The year 2020 has been featured in many predictions and long-term visions in the past, implying not only the terminal point for the forecast or planning period but also a crystal-clear crystal ball. Now that the year 2020 is our present, we can clearly see where these prognostications went wrong and try to understand why they were so cloudy.
Read more here
Recent surveys, studies, forecasts and other quantitative assessments of the progress of AI highlight the increasing presence of AI in the healthcare industry, the assistance AI may provide in the future to workers’ cognitive tasks, and the continuing acceleration in data production and dissemination.
Source: CB Insights
Read more here
Yoshua Bengio speaking with Karen Hao at the EmTech MIT conference, September 18, 2019
[September 20, 2019]
Asked what is the biggest misconception about AI, Yoshua Bengio answered without hesitation “AI is not magic.” Winner of the 2018 Turing Award (with the other “fathers of the deep learning revolution,” Geoffrey Hinton and Yann LeCun), Bengio spoke at the EmTech MIT event about the “amazing progress in AI” while stressing the importance of understanding its current limitations and recognizing that “we are still very far from human-level AI in many ways.”
Deep learning has moved us a step closer to human-level AI by allowing machines to acquire intuitive knowledge, according to Bengio. Classical AI was missing this “learning component,” and deep learning develops intuitive knowledge “by acquiring that knowledge from data, from interacting with the environment, from learning. That’s why current AI is working so much better than the old AI.”
At the same time, classical AI aimed to allow computers to do what humans do—reasoning, or combining ideas “in our mind in a very explicit, conscious way,” concepts that we can explain to other people. “Although the goals of a lot of things I’m doing now are similar to the classical AI goals, allowing machines to reason, the solutions will be very different,” says Bengio. Humans use very few steps when they reason and Bengio contends we need to address the gap that exists between our mind’s two modes of thought: “System 1” (instinctive and emotional) and “system 2” (deliberative and logical). This is “something we really have to address to approach human-level AI,” says Bengio.
To get there, Bengio and other AI researchers are “making baby steps” in some new directions, but “much more needs to be done.” These new directions include tighter integration between deep learning and reinforcement learning, finding ways to teach the machine meta-learning or ”learning to learn”—allowing it to generalize better, and understand better the causal relations embodied in the data, going beyond correlations.
Bengio is confident that AI research will overcome these challenges and will achieve not only human-level AI but will also manage to develop human-like machines. “If we don’t destroy ourselves before then,” says Bengio, “I believe there is no reason we couldn’t build machines that could express emotions. I don’t think that emotions or even consciousness are out of reach of machines in the future. We still have a lot to go… [to] understand them better scientifically in humans but also in ways that are sufficiently formal so we can train machines to have these kinds of properties.”
At the MIT event, I talked to two Intel VPs—Gadi Singer and Carey Kloss—who are very familiar with what companies do today with the current form of AI, deep learning, with all its limitations. “Enterprises are at a stage now where they have figured out what deep learning means to them and they are going to apply it shortly,” says Singer. “Cloud Service Providers deploy it at scale already. Enterprise customers are still learning how it can affect them,” adds Kloss.
Many of these companies have been using for years machine learning, predictive analytics, and other sophisticated techniques for analyzing data as the basis for improving decision-making, customer relations, and internal processes. But now they are figuring out what deep learning, the new generation of machine learning, can do for their business. Singer has developed what he calls the “four superpowers framework” as a way of explaining what’s new about deep learning from a practical perspective, the four things deep learning does exceptionally well.
Deep learning is very good at spotting patterns. It first demonstrated this capability with its superior performance in analyzing images for object identification, but this exceptional capability can be deployed to other types of data. While traditional machine learning techniques have been used for years in fraud detection, for example, deep learning is very powerful in “identifying remote instances of a pattern,” says Singer.
The second “superpower” is being a universal approximator. Deep learning is very good at mimicking very complex computations with great accuracy and at a fraction of the power and time of traditional computation methods. “Whatever you can accelerate by 10,000x might change your business,” says Singer.
Sequence to sequence mapping is the third exceptional deep learning capability. An example would be real-time language translation. Previously, each word was translated in isolation but deep learning brings the “depth of context,” adding a time dimension by taking into account the entire sequence of words.
Last but not least is generation based on similarities. Once a deep learning model learns how a realistic output looks like, it can generate a similar one. Generating images from text is an example. Another one is WaveNet, a speech generation application from Google, mimicking the human voice. Yet another example is medical records anonymization, allowing for privacy-preserving sharing, research, and analysis of patient records.
EmTech 2019 also featured MIT Technology Review’s recent selection of “35 innovators under 35.” A few of these innovators got on the list because they developed and demonstrated a number of practical and successful applications of deep learning. These included Liang Xu and his AI platform that helps cities across China improve public health, reduce crime, and increase efficiency in public management; Wojciech Zaremba, using deep learning and reinforcement learning to train a robot hand to teach itself to pick up a toy block in different environments; and Archana Venkataraman who developed a deep learning model that can detect epileptic seizures and, as a result, limit invasive monitoring and improve surgical outcomes.
There is no doubt that Bengio and Hinton and LeCun have created in deep learning a tool with tremendous positive social and economic value, today and in the future. But they—and other AI researchers—insist on the ultimate goal being the creation of “human-level intelligence” or even human-like machines. Why do these experts in machine learning refuse to learn from history, from seven decades of predictions regarding the imminent arrival of human-level intelligence leading only to various “AI winters” and a lot of misconceptions, including unfounded fear and anxiety about AI? And why aren’t goals such as curing diseases, eliminating hunger, and making humans more productive and content sufficient enough to serve for them as motivating end-goals?
Originally published on Forbes.com
[September 27, 2019]
Language is imprecise, vague, context-specific, sentence-structure-dependent, full of fifty shades of gray (or grey). It’s what we use to describe progress in artificial intelligence, in improving computers’ performance in tasks such as accurately identifying images or translating between languages or answering questions. Unfortunately, vague or misleading terms can lead to inaccurate and misleading news.
Earlier this month we learned from the New York Times that “…in just the past several months researchers have made significant progress in developing A.I. that can understand languages and mimic the logic and decision-making of humans.” The NYT article reported on “A Breakthrough for A.I. Technology” with the release of a paper by a team of researchers at the Allen Institute for Artificial Intelligence (AI2), summarizing their work on Aristo, a question answering system. While 3 years ago the best AI system scored 59.3% on an eight-grade science exam challenge, Aristo recently correctly answered more than 90% of the non-diagram, multiple choice questions on an eighth-grade science exam and exceeded 83% on a 12th-grade science exam.
No doubt this a remarkable and rapid progress for the AI sub-field of or Natural Language Understanding (NLU) or more specifically, as the AI2 paper states, “machine understanding of textbooks…a grand AI challenge that dates back to the ’70s.” But does Aristo really “reads,” “understands” and “reasons” as one may understand from the language used in the paper and similar NLU papers?
“If I could go back to 1956 [when the field of AI was launched], I would choose a different terminology,” says Oren Etzioni, CEO of AI2. Labeling “anthropomorphizing” this “unfortunate history,” Etzioni clearly states his position about the language of AI researchers:
“When we use these human terms in the context of machines that’s a huge potential for misunderstanding. The fact of the matter is that currently machines don’t understand, they don’t learn, they aren’t intelligent—in the human sense… I think we are creating savants, really really good at some narrow task, whether it’s NLP or playing GO, but that doesn’t mean they understand much of anything.”
Still, “human terms,” misleading or not, is what we have to describe what AI programs do, and Etzioni argues that “if you look at some of the questions that a human would have to reason his or her way to answer, you start to see that these techniques are doing some kind of rudimentary form of reasoning, a surprising amount of rudimentary reasoning.”
The AI2 paper elaborates further on the question “to what extent is Aristo reasoning to answer questions?” While stating that currently “we do not have a sufficiently fine-grained notion of reasoning to answer this question precisely,” it points to a recent shift in the understanding by AI researchers of “reasoning” with the advent of deep learning and “machines performing challenging tasks using neural architectures rather than explicit representation languages.”
Similar to what has happened recently in other AI sub-fields, question answering has gotten a remarkable boost with deep learning, applying statistical analysis to very large data sets, finding hidden correlations and patterns, and leading to surprising results, described sometimes in misleading terms.
What current AI technology does is “sophisticated pattern-matching, not what I would call ‘understanding’ or ‘reasoning,’” says TJ Hazen, Senior Principal Research Manager at Microsoft Research.* Deep learning techniques, says Hazen, “can learn really sophisticated things from examples. They do an incredible job of learning specific tasks, but they really don’t understand what they’re learning.”
What deep learning and its hierarchical layers of complex calculations, plus lots of data and compute power, brought to NLU (and other AI specialties) is unprecedented level of efficiencies in designing models that “understand” the task at hand (e.g., answering a specific question). Machine learning used to require deep domain knowledge and a deep investment of time and effort in coming up with what its practitioners call “features,” the key elements of the model (called “variables” in traditional statistical analysis—professional jargon being yet another challenge for both human and machine language understanding). By adding more layers (steps) to the learning process and using vast quantities of data, deep learning has taken on more of the model design work.
“Deep learning figures out what are the most salient features,” says Hazen. “But it is also constrained by the quality and sophistication of the data. If you only give it simple examples, it’s only going to learn simple strategies.”
AI researchers, at Microsoft, AI2, and other research centers, are aware of deep learning’s limitations when compared with human intelligence, and most of their current work, while keeping within the deep learning paradigm, is aimed at addressing these limitations. “In the next year or two,” says Etzioni, “we are going to see more systems that work not just on one dataset or benchmark but on ten or twenty and they are able to learn from one and transfer to another, simultaneously.”
Jingjing Liu, Principal Research Manager at Microsoft Research also highlights the challenge of “transfer learning” or “domain adaptation,” warning about the hype regarding specific AI programs’ “human parity.” Unlike humans that transfer knowledge acquired in performing one task to a new one, a deep learning model “might perform poorly on a new unseen dataset or it may require a lot of additional labeled data in a new domain to perform well,” says Liu. “That’s why we’re looking into unsupervised domain adaptation, aiming to generalize pre-trained models from a source domain to a new target domain with minimum data.”
Real-world examples, challenges, and constraints help researchers address the limitations of deep learning and offer AI solutions to specific business problems. A company may want to use a question answering system to help employees find what they need in a long and complex operations manual or a travel policy document.
Typically, observes Hazen, the solution is a FAQ document, yet another document to wade through. “Right now, most enterprise search mechanisms are pretty poor at this kind of tasks,” says Hazen. “They don’t have the click-through info that Google or Bing have. That’s where we can add value.” To deploy a general-purpose “reading comprehension” model in a specific business setting, however, requires successful “transfer learning,” adapting the model to work with hundreds of company-specific examples, not tens of thousands or even millions of examples.
Microsoft researchers encounter these real-world challenges when they respond to requests from Microsoft’s customers. A research institute such as AI2 does not have customers so it created a unique channel for its researchers to interact with real-world challenges, the AI2 Incubator, inviting technologists and entrepreneurs to establish their startups with the help of AI2 resources. Lexion.ai is one of these startups, offering NLU software that organizes and reads contracts, and extracts the specific terms employees need for their work.
Unfortunately, human ambition (hubris?) hasn’t stopped at solving specific human challenges as sufficient motivation for AI research. Achieving “human-level intelligence” has been the ultimate goal for AI research for more than six decades. Indeed, it has been an unfortunate history, as a misleading goal has led to misleading terms which in turn lead to unfounded excitement and anxiety.
Fortunately, many AI researchers continue to expand what computers could do in the service of humanity. Says TJ Hazen: “I prefer to think about the work I’m doing as something that will help you do a task but it may not be able to do the full task for you. It’s an aid and not a replacement for your own capabilities.” And Oren Etzioni: “My favorite definition of AI is to think of it as Augmented Intelligence. I’m interested in building tools that help people be much more effective.”
*Opinions expressed by Microsoft’s researchers do not necessarily represent Microsoft’s positions.
Originally published on Forbes.com
IDC predicts that 175 trillion gigabytes of new data will be created worldwide in 2025
[July 23, 2109]
Data is eating the world. All businesses, non-profits, and governments around the world are now in full digital transformation mode, figuring out what data can do to the quality of their decisions and the effectiveness of their actions. In the process, they tap into IT resources and landscape that have changed dramatically over the last decade, offering unprecedented choice, flexibility, and speed, facilitating the management of data eating work.
Launched in 2012, DraftKings is a prime example of a new breed of data-driven, perpetually-learning companies. One of the few players in the market for fantasy sports, it has faced “unique challenges that haven’t been solved by other businesses yet,” says Greg Karamitis, Senior Vice President of Fantasy Sports. To solve these challenges, “we have to lean on our analytical expertise and our ability to absorb and utilize vast amounts of data to drive our business decisions.”
Founded by serial entrepreneur Ash Ashutosh 10 years ago, Actifio is a prime example of the new breed of IT vendors transforming the IT landscape from a processor-centric to data-centric paradigm, from a primary emphasis on the speed of computing to a new focus on the speed of accessing data. “It used to be that only the backup people cared about data,” says Ashutosh. “Then it was the CIO, and later, the Chief Data Officer or CDO. Now, every CEO is a data-driven CEO. If you are not data-driven, most likely you are not the CEO for long.”
Data “as a strategic asset” was the vision driving Ashutosh and Actifio in 2009, bringing to the enterprise the same attitude towards data that has made the fortunes of consumer-oriented, digital native companies such as Amazon. “We wanted to facilitate getting to the data as fast as possible, to make it available to anybody, anywhere,” recalls Ashutosh. Since only backup people cared about enterprise data 10 years ago, they were the customers Actifio initially targeted.
The value proposition for these customers centered on reduced cost, as Actifio helped them maintain only one copy of any piece of data, available for multiple uses, instead of maintaining numerous copies, each one for a specific application or data management activity. Actifio achieved this magic trick (e.g., reducing 50 terabytes of data to only 2 terabytes) by capitalizing on another trend shaking the tech world 10 years ago, virtualization. Replacing analog data with digital data gave rise to data-driven companies and replacing the physical with the logical—virtualization—created a new IT landscape.
It took a while for enterprises to adapt to the new IT realities. But around 2015 the cloud flood gates opened because of the business pressures to do everything faster and faster, especially the development of new (online) applications, and the fact that more and more enterprise roles and activities required at the very least some creation, management, manipulation, analysis, and consumption of data. These changes manifested themselves in Actifio’s business. In its most recent quarter (ended April 30, 2019), “60% of our customers used Actifio to accelerate application development, up from close to 0% in 2016,” reports Ashutosh, “and over 30% of our workloads today are in cloud platforms.”
The new attitude towards data as a strategic asset and the widespread availability of cloud computing have opened up new uses for Actifio’s offerings such as compliance with data regulations and near real-time security checks on the data. But possibly the most important recent development is the increased use by data scientists for machine learning and artificial intelligence-related tasks.
A very significant chunk of data scientists’ time (and their most popular complaint) is the time they spend on data preparation. And a significant chunk of the time spent on data preparation is simply waiting for the data. In the past, they had to wait between eight to forty days for the IT department to deliver the data to them. Now, says Ashutosh, “they have an automated, on-demand process,” providing them data from the relevant pool of applications, in the format they require. Bringing up the new term “MLOps” (as in machine learning operations), Ashutosh defines it as “allowing people to make decisions faster by not having data as a bottleneck.” The end result? “The more you give people access to data in self-service way, the more they find new and smarter ways of using it.”
As 40% of Actifio’s sales come from $1 million-plus deals, these new uses of data are not “tiny departmental stuff,” says Ashutosh. “Large enterprises are beginning to use data as a strategic asset, as a service, on premises or in multi-cloud environments.”
Large enterprises today learn from, compete with, and often invest in or acquire the likes of DraftKings, a startup that runs on data. Growing the business for DraftKings means making their contests bigger and bigger. “But if we make our contests too big and we don’t get enough users to fill 90% of the seats, we start to lose money really fast,” says DraftKings’ Karamitis. Balancing user satisfaction and engagement with the company’s business performance requires accurate demand predictions for each of the thousands of contests which DraftKings runs daily in a constantly changing sports environment.
“We need to absorb tons of data points to figure this out on a daily basis,” explains Karamitis. But these data points are not only based on DraftKings accumulated experience with past contests. Befitting a business living online, another important data source is social networks—“our users are giving us enormous amount of data in terms of what they are twitting to us, what they engage with and what they don’t, allowing us to understand better which ways we want to shift as a company and which way we want to build a product,” says Karamitis.
There is yet another source of the data that is driving decisions at DrafKings, possibly the most important one: The data DraftKings creates by constantly experimenting. “We create data points by running structured tests,” says Karamitis. They cannot run A/B tests, he explains, because running smaller size contests will not produce the same effect of larger size contests and because their users tend to communicate a lot among themselves and compare notes about their experiences. “We are willing to take the risk testing our underlying beliefs around user behavior,” says Karamitis, by changing the top prize or changing the marketing treatment, for example.
In the past, domain expertise was the key to a company’s success. Today, it is data expertise, and the skill of applying it instantly to new opportunities as they arise. “Our analytical expertise allows us to learn really fast, learn in the traditional meaning of learning from experience and figuring out what matters to our users and also learn in the more modern sense of building machine learning algorithms around the key principles our users care about,” says Karamitis. Learning from data has become a core competency that can be applied to new markets, a competency that should serve DraftKings well as it pursues the new business opportunity of legalized sports betting.
DraftKings—and the new breed of data-driven companies—are data science labs, creating data and acquiring new insights with continuous experimentation. Like good scientists, they test their hypotheses through carefully structured experiments, challenging their own assumptions about customers and markets. Karamitis recalls the start of the 2017 NFL season when DraftKings offered a new contest site: ”We had a very specific expectation as to who it’s going to appeal to and how big it will be. We were totally wrong, 100% wrong.” But because the new product was developed through experimentation, the data led to what is now a “super valuable product, so different from what we initially offered.”
Ashutosh predicts that over the next few years we will see a bifurcation of the economy into two segments, one focused on producing physical assets and the other comprised of data-driven companies. Like DraftKings, these companies, whether startups or established enterprises, will view data as a strategic asset and its analysis as a core competency and a key competitive differentiator.
And like DraftKings, these companies will increasingly resemble scientific labs, continuously learning through experimentation and creating new data points. Data growth will drive business growth as data continues eating the world.
Originally published on Forbes.com
Developed at the Artificial Intelligence Center of the Stanford Research Institute (SRI) from 1966 to 1972, SHAKEY was the world’s first mobile intelligent robot. According to the 2017 IEEE Milestone citation, it “could perceive its surroundings, infer implicit facts from explicit ones, create plans, recover from errors in plan execution, and communicate using ordinary English. SHAKEY’s software architecture, computer vision, and methods for navigation and planning proved seminal in robotics and in the design of web servers, automobiles, factories, video games, and Mars rovers.”
Read more here
[April 29, 2019]
Opening the CannaTech conference earlier this month, former Israeli prime minister Ehud Barak quipped that Israel is now the “land of milk, honey and cannabis.” Given the recent performance of the cannabis-related stocks traded on the Tel-Aviv stock exchange (Barak is Chairman of InterCure whose stock appreciated 1000% in 2018), are investors getting high on nothing more than a buzz bubble?
Behind the buzz about “marijuana millionaires,” Yuge market potential, and volatile stocks (InterCure’s stock nearly tripled earlier this year but is now 25% off its peak), is a serious 55-year-old Israeli enterprise of pioneering interdisciplinary research into the medical benefits of cannabis. Supported by a perfect climate for growing cannabis, it has led to a very supportive climate—academic, regulatory, and entrepreneurial—for developing botanical-sourced pharmaceutical-grade products. Like the rest of the world, Israel has considered cannabis (and still does) to be a “dangerous drug,” but unlike the rest of the world, it has not let the stigma deter its insatiable curiosity about cannabis’s therapeutic potential.
The entrepreneurial poster child for this long-held belief in the efficacy of medical marijuana is Breath of life (BOL) Pharma. Founded in 2008, today it is “the only company in Israel that is fully integrated throughout the value chain,” says its CEO, Tamir Gedo.
This means BOL Pharma is compliant with the GAP and GMP standards of the global pharmaceutical industry, governing all stages of production and distribution, from cultivation to processing to marketing of finished products such as tablets, capsules, inhalers, creams and oils. This unique competitive advantage is buttressed by BOL’s R&D function, currently involved with 32 Phase 2 clinical trials, and a 65,000 square feet production plant and one million square feet of cultivation facilities.
“You don’t see this kind of consistency in products around the world,” says Gedo. “Flowers are not consistent and if you don’t have consistency, you run the risk of having side effects at different times.” The need to overcome the challenge of developing medicine from an inconsistent botanical source is why 60% of BOL’s 200 employees have worked before in the pharmaceutical industry and why Gedo insists on staying focused on the company’s medical cannabis vision and not developing products for recreational use. “Our advantage is time,” says Gedo, “we’ve been doing it for many years.”
This time- and experience-based competitive advantage applies to the Israeli cannabis ecosystem as a whole. In the early 1960s, looking to make his mark in the academic world, Israeli chemist Raphael Mechoulam decided to focus on cannabis research because “in a small country like Israel, if you want to do significant work, you should try to do something novel.” Moreover, “a scientist should find topics of importance,” he says in the documentary The Scientist. “Cannabis had been used for thousands of years both as a drug [and] as a recreational agent, but surprisingly, the active compound was never isolated in pure form.”
Mechoulam and his colleagues isolated the chemical compounds of cannabis (which he called “cannabinoids”), specifically CBD (the main non-psychoactive component) and THC (the psychoactive component). In the early 1990s, they discovered the endocannabinoid system in the human body which is involved in regulating a variety of physiological and cognitive processes (including mood and memory), and in mediating the pharmacological effects of cannabis.
These discoveries have led to a vast body of research conducted in Israel and around the world on various aspects, medical and otherwise, of cannabinoids (see here, for example). With government support, both in terms of funding and regulation, Israel has become a center for medical cannabis R&D, with many academic institutions and companies “offshoring” research and clinical trials to Israel, having been prevented from doing it in the US and elsewhere.
Gedo calls this R&D-and-clinical-trials-as-a-service “open innovation,” providing the research and regulatory infrastructure for others to innovate and produce their own IP. But the infrastructure and accumulated experience and expertise also help BOL Pharma and other Israeli companies develop their own unique cannabis-related IP. For example, BOL has been working on unique new formulations which make the medical cannabis more effective by increasing its “bio-availability” (rate of absorption in the body), thus reducing the cost to the consumer and potential side effects.
When BOL entered the medical cannabis market a decade ago it did not have a lot of local (or global, for that matter) competition. Today, according to a recent survey published in Israeli business publication Globes, there are more than 100 Israeli companies contributing to the “current boiling point” of this market. These include companies growing and processing cannabis, or running pharma production facilities, or exporting Israeli know-how, or developing drug delivery mechanisms.
These companies are going after a worldwide medical cannabis market estimated to grow rapidly to $100.03 billion in 2025, according to Grand View Research. Most are private companies, but some may test the public markets before long—BOL Pharma may list on the Canadian stock exchange or in the US and Canndoc, another pharma-grade medical cannabis pioneer (acquired last year by InterCure), has recently submitted a confidential prospectus for a Nasdaq IPO.
The ever-growing market size estimates and increased activity in the public markets have drawn the attention of venture capital firms. Funding for cannabis startups in the US more than doubled from 2017 to 2018, reaching more than $1.3 billion, according to Crunchbase. The first quarter of 2019 saw funding more than double year-over-year, and earlier this month, Pax Labs (vaporization technologies and devices) raised $420 million at a valuation of $1.7 billion.
The most recent funding data for cannabis-related Israeli startups shows that only $76 million have been raised from 2013 to 2017, according to IVC Research. That number has probably increased considerably by now as just one cannabis-related startup, Syqe Medical (inhalers), has raised $50 million in its second round of funding at the end of 2018.
And there’s more to come, including Israel-US collaborations. OurCrowd, Israel’s most active venture investor (including in Syqe Medical), announced in January that it will partner with Colorado-based 7thirty to create a new $30 million fund focused on emerging cannabis technology companies in Israel, Canada and the United States.
At its annual conference last month, OurCrowd awarded 88-year-old Professor Raphael Mechoulam its Maimonides Lifetime Achievement award (the other winner was 100-year-old Professor Avraham Baniel, the inventor and co-founder of DouxMatok). In accepting the award, Mechoulam talked about his current work, predicting that “within the next decade, maybe less, we shall have drugs for a variety of diseases based on the compounds, the constituents of the [cannabis] plant and the constituents of our own body, the endogenous cannabinoids, and compounds that we make, that will be effective for a large number of diseases.”
Originally published on Forbes.com
“AI Career Pathways” is designed to guide aspiring AI engineers in finding jobs and building a career. The table above shows Workera’s key findings about AI roles and the tasks they perform. You’ll find more insights like this in the free PDF.
From the report:
People in charge of data engineering need strong coding and software
engineering skills, ideally combined with machine learning skills to help them
make good design decisions related to data. Most of the time, data engineering is done using database query languages such as SQL and object-oriented programming languages such as Python, C++, and Java. Big data tools such as Hadoop and Hive are also commonly used.
Modeling is usually programmed in Python, R, Matlab, C++, Java, or another language. It requires strong foundations in mathematics, data science, and machine learning. Deep learning skills are required by some organizations, especially those focusing on computer vision, natural language processing, or speech recognition.
People working in deployment need to write production code, possess strong back-end engineering skills (in Python, Java, C++, and the like), and understand cloud technologies (for example AWS, GCP, and Azure).
Team members working on business analysis need an understanding of
mathematics and data science for analytics, as well as strong communication skills and business acumen. They sometimes use programming languages suchas R, Python, and Tableau, although many tasks can be carried out in a spreadsheet, PowerPoint or Keynote, or an A/B testing software.
Working on AI infrastructure requires broad software engineering skills to write production code and understand cloud technologies.
[March 12, 2019] Tim Berners-Lee liberated data so it can eat the world. In his book Weaving the Web, he wrote:
I was excited about escaping from the straightjacket of hierarchical documentation systems…. By being able to reference everything with equal ease, the web could also represent associations between things that might seem unrelated but for some reason did actually share a relationship. This is something the brain can do easily, spontaneously. … The research community has used links between paper documents for ages: Tables of content, indexes, bibliographies and reference sections… On the Web… scientists could escape from the sequential organization of each paper and bibliography, to pick and choose a path of references that served their own interest.
With this one imaginative leap, Berners-Lee moved beyond a major stumbling block for all previous information retrieval systems: The pre-defined classification system at their core. This insight was so counter-intuitive that even during the early years of the Web, attempts were made to do just that: To classify (and organize in pre-defined taxonomies) all the information on the Web.
Thirty years ago, Tim Berners-Lee circulated a proposal for “Mesh” to his management at CERN. While the Internet started as a network for linking research centers, the World Wide Web started as a way to share information among researchers at CERN. Both have expanded to touch today more than half of the world’s population because they have been based on open standards.
Creating a closed and proprietary system has been the business model of choice for many great inventors and some of the greatest inventions of the computer age. That’s where we were headed towards in the early 1990s: The establishment of global proprietary networks owned by a few computers and telecommunications companies, whether old or new. Tim Berners-Lee’s invention and CERN’s decision to offer it to the world for free in 1993 changed the course of this proprietary march, giving a new—and much expanded—life to the Internet (itself a response to proprietary systems that did not inter-communicate) and establishing a new, open platform, for a seemingly infinite number of applications and services.
As Bob Metcalfe told me in 2009: “Tim Berners-Lee invented the URL, HTTP, and HTML standards… three adequate standards that, when used together, ignited the explosive growth of the Web… What this has demonstrated is the efficacy of the layered architecture of the Internet. The Web demonstrates how powerful that is, both by being layered on top of things that were invented 17 years before, and by giving rise to amazing new functions in the following decades.”
Metcalfe also touched on the power and potential of an open platform: “Tim Berners-Lee tells this joke, which I hasten to retell because it’s so good. He was introduced at a conference as the inventor of the World Wide Web. As often happens when someone is introduced that way, there are at least three people in the audience who want to fight about that, because they invented it or a friend of theirs invented it. Someone said, ‘You didn’t. You can’t have invented it. There’s just not enough time in the day for you to have typed in all that information.’ That poor schlemiel completely missed the point that Tim didn’t create the World Wide Web. He created the mechanism by which many, many people could create the World Wide Web.”
Metcalfe’s comments were first published in ON magazine which I created and published for my employer at the time, EMC Corporation. For a special issue (PDF) commemorating the 20th anniversary of the invention of the Web, we asked some 20 digital influencers (as we would call them today) how the Web has changed their and our lives and what it will look like in the future. Here’s a sample:
Howard Rheingold: “The Web allows people to do things together that they weren’t allowed to do before. But… I think we are in danger of drowning in a sea of misinformation, disinformation, spam, porn, urban legends, and hoaxes.”
Chris Brogan: “We look at the Web as this set of tools that allow people to try any idea without a whole lot of expense… Anyone can start anything with very little money, and then it’s just a meritocracy in terms of winning the attention wars.”
Dany Levy (founder of DailyCandy): “With the Web, everything comes so easily. I wonder about the future and the human ability to research and to seek and to find, which is really an important skill. I wonder, will human beings lose their ability to navigate?”
We also interviewed Berners-Lee in 2009. He said that the Web has “changed in the last few years faster than it changed before, and it is crazy to for us to imagine this acceleration will suddenly stop.” He pointed out the ongoing tendency to lock what we do with computers in a proprietary jail: “…there are aspects of the online world that are still fairly ‘pre-Web.’ Social networking sites, for example, are still siloed; you can’t share your information from one site with a contact on another site.”
But he remained both realistic and optimistic, the hallmarks of an entrepreneur: “The Web, after all, is just a tool…. What you see on it reflects humanity—or at least the 20% of humanity that currently has access to the Web… No one owns the World Wide Web, no one has a copyright for it, and no one collects royalties from it. It belongs to humanity, and when it comes to humanity, I’m tremendously optimistic.”
Originally published on Forbes.com
New Zealand flatworm
[January 8, 2019] Today is the first day of CES 2019 and artificial intelligence (AI) “will pervade the show,” says Gary Shapiro, chief executive of the Consumer Technology Association. One hundred and thirty years ago today (January 8, 1889), Herman Hollerith was granted a patent titled “Art of Compiling Statistics.” The patent described a punched card tabulating machine which heralded the fruitful marriage of statistics and computer engineering—called “machine learning” since the late 1950s, and reincarnated today as “deep learning,” or more popularly as “artificial intelligence.”
Commemorating IBM’s 100th anniversary in 2011, The Economist wrote:
In 1886, Herman Hollerith, a statistician, started a business to rent out the tabulating machines he had originally invented for America’s census. Taking a page from train conductors, who then punched holes in tickets to denote passengers’ observable traits (e.g., that they were tall, or female) to prevent fraud, he developed a punch card that held a person’s data and an electric contraption to read it. The technology became the core of IBM’s business when it was incorporated as Computing Tabulating Recording Company (CTR) in 1911 after Hollerith’s firm merged with three others.
In his patent application, Hollerith explained the usefulness of his machine in the context of a population survey and the statistical analysis of what we now call “big data”:
The returns of a census contain the names of individuals and various data relating to such persons, as age, sex, race, nativity, nativity of father, nativity of mother, occupation, civil condition, etc. These facts or data I will for convenience call statistical items, from which items the various statistical tables are compiled. In such compilation the person is the unit, and the statistics are compiled according to single items or combinations of items… it may be required to know the numbers of persons engaged in certain occupations, classified according to sex, groups of ages, and certain nativities. In such cases persons are counted according to combinations of items. A method for compiling such statistics must be capable of counting or adding units according to single statistical items or combinations of such items. The labor and expense of such tallies, especially when counting combinations of items made by the usual methods, are very great.
In Before the Computer, James Cortada describes the results of the first large-scale machine learning project:
The U.S. Census of 1890… was a milestone in the history of modern data processing…. No other occurrence so clearly symbolized the start of the age of mechanized data handling…. Before the end of that year, [Hollerith’s] machines had tabulated all 62,622,250 souls in the United States. Use of his machines saved the bureau $5 million over manual methods while cutting sharply the time to do the job. Additional analysis of other variables with his machines meant that the Census of 1890 could be completed within two years, as opposed to nearly ten years taken for fewer data variables and a smaller population in the previous census.
But the efficient output of the machine was considered by some as “fake news.” In 1891, the Electrical Engineer reported (quoted in Patricia Cline Cohen’s A Calculating People):
The statement by Mr. Porter [the head of the Census Bureau, announcing the initial count of the 1890 census] that the population of this great republic was only 62,622,250 sent into spasms of indignation a great many people who had made up their minds that the dignity of the republic could only be supported on a total of 75,000,000. Hence there was a howl, not of “deep-mouthed welcome,” but of frantic disappointment. And then the publication of the figures for New York! Rachel weeping for her lost children and refusing to be comforted was a mere puppet-show compared with some of our New York politicians over the strayed and stolen Manhattan Island citizens.
A century later, no matter how even more efficiently machines learned, they were still accused of creating and disseminating fake news. On March 24, 2011, the U.S. Census Bureau delivered “New York’s 2010 Census population totals, including first look at race and Hispanic origin data for legislative redistricting.” In response to the census data showing that New York has about 200,000 less people than originally thought, Senator Chuck Schumer said, “The Census Bureau has never known how to count urban populations and needs to go back to the drawing board. It strains credulity to believe that New York City has grown by only 167,000 people over the last decade.” Mayor Bloomberg called the numbers “totally incongruous” and Brooklyn borough president Marty Markowitz said “I know they made a big big mistake.” [The results of the 1990 census were also disappointing and were unsuccessfully challenged in court, according to the New York Times].
Complaints by politicians and other people have not slowed down the continuing advances in using computers in ingenious ways for increasingly sophisticated statistical analysis. In 1959, Arthur Samuel experimented with teaching computers how to beat humans in chess, calling his approach “machine learning.”
Later applied successfully to modern challenges such as spam filtering and fraud detection, the machine-learning approach relied on statistical procedures that found patterns in the data or classified the data into different buckets, allowing the computer to “learn” (e.g., optimize the performance—accuracy—of a certain task) and “predict” (e.g., classify or put in different buckets) the type of new data that is fed to it. Entrepreneurs such as Norman Nie (SPSS) and Jim Goodnight (SAS) accelerated the practical application of computational statistics by developing software programs that enabled the widespread use of machine learning and other sophisticated statistical analysis techniques.
In his 1959 paper, Samuel described machine learning as particularly suited for very specific tasks, in distinction to the “Neural-net approach,” which he thought could lead to the development of general-purpose leaning machines. The neural networks approach was inspired by a 1943 paper by Warren S. McCulloch and Walter Pitts in which they described networks of idealized and simplified artificial “neurons” and how they might perform simple logical functions, leading to the popular description of today’s neural networks as “mimicking the brain.”
Over the years, the popularity of “neural networks” have gone up and down a number of hype cycles, starting with the Perceptron, a 2-layer neural network that was considered by the US Navy to be “the embryo of an electronic computer that.. will be able to walk, talk, see, write, reproduce itself and be conscious of its existence.” In addition to failing to meet these lofty expectations—similar in tone to today’s perceived threat of “super-intelligence”—neural networks suffered from a fierce competition from the academics who coined the term “artificial intelligence” in 1955 and preferred the manipulation of symbols rather than computational statistics as a sure path to creating a human-like machine.
It didn’t work and “AI Winter” set in. With the invention and successful application of “backpropagation” as a way to overcome the limitations of simple neural networks, statistical analysis was again on the ascendance, now cleverly labeled as “deep learning.” In Neural Networks and Statistical Models (1994), Warren Sarle explained to his worried and confused fellow statisticians that the ominous-sounding artificial neural networks
are nothing more than nonlinear regression and discriminant models that can be implemented with standard statistical software… like many statistical methods, [artificial neural networks] are capable of processing vast amounts of data and making predictions that are sometimes surprisingly accurate; this does not make them “intelligent” in the usual sense of the word. Artificial neural networks “learn” in much the same way that many statistical algorithms do estimation, but usually much more slowly than statistical algorithms. If artificial neural networks are intelligent, then many statistical methods must also be considered intelligent.
Sarle provided his colleagues with a handy dictionary translating the terms used by “neural engineers” to the language of statisticians (e.g., “features” are “variables”). In anticipation of today’s “data science” and predictions of algorithms replacing statisticians (and even scientists), Sarle reassured them that no “black box” can substitute for human intelligence:
Neural engineers want their networks to be black boxes requiring no human intervention—data in, predictions out. The marketing hype claims that neural networks can be used with no experience and automatically learn whatever is required; this, of course, is nonsense. Doing a simple linear regression requires a nontrivial amount of statistical expertise.
In his April 2018 congressional testimony, Mark Zuckerberg agreed that relying blindly on black boxes is not a good idea: “I don’t think that in 10 or 20 years, in the future that we all want to build, we want to end up with systems that people don’t understand how they’re making decisions.” Still, Zuckerberg used the aura, the enigma, the mystery that masks inconvenient truths, everything that has been associated with the hyped marriage of computers and statistical analysis, to ensure the public that the future will be great: “Over the long term, building AI tools is going to be the scalable way to identify and root out most of this harmful content.”
Facebook’s top AI researcher Yann LeCun is “less optimistic, and a lot less certain about how long it would take to improve AI tools.” In his assessment, “Our best AI systems have less common sense than a house cat.” An accurate description of today’s not very intelligent machines, and reminiscent of what Samuel said in his 1959 machine learning paper:
Warren S. McCulloch has compared the digital computer to the nervous system of a flatworm. To extend this comparison to the situation under discussion would be unfair to the worm since its nervous system is actually quite highly organized as compared to [the most advanced artificial neural networks of the day].
Over the past sixty years, artificial intelligence has advanced from being not as smart as a flatworm to having less common sense than a house cat.
Originally published on Forbes.com
