Let’s first talk about the overall system.  Watson is a question-answering system that represents 100 person-years effort by a team of 25 IBM scientists.  Watson is running on 750 Linux servers with 2880 cores, 15TB of RAM and providing 80 teraflops of computing.  Its software architecture is based on UIMA and integrates a variety of machine learning algorithms that have been used to develop approximately 3000 predictive models.  These models are running in parallel every time Watson is trying to answer a question. The predictive models were tuned over several months by utilizing Watson’s results from a series of 134 “sparring matches” with past Jeopardy winners. Watson’s “knowledge base” was seeded with 70GB of curated, i.e., noise-free, full-text documents and eventually grew to 500GB, consisting of additional curated documents and information derived from the seed database.  The final knowledge base represented approximately 200M pages of textual content.  The text data was preprocessed using Hadoop.  However, Hadoop was not used while Watson was competing in Jeopardy.  Watson’s knowledge base also included various sets of handcrafted rules including rules that provide clues on what to look for and others that describe strategies on how players select within a category so that they can find Jeopardy’s daily double.

Watson operates through a cycle of hypothesizing an answer, gathering evidence that supports each hypothesized answer, evaluating the probabilities associated with the collected evidence and proposing the final answer.  As audiences discovered over 3 days of play, all this is done in “warp” speed.

There are several reasons of why we should care about Watson as a Big Data analytics system.

1. The interaction with Watson through spoken language will be particularly important for the broader use of analytic applications by business users.  The use of such applications is often limited because business users are often intimidated by their complex interfaces.  Watson demonstrated convincingly that natural spoken language interaction with computers is no longer science fiction or the result of Hollywood special effects teams. In prior posts I wrote about Watson exhibited superior natural language-understanding skills; and this is a really big deal.  It was able to address the inherent ambiguity of natural language and understand word-meaning and context.  Jeopardy’s statements represent much harder problem descriptions than the queries we typically pose to search engines.  They use puns, double meanings and misspellings to convey meaning.  While for a human this may be second nature, for a computer it is a very hard task.
2. Watson is making decisions and is responding to questions by utilizing a knowledge base of textual, unstructured information, rather than a knowledge base of well-structured concepts and rules that has previously digested and represented using a knowledge representation language, such as CYC.  In this respect Watson is significantly different from other question-answering systems such as Wolfram|Alpha.  Stephen Wolfram wrote an interesting post regarding the different approaches taken by Alpha and Watson.
3. While the data sets Watson was analyzing during Jeopardy were relatively small by Big Data standards, being able to quickly and effectively analyze unstructured data is representative of many big data analytics situations, where you don’t always know what data you will need to analyze, where it will come from, how large each data set will be, how clean it will be and how long you will have to provide an answer.
4. Watson concurrently utilizes a large number of predictive models to analyze big data and come up with answers in real time.  This is significant because it provides another important approach to analyzing big data, i.e., rather than parallelizing a single analysis algorithm and then using MapReduce to apply it on a big data set, as is typically done in various Hadoop/MapReduce implementations, Watson applies several different predictive and scoring algorithms concurrently.  Some of these algorithms may be parallelizable and thus able to take advantage of MapReduce.  The application of these algorithms on text data is particularly important to IBM since the majority of its customers possess a lot of such data.  Admittedly the data analyzed by Watson was curated and clean.  As mentioned above, the majority of big data is not of such quality.  IBM will need to test the system’s performance with noisier data, since corporate data is rather noisy, as well as with voice and video data.  Moreover, incorporating online data will introduce additional noise which will undoubtedly impact the Watson’s performance.  Watson cannot deal with incremental additions to its knowledge base regardless of the form these additions take, i.e., text documents or rules.  Such additions necessitate the re-tuning of the predictive models used.
5. According to IBM, the majority of the software used in Watson is open source.  That we can build such a sophisticated system from open source components is a feat in itself.

Of course IBM is very interested in business applications of the technologies that made Watson so successful.  These application areas must involve the use of complex language, including the use of such language in the data to be analyzed, the need for reaching high-precision responses/decisions/actions using ambiguous and noisy data, as well as the need to provide responses with confidence in real-time.  Some initial application ideas we discussed following the show included:

1. Medical diagnosis, including telemedicine.  The body of medical knowledge is becoming large and growing very fast. Being able to work directly from text rather than having to first represent medical knowledge in some intermediate language, as was the case with expert systems in the past, could represent a big breakthrough.
2. Technical support.  Help desks use mostly text data to provide answers to product issues. Increasing the accuracy of these responses while improving the overall user experience is important. 
3. Insurance claims analysis.  This is another area where the captured data is in text form and a system like Watson can be used to analyze it and provide a better user experience when consumers interact with their insurance provider.
4. Other application areas we discussed included online advertising, air traffic control and financial portfolio creation where Watson’s real-time analytics can be the central component of an overall solution.

The business model under which to offer Watson-like technology is another issue IBM is thinking about.  Should such a system be offered as a service or as on-premise software?  Of course it will depend on the application.  One of the ideas we discussed was to run Watson as a service and charge by the difficulty of the question asked or the importance of the answer provided.  This may be something that IBM can do for medical applications, for example, in collaboration with its traditional partners Mayo Clinic and Cleveland Clinic.

The Watson team deserves all the kudos they have received.  Not only for their win in Jeopardy! but, more importantly, for their technical accomplishments, the water-cooler discussions that resulted the day after each broadcast which will hopefully make more people (particularly young people) interested in technology and engineering, and the technical conversations they motivated among engineers and scientists about what is possible regarding big data analytics and human/machine interactions.

In a previous post I wrote that people, universities and research labs, risk capital, and modern infrastructure (legal, financial, physical, technical) make up Silicon Valley’s basic ingredients.  But during this three-day visit I was able to really appreciate:

1. Stanford’s central role in this ecosystem.  Fifty years after Professor Terman started what is we now know as Silicon Valley, Stanford continues to be not a mere member of the Valley’s ecosystem but its core.  It appears that each one of its students is a potential entrepreneur ready to start or participate in an innovative company.  Professors advise students on company concepts, often fund these concepts, and actively participate in the resulting companies.  Science and business classrooms, cafeterias, and the hallways of the campus buildings are the places where ideas are exchanged and where teams are formed.  Recognizing the quality of faculty and students and the IP they create, investors “walk the hallways” in search of the next fundable idea.  The University actively encourages the smooth flow of IP from its labs to the startups knowing that it will benefit from their success in multiple ways, as the founders of HP, Google, Yahoo, and of countless other companies have demonstrated over the years.
2. The ease with which university professors, laboratory researchers and students move between academia and industry, research and development.  These people know that without repercussions they can leave their academic or research positions to start companies and create a product or service out of the IP they created in their labs.  But they also know that they can as easily return to their old positions to continue their academic or research endeavors.
3. The willingness and drive to rapidly prototype an idea, test it with prospective customers, evaluate the results, repeat the success and discard the failure, learn from either outcome, change/pivot as appropriate, and try again.
4. The ecosystem’s completeness, connectedness, efficiency and high execution speed.  Many have spoken about the ecosystem’s characteristics but only as you drive from a meeting in Palo Alto (Trident Capital) to one in Mountain View (Fenwick and West) and back to one in Menlo Park (Silicon Valley Bank) you come to realize the efficiency advantages offered by close proximity (VCs, lawyers and bankers within a 10-mile radius).  You also realize the advantages of being part of a people-network that has been created and strengthened over many startup-related transactions, focuses on the entrepreneur, and is based on openness, trust, sharing knowledge (successes and failures) and optimism.
5. The entrepreneurs’ ability to select investors and advisors not based on the money they can invest, but based on the value they bring (connections to customers, connections to executives, help anytime and anywhere on anything from closing the next customer, interviewing and helping to hire the next executive, to refining the business model).

Several countries from around the world, but also many US regions, are trying to replicate Silicon Valley’s ecosystem and economic model in the process of creating innovation- and entrepreneurship-driven economies.  Successful will be the ones that analyze this ecosystem and model and properly adapt them, rather than trying to directly copy them.  In their analysis and adaptation they must take into account their regional characteristics, peoples’ competencies and temperaments, university environments and IP portfolios, as well as their overall economic environments.  While they will make mistakes, they need to remain optimistic, flexible and be willing to experiment, analyze, pivot if necessary and try again until they get it right, just like everybody in Silicon Valley does.

Starting with a bit of history (you can find a lot more detailed historical accounting in Steve Blank’s blog), Silicon Valley was created over a period of 60 years and today is the home to roughly 10,000 companies employing 1M people.  Its growth and success were based on multiple waves of technological innovations: electronics, semiconductors, computer hardware, software, biotechnology, internet, and more recently cleantech.  This 60-year period has been marked by several ebbs and flows through which we kept our strength and optimism.  The US government played an important role in the creation and success of Silicon Valley.  First, it is important to remember that the US government played a very important role in launching Silicon Valley and keeping it going for at least the first 10-15 years of its existence.  As early as the 1950s several companies were spun out from Stanford University to develop microwave technology for the Cold War under government contracts. During the 60s government-sponsored projects around spy satellites and ballistic missiles gave rise to many successful semiconductor companies.  It was since the late 60s, along with the rise of institutional venture capital, that academic and private sectors drove our region’s success.  However, with subsequent technology waves, the government’s role became one of providing the right incentives for people and companies to jump in, but then become less involved.

The establishment and enduring success of Silicon Valley is based on several factors, all of which comprise important and necessary ingredients:

1. The area is a magnet for top people.  Silicon Valley is not about the people who live here, it’s about the people who are attracted here.  They create a diverse network from a variety of disciplines and backgrounds.  Entrepreneurs, academics, investors, commercial and investment bankers, accountants, and lawyers make up Silicon Valley’s network.  Also, this talent is coming here from around the world.  For example, more than 50% of Silicon Valley’s companies are founded by immigrants, not “local” talent.
2. The Valley’s culture which is rather unique and is the combination of several elements itself.  It combines
1. a risk capital culture, driving venture capitalists to invest where others fear and where financial returns match the risks taken,
2. a risk employment culture,  driving people to forego the financial stability that employment in a large corporation brings and work in a startup in order to achieve technological success and financial independence,
3. a celebrate success culture, under which making money is as important as advancing technology and creating a successful product,
4. a learn from failure culture, where failure is acknowledged as much as success because there are important lessons that entrepreneurs carry from each failure to their next effort,
5. and finally a culture that celebrates technology entrepreneurs as heroes.

3. Silicon Valley is blessed with top academic institutions, such as Stanford University and Berkeley that try to bring on their campuses the world’s best students, professors and researchers, and view commerce with enthusiasm rather than disdain as the only way to sustainably serve the community around them. It is also home to several national research labs, such as the Lawrence Livermore National Laboratory, and SLAC that strive to be the best in their fields and attract government and industry funding to conduct basic research.
4. Silicon Valley is the headquarters of many large technology companies (many of which started as venture-backed startups) that play multiple roles in the ecosystem.  They collaborate and partner with startups providing them with revenue streams, help them shape their product, and give them access to their clients. They train the next generation of executives that will go on to staff the startups.  Oftentimes, they even acquire the startups providing profitable exits for founders, management teams and investors. 
5. The area is at the forefront of modern legislative and legal infrastructure for company creation and dissolution, hiring and rewarding the best talent, capital formation and allocation, IP protection and licensing, and exporting and importing of products and technology.
6. Finally, Silicon Valley’s top quality infrastructure from the multiple means of transportation, to buildings and other facilities suited for every type of technology and budget, to telecommunications and reliable electricity supply, and its temperate climate that provides an environment where people want to live and work.

These characteristics have led to the creation of a virtuous cycle of continuous innovation. Top talent is coming to Silicon Valley and become engaged in an organic, Darwinian process that results in the best companies rising to the top, has led to the emergence of global companies like HP, Intel, Genentech, Apple, and Google, to name but a few, and has made California’s economy one of the top in the world.  Many of these characteristics are importable but others may not be.  As Margaret O' Mara states here it is not enough to just state that you want to establish a Silicon Valley-like region and expect to automatically have its success.  The success of the various Silicon Valley-like efforts around the world will largely depend on deciding whether these characteristics are "importable" by a country, i.e., do they fit its culture, as well as on how well and how quickly these characteristics can be adopted and properly adapted to those environments.

Greece recognizes that entrepreneurship-driven innovation is a major driver in modern economies.  The importance of innovation-driven economies and the emulation of Silicon Valley’s successful model in creating such economies have been recognized as early as the 80s by Israel (see here for an overview) but later on by several other countries, and most recently by Russia, Saudi Arabia, and Jordan.  All these countries are trying to learn from Silicon Valley’s successes and failures, identify their regional advantages, and then adapt Silicon Valley’s model to enhance these advantages.  First, it is important to stress that Silicon Valley was created over a period of 50 years.  So any such initiative must be given the time to flourish and must become a national goal.  Second, one should be careful to properly adapt the Silicon Valley’s model to the region’s characteristics.  By drawing lessons from Silicon Valley and other similar efforts and tailoring them appropriately to Greece’s innovation model and Thessaloniki’s regional characteristics, Greece can make the Alexandrian Innovation Zone succeed and broaden its impact to its constituents and beyond.

...

I would like to now propose a few make easy steps that Greece will need to consider taking around the Alexandrian Innovation Zone.  Initially I propose that Greece applies these steps around only to the high-tech sector.  The goal would be to create some quick successes that can be used as a springboard for additional more fundamental and broader changes.

1.       Make easy and inexpensive the creation, financing and dissolution of high tech companies.  Starting up a business and getting the proper regulatory permits should be quick, and inexpensive; not rocket science.  Look at the examples of countries like Singapore where it is possible for a company to be established and funded within weeks.  Moreover make easy for small companies to obtain lines of credit from local or foreign banks so that they will have appropriate working capital.  Offer incentives to the banks to provide such loans.  Look at banks like the Silicon Valley Bank that specialize in this area.

2.       Teach entrepreneurs that from the beginning of their venture they must consider the world markets for their products rather than the Greek market alone and make that easy.  We all understand that the Greek market is small.  In this respect it is not dissimilar from the Irish, Israeli and Singaporean markets.  In these countries new companies learn to develop products with global appeal and start marketing them internationally very early on in the product development cycle.

3.       Make it easy to captivate the local talent as well as attract the best people possible from around the world for the technology sectors being pursued and the companies being formed.  Understand that Silicon Valley is not about the people who live here, it’s about the people who are attracted here. China, and India but also Singapore and Saudi Arabia are doing exactly that.  It is not important to only have the right technology innovators but you need to staff companies with people from other disciplines to make them successful.  Product managers who shape the product to address the market’s needs, engineering executives who lead the teams that transform an innovative technology to a product the market can use, sales and marketing executives who propel the product to the international markets, and CEOs that lead the companies to great outcomes must be part of any such effort.

4.       Establish intellectual-property offices in the country’s main universities and make easy and fast the spinning out of research from universities and national labs.  Negotiation of IP rights and licensing of patents should be done in weeks not months or years.  Make easy for professors and students to leave the university to start a company and then return back to the university.  The constant flow between academia and industry has been rejuvenating for many academics in the US, Israel, India, Brazil and Singapore. 

5.       Make easy the importation and exportation of foreign capital.  Learn from Israel, India, China and Russia, on how their governments continue to work with American venture capital firms to establish ways to start new investment funds in these countries and jump start their innovation-driven economies.  

6.       Make it easy for high tech startups to operate by offering to lower their taxes, and collect these taxes only after the startup’s losses are recouped and offer an additional tax-free period of a few years during which the company is profitable.

7.       Make easy the provisioning of infrastructure, from signing office space leases, to getting various permits, telephone lines and broadband internet.  Learn from best practices in Singapore, and Silicon Valley.

These are some of the short-term steps that you should consider.  These are some of the short-term steps that you should consider.  For the medium term you should consider providing a more growth-friendly corporate tax environment, as well as other incentives that will cause particularly large foreign high tech companies to come and establish technology operations in Greece, and making easy for entrepreneurs of failed companies to start new ventures.  Longer term you will need to work on providing a transparent financing environment, an area we are now advising Russia on, creating special incentives to attract larger pools of risk capital and making corporate M&A easy.

You will know that your initial steps are having the right impact when the pace of high tech company creation increases, the products of the financed companies start to attract the attention of the international markets, and capital begins to flow into Greece to be invested in these companies.

…

This will be a multi year, national, and difficult effort that must involve a unique partnership between government, academia, industry and people.  In order to succeed, Greece will need to execute flawlessly and consistently while competing successfully at the international stage.

…

I look forward to your comments on the above and additional suggestions for steps that will be important for countries like Greece to take in order to spur entrepreneurship-driven innovation and economic growth.