February 2017

small washpo

The term “artificial intelligence” is widely used, but less understood. As we see it permeate our everyday lives, we should deal with its inevitable exponential growth and learn to embrace it before tremendous economic and social changes overwhelm us.

Part of the confusion about artificial intelligence is in the name itself. There is a tendency to think about AI as an endpoint — the creation of self-aware beings with consciousness that exist thanks to software. This somewhat disquieting concept weighs heavily; what makes us human when software can think, too? It also distracts us from the tremendous progress that has been made in developing software that ultimately drives AI: machine learning.

Machine learning allows software to mimic and then perform tasks that were until very recently carried out exclusively by humans. Simply put, software can now substitute for workers’ knowledge to a level where many jobs can be done as well — or even better — by software. This reality makes a conversation about when software will acquire consciousness somewhat superfluous.

When you combine the explosion in competency of machine learning with a continued development of hardware that mimics human action (think robots), our society is headed into a perfect storm where both physical labor and knowledge labor are equally under threat.

The trends are here, whether through the coming of autonomous taxis or medical diagnostics tools evaluating your well-being. There is no reason to expect this shift towards replacement to slow as machine learning applications find their way into more parts of our economy.

The invention of the steam engine and the industrialization that followed may provide a useful analogue to the challenges our society faces today. Steam power first substituted the brute force of animals and eventually moved much human labor away from growing crops to working in cities. Subsequent technological waves such as coal power, electricity and computerization continued to change the very nature of work. Yet, through each wave, the opportunity for citizens to apply their labor persisted. Humans were the masters of technology and found new ways to find income and worth through the jobs and roles that emerged as new technologies were applied.

Here’s the problem: I am not yet seeing a similar analogy for human workers when faced with machine learning and AI. Where are humans to go when most things they do can be better performed by software and machinery? What happens when human workers are not users of technology in their work but instead replaced by it entirely? I will admit to wanting to have an answer, but not yet finding one.

Some say our economy will adjust, and we will find ways to engage in commerce that relies on their labor. Others are less confident and predict a continued erosion of labor as we know it, leading to widespread unemployment and social unrest.

Other big questions raised by AI include what our expectations of privacy should be when machine learning needs our personal data to be efficient. Where do we draw the ethical lines when software must choose between two people’s lives? How will a society capable of satisfying such narrow individual needs maintain a unified culture and look out for the common good?

The potential and promise of AI requires a discussion free of ideological rigidity. Whether change occurs as our society makes those conscious choices or while we are otherwise distracted, the evolution is upon us regardless.

This column originally appeared in The Washington Post.

 

Read more

small washpo

We all know that there are two main ways to make money in business: by selling yourself (a “service”) or selling something you create (a “product”).

I work daily on growing the greater Washington region’s innovation community, and I am constantly reminded of the fundamentally important distinction between these two business approaches. Our region’s ability to produce rapidly growing innovation companies depends on how successfully we make the distinction.

To illuminate the differences between service- and product-based businesses, I turned to members of the entrepreneurial networking group FounderCorps who have successfully built technology product businesses.

They all agree that a product-based business can grow more rapidly and achieve greater scale. “If you can find something that people want and can afford, it is easy to make lots of them,” said Dendy Young, principal of McLean Capital. This was echoed by Stefan Midford, chief executive of Natural Insight, a Sterling company that makes software to manage retail workforces, who said that a product, once created “can be sold to many companies and individuals.”

Once you understand the benefit that comes from making something once and duplicating it without customization, it’s easy to understand why “massive scale and rapid growth almost always comes from product companies,” observes Eric Koefoot, CEO of PublicRelay, which sells a media monitoring product.

Why can’t service-based companies flourish as fast? Jim Condon, principal of the consultancy CC Group, believes there is a very simple explanation. For a service business, he says, growing revenue generally requires “a commensurate growth in staff.” Ben Foster, an advisor to many local startups, adds another dimension: “service companies have to invest significant time and energy into every incremental sale, understanding the unique customer needs and building a custom solution to address them.”

So how does this affect our region?

Koefoot describes the greater Washington innovation community as being “disproportionately deep in services and businesses that deliver services.” Indeed, market data support this statement. For example, a recent survey by TandemNSI revealed that of almost 1,000 regional cybersecurity companies, only 1 in 20 is a product-based business. Similar patterns are found in the software and healthcare industries.

Because service-based entrepreneurship dominates the greater Washington region, our innovation community’s ability to provide new high-value jobs and wealth creation is constrained. We must shift to creating more product-based companies to achieve higher growth from our knowledge work.

The good news is that the greater Washington region is home to many successful product companies. Companies such as Cvent, AOL, Opower, Sourcefire, Blackboard, Weddingwire, Broadsoft and others. The challenge is to create enough density of success to allow innovators to learn how to change from a service to product-based business model.

“Many service companies fail when trying to convert to a product company – they need to be run, marketed and priced very differently,” Elizabeth Shea, CEO of SpeakerBox Communications, pointed out.

Conversion from service to product requires the development of an unexpected skill: the ability to say “no” to a customer that asks for customization, Foster added. That’s tough if you have been in a service business, where each customer relationship is based upon a unique transaction.

So local entrepreneurs agree that building a product-based business is a skill that can be taught, but that we must make an effort as a region in providing these lessons. And by acquiring this skill, we just might be able to reach our potential.

This column first appeared in The Washington Post.

Read more

small washpo

While Washington area success story Invincea being sold for more than $100 million last week is newsworthy, founder Anup Ghosh’s journey is in some ways even more important. To provide some lessons learned in the trenches, I asked him to share highlights of his experience.

Ghosh described a product discovery process that began long before he even thought about starting a business. As a program manager at the Defense Advanced Research Projects Agency, Ghosh realized that when it came to cybersecurity “as good as we were on offense, we were equally vulnerable to attacks from our adversaries. In other words, we were throwing rocks from glass houses and we needed to re-think defense.”

Over the years, I have heard from many current and former program managers that DARPA is an atypical government entity. It is tasked with creating advanced technologies to avoid technological surprise from our international adversaries. It moves fast, and demands a high level of commitment and engagement. “Sprint” is the word I hear most frequently when DARPA employees describe their work pace.

Program managers like Ghosh are term-limited and given a short number of years to identify new approaches to solve big problems. Ghosh’s description of working at DARPA was very consistent with what I had heard from others. When he completed his term, Ghosh had accumulated expertise and insight into an area of technological importance. Unlike some who returned to research, he wanted to develop a commercial product.

Conventional wisdom states that entrepreneurs go to angel investors to get capital to create a product. While this approach works well in markets with ample risk capital, such as Silicon Valley, it isn’t usually successful in our region, where such capital is hard to find.

So instead, Ghosh got creative and used resources more readily available locally. He joined George Mason University as a research professor, and sought out federal research and development funding to advance his technology vision. In effect, a university community and the federal government became his angel investors!

The tactic worked. Ghosh created promising new cybersecurity technology without outside capital. He was able to show venture capitalists a completed product and get the first of a number of rounds of funding.

With the capital, Ghosh grew an impressive team including Steve Taylor, a Mason alumnus now Invincea’s longest-tenured employee, experienced cybersecurity entrepreneur Norm Laudermilch as chief operating officer and head of product, and Mike Daniels, Network Solutions founder as chairman of the board.

It was an uphill climb as he hand-picked people who could grow a software product company. Finding these workers, and having them collectively win, is to Ghosh the most important part of the Invincea story. And that success deepens the talent pool of experienced software product entrepreneurs in the greater Washington region, thereby creating a fertile environment for a new generation of successful cybersecurity product companies.

Invincea’s story arc is a reminder that by watching how a creative individual took advantage of the region’s distinct advantages, we learn about a model for a software product start-up path that is easily repeated and scaled.

For the benefit of our nation’s economic future, as well as for our region’s economy, it is a lesson worth learning.

This column first appeared in The Washington Post.

 

Read more

small washpo

Success comes from awareness of one’s limitations and opportunities, and the willingness to see how actions influence others. Now is the time for the greater Washington region to challenge itself to look in the mirror and spot its weaknesses and act accordingly.

Yes, the business community must get ready to pivot.

We know all too well the strong connection between our local economy and the federal government. The proximity has been our region’s biggest strength, and yet sometimes its largest burden. The magnitude and diversity of our business community is often obscured by the “business of government” when outsiders view our region. The ability to generate significant wealth by serving the government represents a large portion of our activity, and we have embraced that.

However, for those concerned with the future of our region and its ability to generate high-paying jobs, the complex relationship between our business community and the government has long been a concern. For years, we’ve been identifying and promoting opportunities to divert our economy away from the shadow of government.

Sequestration was a strong reminder that our reliance on the federal government could have adverse effects. As federal spending fell, our region’s high value job employment stagnated. For a time, new employment in greater Washington lagged behind other metropolitan areas, and that’s when a broad cross-section of business leaders and politicians finally stood up and took notice.

The message was clear: greater Washington needed to diversify its economy and lessen its reliance on and interdependence with the federal government.

Business groups took up this message, reports were commissioned, plans were made and everyone agreed that something had to be done to address this pressing issue.

Sadly, as the specter of sequestration passed from the public consciousness, the urgency to make changes also passed. Some argue that our region has diversified. Others believe the lesson to be learned is that even if federal spending shrinks for a time, things eventually return to normal.

 

Professor Stephen Fuller, one of the leading thinkers on our local economy, last week pointed out that economic diversification hasn’t accelerated. In fact, the data show that the reverse is true. In a sobering report, he describes how the seven industrial clusters most likely to create new high value-added jobs in our community are not growing the way they must. Our reliance on federal government spending continues, and we are falling behind in our need to create high-value employment.

Meanwhile, there are clear indications that the federal government’s spending in the region could be at risk. The Trump administration has signaled plans for a federal hiring freeze. Entire government agencies could be downsized or eliminated. True, indications are that national security spending will increase. However, it is impossible to predict whether the net effect of these trends will create more or less federal spending in our region.

I believe that questions of how to cut our dependency on the federal government will become more pressing. Fuller’s research reminds us that we as a region must be more nimble — ready to adapt and embrace employment not necessarily linked to the federal government.

Greater Washington is about to have its pivotal moment. Will it be ready?

This column was originally posted at WashingtonPost.com.

Read more