and 80s in the mini-computer revolution, they did not adapt to the consumer internet.
Because of that, fewer people became angel investors, and so there was never a virtuous cycle in the 90s like there was in SF where people who made money invested it back in the ecosystem.”
Maybe Rana was right in stating that the consumer internet wasn’t a cultural fit for Boston in the first place, and Boston is better off focusing its VC funds on the B2B domains where it’s deepest strengths lie. This takes us to our next point.
More Focus on B2B
Over and over again, I heard companies say that their decision to stay in Boston was because they primarily sold to businesses, not consumers. Tomás Ratia García-Oliveros of Frase.io expressed that this focus extends to investment capital as well:
“Boston actual has some pretty robust venture capital, the volume maybe one third of SF, but there are VCs focused on AI here now, so if you have a market fit, there is really no shortage of funding.
If we are talking about B2B applications. If we talk about B2C, they are better suited for SF.”
Brennan from Cortex wasn’t the only interviewee to mention that Boston’s “conservatism” tends to make for longer fundraising cycles, and potentially lower valuations.
“I can only speak to early stage fundraising, but for me an early stage conversation… should be mostly based on the team and vision and maybe early traction and tech… that appears the be how it’s done in SF / Toronto / etc… that’s the level of conversation you have… it’s more contextually relevant (to seed stage).”
Brennan told me that Boston VCs tends to ask questions that would normally be asked of later stage companies. While this might be a sign of prudence, it might also be seen as a hinderance to rapid iteration and growth, and a potential fear of risk (which might have both pros and cons).
“I talk to my friends – it is universally true that we have a hard time getting term sheets around here [Boston], and it takes much longer. We have a much shorter conversations on the West Coast. One meeting on the West Coast versus six meetings here – and the West Coast will often also have much higher valuations.”
Silicon Valley is renown for this. Critics would say that it’s imprudent keeping up with the joneses in the form of venture capital. Admirers might say that Silicon Valley VCs have a better sense of what a good idea and a good team is – and are willing to risk big.
No matter who’s right, the impact on SF seems to be higher prices for everything, including for equity in startups.
There have been a lot of good comparisons between these two cities in the past.
They’re both university heavy cities with a larger neighbor city below them (NYC and LA, respectively). They’re both known for tech and innovation. They’re also quite different.
Instead of re-hashing what’s been said before, I’ll address the classic points often brought up in the Boston vs SF debate, and then I’ll add a series of factors that might be specifically relevant for business leaders working on AI.
We’ll take a quick look at some of the classic Boston vs SF comparisons, and then move quickly into how this might be relevant for AI founders and AI businesses.
Classic Comparisons Revisited
One of the best side-by-side comparisons of culture what I’ve seen was from an article in Xconomy:
The dichotomy of “Idea-driven” (Boston) versus “Money-driven” (SF) seems to map over the trend I heard with a lot of Boston founders who said that Boston likes “hard problems.” The idea of Boston being “Prone to testing and debate” also seems in line with this preference for sticking the science before zipping off to make billions.
We see “conservatism” come through in this comparison as well, with Boston leaning to the past and the known, and San Francisco leaning to the new and away from the known. Culturally, I’d say this rings true.
The idea of loyalty is mentioned here as well, and that’s something that we’ll get into in the sections that follow.
The Interesting Issue of “Talent Loyalty”
This is a point I want to address a point that I haven’t heard addressed in depth – but I heard it peek it’s way into nearly a dozen of my Boston conversations.
Rana of Affectiva put it well:
“A lot of our investors are West Coast-based… and I think it [the prospect of moving to the West Coast] comes up occasionally – I wouldn’t say there’s a lot of pressure. I also find that the talent here is very loyal and I hear that’s not the case on the West Coast.”
The first four or five times I heard this kind of statement, it felt like a weakness.
It seems hard to state that SF people are lacking a “loyalty” virtue that Boston people have in spades. It seems obvious, in fact, that there is simply less competition for out-of-this-world salaries and exciting companies to join.
From the perspective of technical employees, it’s simple: When $50MM dollars is getting raised here, $100MM is being raised there, and the hottest companies from all over the world at moving to your city or opening offices there, you have options.
In that sense, it makes Boston’s fears about “not having enough loyalty” from SF talent as really meaning “not a big enough deal for the best talent to pick us over the powerful, ever-growing competition in SF.”
Rohit from Sentenai:
“The challenge in the Bay Area is that that talent gets scooped up by Apple, Google, Facebook… because they can pay enough to live in the valley and still work on science.”
Interesting problems + good-enough-to-live-in-SF salaries = only possible by a select few firms. That seems to be much more the reality.
To some degree, I think this is true. Not all companies can deal with it.
While it’s feeble to suspect a lack of virtue in Bay Area employees, it’s not to say that if Boston companies were “cool enough” they’d all move to the valley. Of course there are advantages to Boston, too – the cost of living being one of them.
Jeremy Achin, founder and CEO of DataRobot (a firm with over 300 employees now, almost all of whom are in Boston), mentions the same sentiment, with a positive spin:
“I think the silicon valley is not even a realistic place to have a company because of the competition in terms of costs, the amount of recruiting needed (high attrition), people don’t seem to be as hard working because there are so many options.
You take Boston, I can get so much done here for the same amount of money; we [DataRobot] have a great competitive advantage by being here in Boston.”
However, I do believe there’s more to the issue of “loyalty” than just competition.
There’s also the culture of “the next big thing.” An eagerness to hop into what’s hottest… to find the next Uber, the next Facebook, and get in early. That kind of antsy excitement is fueled by opportunity, yes, but also by a kind of neurotic fear of missing out. I’m sure this tendency has in some ways encouraged big companies to be birthed in the Bay Area, but it’s also a kind of risk that isn’t something all companies want to (or need to) endure in order to grow and prosper.
All in all while I don’t see Boston humans as any more virtuous than those in the Bay Area, it’s true that SF has a feverish tendency to leap to what’s “next”, and it’s an environment where “so-so” companies will be hiring mediocre talent for a massive fee, and where a few bad press releases might mean your top talent jumps ship.
SF is B2C, Boston is B2B
Rudina Seseri is founder and Managing Partner at Glasswing Ventures:
“The way I wrap my head about this regional question is that when it comes to consumer plays… the West Coast has the DNA, the UI, UX talent that’s so hard to find and has quite an advantage.
On the enterprise security-type plays… I’d venture to say we have quite an advantage and strength here on the East Coast.”
Anecdotally I resonate with this idea. To some degree, there’s a level of design and UI that reaches it’s pinnacle in SF, where apps and digital experiences are beautiful and smooth – not just pragmatically workable. A Boston company website and an SF company website will often reflect this kind of local social pressure and level of expertise.
Abhi Yadav is co-founder and CEO of Zylotech is one of many founders who echoes this B2C or B2B divide.
“Being in the Bay area may not matter to you if you are in the B2B enterprise space. But if you have a mobile app or a B2C application, San Francisco is the place to be in mainly from a talent diversity point of view.”
Abhi told me that SF is an important place to hire for specific kinds of technical talent. Other founders mentioned the same. While funding was a first reason to visit or open an office in SF, talent acquisition was mentioned quite often as well.
As Marco Lagi put it, if you want to meet founders and investors, to share ideas about growing companies and raise the “big bucks” to make that growth happen, you’re best suited for SF:
“It [Boston] has easy access to the best talent, salaries are still lower than the West Coast – it’s easier here because there is less competition. Boston is a network of talent and not VCs or founders.”
Cost of Living – A Deeper Look
You can do a lot in San Francisco. If you want to become a billionaire before age 40 it’s arguably the best place on earth.
One thing you can’t do in the Bay Area – in your right mind – is bootstrap your company. Why?
- Rent is astronomical, as a founder your runway will be significantly shorter than it would be anywhere else
- The best talent isn’t interested in most startups. If you’ve built momentum as “the next big thing” (maybe 3% or less of all startups), you convince top talent to jump ship from tech giants like Google or Facebook, but without massive salaries or huge promise of excitement and success, you’ll be hiring mediocre talent for astronomical salaries
- The VS money is right there in SF… being there without using it is almost nonsensical. All you competitors are hiring talent and growing and paying massive rent with other people’s money – are you going to try to catch up with them running on profit?
Rohit Gupta of Sentenai put it succinctly:
“I have a son and when I think about education, I can actually raise a family and not worry about school and family support.
…The people with an entrepreneurial bent can actually make a case to work on science at a startup and be able to do that without living with 3 other people in a Mountain View apartment.”
(Fun fact: I lived in Mountain View for nearly a year, and it’s indeed common to see people with six figure salaries being crammed into small $3,500/mo apartments to save costs.)
The following cost-of-living calculator at Numbeo.com puts some of the broad cost comparisons in full view:
Consumer prices, groceries, and (most importantly) rent is clearly higher. Boston is no Kansas City, but it’s clearly on a different level that SF. Even this chart doesn’t capture many of the “costs of living” factors that matter for founders and their teams in Boston.
The challenges of “raising a family” in SF were brought up somewhat often:
- It’s possible to afford a home (with… no way… a yard) just 25 minutes from Boston for $450,000. There is nothing within 2 hours of SF (that’s livable) for under a million bucks, and even then your commute is 90 minutes both ways and you have little actual space.
- Many people with young children feel uncomfortable with the homelessness in San Francisco. Even on Patricia Green in Hayes Valley (where I sometimes go for walks with my wife), I’ll see a homeless person babble incessantly at a woman walking a stroller as she has to smile and scurry away. Many SF-ers have sympathy for the homeless and an eagerness to help, but still have a fear for their own safety – or that of their children.
These factors by no means take away from the myriad benefits of the Bay Area, but they are factors to consider for early-stage companies looking to pick their HQ location.
You’re not going to see airplanes delayed much in San Francisco. Good luck with Boston in the winter. In the last 5 years the city has been hit with a number of brutal storms – bringing feet of snow and slowing down travel, commutes, and commerce in the entire metro area.
While some might argue that snow is a plus for people who enjoy skiing, there’s positively zero business value to the interruptions and perturbations of the winter. The physical chill can be heart-warming for some, as snow brings with it:
- A real Christmas (no offense to the rest of America)
- A coziness of fairy-tale wonder of a New England winter, as encapsulated by one of our country’s greatest men of letters:
“Announced by all the trumpets of the sky,
Arrives the snow, and, driving o’er the fields,
Seems nowhere to alight: the whited air
Hides hills and woods, the river, and the heaven,
And veils the farm-house at the garden’s end.
The sled and traveller stopped, the courier’s feet
Delayed, all friends shut out, the housemates sit
Around the radiant fireplace, enclosed
In a tumultuous privacy of storm.”
– The first 8 lines from Snow Storm by Ralph Waldo Emerson
If you’ve lived in New England and that prose doesn’t tug at you little bit, you are probably a cruel and heartless person.
Romantic coziness aside, the bottom line for business is that Boston weather is bad, and in the dead of winter one could almost forgive the egregious relative cost of living in SF for the fact that you can wear sandals and bike to work year-round.
Location Choice – When to Pick Boston, When to Pick SF
I did my best to take all of my conversations with founders and turn them into a “when you might want to pick SF, and when you might want to pick Boston” chart:
To put a bit of color on some of the comparative points above:
- Companies that have tons of cash on hand already are more likely to put up with the costs of SF, adn survive the burn rate until the next fundraise
- Companies that need to raise big money might also be served by SF because of the density of massive VC firms, while companies that might not need to raise as much early on may not have those same needs
- Companies who can grow steadily and slowly might be better served in Boston, while companies who need to grow fast (for example, platform-based businesses that need to achieve a certain degree of market share or network effects fast) are almost inevitably best served by the Bay Area
It goes without saying that these aren’t “rules.” There are B2B companies who by all means should go to San Francisco. There are companies that need lots of funding that might stay in Boston.
These are simply aggregated ideas and preferences from dozens of AI investors, execs, and founders in the Boston area – put into a chart. Companies considering opening an office – or founders considering picking a “home base” – will hopefully find this little aggregation useful.
Those of you who have followed Emerj over the years are aware that, ultimately, my own interest in AI is predicated on its potentially massive ethical implications in the long term.
When I left Boston area in 2015, there was practically no focus on the social implications of AI (or, nothing compared to the kind of focus we see today) – so I couldn’t leave Boston without checking in with some of the folks on the ground with the ethical AI initiatives in the area.
I sat down in Ashley Heacock, Brendan Roach, Esam Goodarzi, and Marie-Therese Png, thanks to a connection from Cyrus Rhodes at the Harvard Future Society. The group is all in some way associate with either The AI Initiative at the Harvard Kennedy School, or The Future Society. In my open dialogue with them all, I tried to grab all the quotes I could about how the ethics and AI conversation is overlapping.
Some of the main point that the group brought up about the “pro” of the recent AI hype on the AI ethics conversation:
- The “hype” around AI has put the topic on the map, and made the ethical concerns of AI become more pressing and interesting for people in government, not just people in basic research and the business of AI.
An overview of the “con” of that same hype:
- Some of the AI-related panels and discussions involve people who don’t know what they’re talking about. Marie-Therese Png mentioned that having experts from multiple backgrounds is a good thing (philosophers, economists, machine learning researchers, etc…), but when economists are doing the explaining about how AI works, and AI researchers are explaining economics, we might be going about things incorrectly.
In Marie-Therese’s words:
“We use the hype just to start getting people aware of what this [AI] is, in terms of who has the right to talk about it and how should it be talked about, I think that it is multidisciplinary, engineers are good at getting things to work well, but might not be thinking about how things might affect human behavior.”
Esam Goodarzy brought up a point about the potential downside of adding a lot of enthusiasm atop a complex and challenging subject:
“It’s natural when something gets popular the conversation gets oversaturated with information; what I’m worried about the hype is that there’s so many topics that everything gets leveled out and in terms of what’s the most important aspects to be focusing our efforts on from an altruistic perspective.”
Admittedly, it’s challenging to know whether privacy, racial bias, or superintelligence are the most important on a utilitarian scale. Our own polling of AI researchers seems to suggest that there’s much disagreement about “What to focus on” – even among PhDs. The point is valid, however, and I suspect that Harvard’s effective altruism group (among others) is fleshing out some useful ideas about relative “importance”.
My hope is that a myriad of ideas will eventually separate the wheat from the chaff, and that while mislead conversations might do hard, an abundant of smart and well-intended conversations. Too much discussion is probably a good problem to start with. May the best ideas win.
Edmond Awad is a postdoctoral associate at MIT Media Lab, and he runs a project called Moral Machine, aggregating human ethical sentiments around AI moral dilemmas (such as self-driving cars having to choose between crashing, or hitting pedestrians).
“My impression is that – this topic when we started was mostly theoretic exercise, hypothetical. Most car companies didn’t take it into serious considerations…
In 2016 this was more about thought experiments. Since then, car companies seem to have taken this more seriously. The Department of Transportation deployed 15-point checklist… included documentation of ethical considerations, then they removed it and added it in a footnote.”
He mentioned that Germany formed a committee and drafted a set of rules for how driverless cars should behave.
The MediaLab and Harvard scholarship for AI governance, and now there’s an AI ethics branch, too.
There are LOTs of workshops on this now. For the first time they have a small workshop on AI ethics and society.
The recent AI collaboration between IBM and MIT (establishing the new MIT–IBM Watson AI Lab, thanks to a 10-year plan for IBM to invest $240MM in the lab) began with a call for papers, including papers about “advancing shared prosperity through AI”, a topic we might not have suspected to see just a few years ago.
The Future of Life Institute is based in Boston, and given the city’s academic leaning and historical reputation as the “Athens of America,” it’s likely to remain a hub for rich thought about AI’s ethical considerations.
While it’s hard to predict the future of Boston’s AI scene, a few near-term predictions seem safe to make:
- More and more Boston universities are likely to develop and expand their AI-related programs and courses
- The hype wave of AI isn’t dead yet, and using “AI” in pitch decks will likely be the norm in the few years ahead
- Staple industries like life sciences and robotics are likely to remain major areas of focus for applying machine learning and AI in the years ahead
Other than that, it’s anyone’s guess. Boston founders seem to have their fingers crossed that the “brain drain” won’t deprive them of the talent they need – and that hopefully Boston’s ecosystem can actually attract top tech and data science talent, in addition to training it.
I’d like to thank all of the interviewees who took the time to be interviewed for this piece, most of whom made time to catch up in person in the middle of a workday during my hectic 10-day Boston trip.
Interviews for the article were conducted by Daniel Faggella. Quote transcription and selection completed by Emerj Content Lead Raghav Bharadwaj.
Header image credit: Amtrak Downeaster