4
The New Machine
Systems of Intelligence
Perhaps you sometimes wonder the same things we do: “How can Uber instantly connect me to a car when I'm on a random street corner some 500 miles from home and then automatically bill my credit card, send me an invoice, and capture my passenger rating in seconds?” “How can I be watching a YouTube video on a mobile device while moving at 130 mph on a train?”
These two experiences, moments of “magic” that have become almost mundane, would have been impossible even a few years ago. What's surprising is that Uber and YouTube, in spite of offering dissimilar services, both run their operations on a “machine” with essentially the same components. This new machine, what we call “a system of intelligence,” is rapidly becoming the cornerstone for companies that compete on knowledge. It sits at the center of Facebook, Instagram, Google, e*Trade, Betterment, and all the other examples of today's digital leaders.
Yet, for all its importance, the new machine is still largely misunderstood. Most of us actively consume the output of these systems of intelligence without slowing down to think about how such real-time, personalized, and curated experiences are actually created and delivered.
To that end, in this chapter we will explain what these new machines are—what the technology components are, how they fit together, what a good one looks like, what they mean to enterprise work, and how they will profoundly impact the future of your work.
We know; this overview may seem akin to when you were learning to drive as a teen and your uncle popped the hood on the car, and started to explain exactly how everything worked. Some of the lesson can be dry (e.g., “this is the carburetor; these are the spark plugs”), but as we now consume these systems of intelligence on a continual basis and need to build and implement them in our own companies to seize competitive advantage, a working knowledge is important.
Defining the New Machine
Let's start with a simple definition; then we'll unpack it a bit.
This may sound like “a bunch of hardware, software, and data is combined, and then a miracle occurs.” So let's touch on the three key attributes that make a system of intelligence so special:
- Software that learns: The software at the core of the new machine is like nothing we've seen before. For the first time in human history we have a tool that can make itself. With machine-learning software, these systems improve on their own over time. The system learns how to recognize patterns and how to find hidden insights in the data, all without being explicitly programmed on what to do or where to look. This is the way, for example, Uber knows how to match the right driver with the right passenger or how Facebook populates your personalized News Feed. There are few humans at these companies figuring these things out. That would be impossible as, in the case of Facebook for example, there are more than one billion daily customer logins. 1 Instead, the machine learns about each and every one of those sessions, continually getting smarter.
- Massive hardware processing power: During the past several decades, we've seen the power of hardware technology increase exponentially. No other innovation in history has both improved and diffused this rapidly. Moore's Law, which finds that the number of transistors on a chip (and therefore processing power) doubles approximately every two years, persists in spite of having recently celebrated its 50th birthday. But lately, this has been turbocharged by the cloud, which enables hyper-powerful computers to be tied together. By way of comparison, a muscle car may have impressive horsepower, such as the 435 horses under the hood of a Ford Mustang GT, but you cannot glue two Mustangs together to go twice the speed. By contrast, with cloud computing one can tap into multiple computers to get blazing-fast performance. Thus, each time you use Google, Facebook, or Amazon, you are tapping into their hardware farms of connected, super-fast servers.
- Huge amounts of data: Data is the fuel of the new economy. Relate this fact to the taxi dispatcher example cited previously. In the good old days of, say, 2012, maybe three points of “data” about your ride would be captured: a record of your phone call requesting the cab, the handwritten notes of the dispatcher and the driver, and your payment details. (Of course, rarely if ever would these manual records be reviewed or analyzed.) Compare that with a typical Uber ride, which will capture data on your request, location, time, route taken, device used, payment and tip, driver, passenger, driver rating, and passenger rating. And then multiply that by the more than two billion rides taken (through mid 2016) via Uber.
In short, these three special features—software that learns, massive processing power, and enormous amounts of data—are combining to bring systems of intelligence to life. (As an aside, in some circles, these are now being referred to as software “platforms”; for clarity and consistency, we will refer to them as systems of intelligence). Further along in the chapter, we will outline how these pieces all fit together. Before we get there, it's useful to provide some definitions on the most controversial and misunderstood part of the machine—artificial intelligence.
Artificial Intelligence: Why a Narrow View Is Best
The term “artificial intelligence” has become so overused that it's actually causing more confusion than clarity. There are many definitions in the market, and almost all of them emphasize a comparison to humans. Such definitions—for example, Merriam-Webster's (“the capability of a machine to imitate intelligent human behavior”)—immediately send many of us down the wrong path, for we start thinking, “What human intelligence can and will be imitated?” We think this is wrong.
Our definition is simpler:
It's that clear-cut. The anthropomorphically biased definitions of AI are wrong for two basic reasons:
- AI that creates business results will focus on what machines do really well, rather than trying to replicate what humans already do well.
- Humans have long ago proven themselves to be imperfect “machines” (just watch the six o'clock news). It seems a bit narcissistic that the design-point of the new machine should be that of a human.
Thus, AI is not about building robots that ape human form and behavior. Instead, practically applied AI represents the next generation of computer systems that, like the systems of old, are housed in air-conditioned computer rooms and accessed through networks and systems (like those apps on your smartphone) that you may not see but consume on a regular basis.
But such a definition is only the beginning. We've found it very helpful in slicing through the definitional clutter to divide AI into three subsets: 2
- Narrow AI
- General AI
- Super AI
Narrow AI, which is also referred to as “applied AI” or “weak AI,” is our default definition for this book. It is important to note that all AI today—and for at least the next decade—is narrow (also termed “artificial narrow intelligence” or ANI). Such AI is purpose-built and business-focused on a specific task (e.g., driving a car, reviewing an X-ray, or tracking financial trades for fraud) within the “narrow” context of a product, service, or business process. It's what the FANG vendors utilize today in delivering their digital experiences. Thus, while it appears that the new machines can do everything, they actually focus on doing just one particular thing very well. As such, these ANI systems would be hopeless in any pursuits beyond those for which they were specifically designed (just ask your Waze GPS if that onion bagel with cream cheese fits your current diet). ANI is simply a tool, albeit a very powerful one, that provides the basis for all we will explore in the coming pages.
General AI, also referred to as “strong AI” or AGI, is what is fueling the fears of the Singularity crowd, and has been highlighted in the previously referenced films Her and Ex Machina. 3 Strong AI is the pursuit of a machine that has the same general intelligence as a human. For example, just as you in a span of a few moments can discuss politics, tell a joke, and then hit a golf ball 150 yards, the AGI computer will have the general intelligence to perform these activities as well.
Ben Goertzel, Chairman of the Artificial General Intelligence Society, points to the Coffee Test as a good definition for AGI. That is, “Go into an average American house and figure out how to make coffee, including identifying the coffee machine, figuring out what the buttons do, finding the coffee in the cabinet, etc.” 4 This set of tasks that is seemingly easy for almost any adult to perform is currently insanely difficult for a computer. Creating AGI is a dramatically harder task than creating ANI; by most estimates we are still more than two decades away from developing such AI capabilities, if ever.
That said, it's easy to scare ourselves with general AI for two reasons, one practical and the other theoretical. On the practical front, we see examples today of narrow AI that appear to be general AI. This might be your Amazon Alexa home appliance, which passes the Turing test (of acting in a way that's indistinguishable from a human). It may feel as though it's moving toward AGI, but it's actually just a brilliantly elegant vocal interface to the Internet search capacity we've known for over 15 years.
On the theoretical front, computer scientists look at the human as a machine in and of itself—one with very distinct limitations. Human IQs generally range from 80 to 150, an incredibly small scope in computing terms. If AGI becomes a possibility from a software standpoint, why would we limit a machine's “intelligence” to, say, an IQ of 150? Why not 300 or 3,000 or 30,000? None of us would even begin to understand what such an IQ would be or could do, but when it's a simple matter of stringing together more servers in the cloud to add more processing power, where would this take us?
This leads to the third definition. Super AI , in essence, is the technical genie being let out of the bottle. In such a scenario, would humans even know how to stop such a machine? It would run circles around our collective intellect (and, as we know, whenever 10 reasonably smart people are put in a room, the collective IQ is not 1,200 but actually somewhere around 95 once one accounts for the different opinions and objectives that people always bring). How could we then turn the machine off when it's always 10 (or 1,000) steps ahead of us?
This is interesting stuff, particularly for cocktail party conversations. Yet, to reiterate, based on our research, a post-Singularity future with super AI-based Terminators running amok is a mirage. Serious people—the people building these systems today—are quite modest about whether such scenarios are even possible in the next 100 years, much less in the next 5 to 10. Andrew Ng, the chief scientist at Baidu Research, put it succinctly when he said that “worrying about [general or super] AI is like worrying about overpopulation on Mars before we've even set foot on it.” 5
Thus, our focus in this book is specifically on ANI because here, in the real world, we are more concerned about effective tool use for good business outcomes in the modern enterprise. Although some will continue fretting away, worrying about such things as super AI, their competitors will put them out of business with practical applications of narrow AI. With that as a definition of AI, let's go deeper into the new machine.
Meet the Machine: Anatomy of a System of Intelligence
Each system of intelligence can do vastly different things, but they all share a similar basic anatomy. In fact, if you are familiar with enterprise technology and the prior generation of systems of record (such as ERP and CRM systems), many of the constituent parts will seem familiar. After all, the technology “stacks” of systems of record and systems of intelligence share many common elements, such as user interfaces, application logic, process flows, databases, and infrastructure.
However, critical differences exist up and down each layer of the stack, the most important focusing on the three distinguishing characteristics of the new machine highlighted earlier: systems that learn, massive processing power, and huge amounts of data. In Figure 4.1 we highlight several of the key differences at each layer of the software stack. Next, we'll work top-down through the various components
Figure 4.1 Systems of Record vs. Systems of Intelligence
Figure 4.2 Anatomy of a System of Intelligence
common to every system of intelligence as shown in Figure 4.2 .
Users, Customers, and Employees
Regardless of how digitized our world becomes, this shift is primarily oriented around people; that is, us—carbon-based life forms without an on/off switch. Thus, with successful digital solutions, the more technical they are, the more human they feel. The best digital solutions don't slow us down; instead, they get out of the way, seamlessly helping us to achieve our end goals. We don't want to learn the systems, we just want the results. With the Waze app, it's how to get somewhere by the fastest possible route; with GE's Predix platform, it's the status of our windmill farm; with the Lex Machina legal analytics service, it's the track record of a particular judge. The machines may be able to do amazing things, but the trait we've seen successful systems share is that they put the human experience at the center of the design. Oftentimes, even in the most automated, seemingly AI-rich environments, there is still a considerable amount of human intervention involved.
The App Interface
In our personal lives, we think of Netflix, Strava, LinkedIn, and so on as “apps.” Most of us don't even call them “software” anymore. What you touch, the content you share, the information or insight you get—all of this flows through the app (or the application layer), which is the front door to the rest of the new machine. For years, you've downloaded apps onto your PC, your smartphone, or your tablet, and now they are being embedded into industrial machines (such as cars). What's essential here is that it's the app that frames our experience. The rest of the system of intelligence, if it's any good, is invisible to us.
Additionally, as we outlined in our book Code Halos, the app has to provide a beautiful experience for users. The app has to hit the FANG benchmark for elegance and ease of use. This explains the explosion of “design thinking” in corporate IT circles, given that these interfaces must fit the way in which your customers, partners, and employees wish to use them in the midst of their activities. (No owner's manuals or user training allowed!) The key to success here is that the app interface must be simple and intuitive, seamlessly fitting into the context of the user's needs.
The AI “Innards”
Beyond all the hype, ANI is modern, complex, adaptive software that is the essential core of a system of intelligence. What we think of as AI really comprises three main elements:
- Digital process logic: Whenever we log onto a system of intelligence, we are engaging in a process: booking a car, processing an insurance claim, making a financial trade, or checking the status of an MRI machine. There's nothing terribly new here. An IT consultant from 1990 would recognize the technical portions managing the process flows in a system of intelligence. What is revolutionary is that systems of intelligence transform many manual processes into automated ones. Think again of the Uber versus taxi example. The car-dispatch process for a traditional taxi company is manual (e.g., customer calls in, agent informs dispatcher, dispatcher radios the appropriate cabbie, etc.). At Uber, that entire chain is automated. When that digitized process is multiplied over millions of transactions, an industry is revolutionized. We outline how to do this in the second half of the book. But for now, the key point is that while the technology that runs the process layer of a system of intelligence is rather simple, getting the underlying process structured properly is very hard work.
- Machine intelligence: This is what's genuinely new and different about the technology of the new machine. Through a combination of algorithms, automation processes, machine learning, and neural networks, the system mimics “learning” through experience, which is really just a richer data set. This is how it can automate work processes (e.g., reading an X-ray), instruct an employee on the next best action (e.g., a salesperson knowing the precise pricing that will most likely close the next deal), and recognize market trends that will lead to the next breakthrough product. The internal “clockworks” (the software mechanism inside the system of intelligence) is machine intelligence, the true heart of AI. When viewed this way, it's not scary, arcane, or hopelessly complex. Don't get us wrong; we're not trivializing the technical complexity of creating these works. But it's not anything mystical either. For all the talk of a “ghost in the machine,” 6 in the real world, that's merely applied ANI. It's nothing more and certainly nothing less, because in the corpus of the new machine, AI is the heart.
- Software ecosystem: Our apparently magical experiences with systems of intelligence may appear seamless, but we are never interacting with one piece of software. These systems are usually made up of an ecosystem of dozens of other tools connected by application programming interfaces (APIs) that are bits of software that link one tool to another like Lego blocks. Uber, for example, leverages a rich fabric of tools, including Twilio for cloud-based communications, Google for maps, Braintree for payments, SendGrid for e-mail, and so on. We each have an elegant personal experience thanks to the Uber software interface, but we are actually engaging with an entire ecosystem of tools and services.
This is a business book for a technical age. We're not going to go into the ins and outs of the specific technologies that comprise today's artificial intelligence. We could write an entire second book on machine learning, deep learning, and neural networks (these are the hot topics in leading universities around the world), but candidly that would be redundant because there are some terrific learning resources in the market already.
Data from Things and Systems of Record
The inputs to the new machine are many and varied. Some will be mature, ERP-based systems; some will be real-time data pouring in from instrumentation—the Code Halos of data surrounding products, people, and places—that is continually informing this nerve center as to what is going on all around it. Over time, these inputs will evolve, sometimes both quickly and radically. All those inputs are responsible for creating that contextualized, highly valuable data. Without those new data sources, it will be difficult to build or fuel your new machines. Sensors in mobile devices, apparel, sporting equipment, cars, roads, and virtually every other physical entity will be responsible for generating the information—the code. Connecting those sensors to a system of intelligence is the Internet of Things coming to life.
Data may seem like a weightless abstraction. (If you're interested, all the electrons in the Internet together apparently weigh about as much as two strawberries.) 7 So while it doesn't weigh much, data has a huge amount of value when it's deployed in the right place at the right time. Massive amounts of data must be captured, stored, maintained, analyzed, and made accessible. That's why we need large database systems that that are stable, scalable, and tested (no matter how cool the shiny new tools may seem). A new generation of databases (e.g., Hadoop) are finding favor, but Oracle and SAP, with a combined customer base of more than 740,000 companies, are not going anywhere. Neither are IBM, Microsoft, or any of the other major enterprise software “arms dealers.” In the digital economy, we'll still need high-quality systems of record (both traditional and emerging ones) as much as we need our AC power grid.
Infrastructure
As with any machine on a factory shop floor, the new machine needs power and “plumbing.” Infrastructure includes all the networking, servers, power sources, and so forth that make the machine hum. For existing systems, many of the elements are already managed either directly by the IT department, by an external service provider, or—more common now—by a cloud provider. Mobile networking, generally running on a major carrier, is imperative. For computing power, regardless of whether your system is running on Amazon's servers, somewhere in the Googleplex, or in your own data center, systems of intelligence all need highly efficient, always-on plumbing.
Systems of Intelligence in Action
The “anatomy” of the new machine can seem somewhat abstract unless the parts are linked together into actual new machines visible in the real world that we conceptually understand. Much has already been said about Netflix, but our take is different. Because we are all familiar with the media-streaming platform, it provides a great illustration of how a company that is leveraging the new machine, the new raw materials, and a business model oriented around them is upending business as usual.
The Machine That Is Netflix (Thanks to AI)
As of 2016, Netflix represented about 35% of all Internet traffic in North America and has the incumbent TV networks in quite a spin. 8 By dissecting Netflix a bit, we can see the anatomy of the new machine in action. (See Figure 4.3 .)
Figure 4.3 The Anatomy of the Netflix System of Intelligence
| Anatomy Element | How It Works (continued ) |
| Users/Customers | This is all of us as consumers. Netflix now has about 75 million subscribers worldwide (and growing), and we all want what we want (and we want it now, on every device, always). 14 |
| The App | Most of us experience Netflix as an app. Regardless of where we are or the hardware we use—tablet, mobile device, laptop, set-top box, or VR headset—the only part of the Netflix system of engagement we touch is the app that connects us to Orange Is the New Black (and many more). |
| Process Logic | Do you remember using the print TV Guide to find out what was on television and then having to time your life around when a show was airing? Um, neither do we. In those bygone days, there was really only one process for consuming media via television or movie screen, and it wasn't great. Now, of course, whether we are watching on a mobile device on the road or next to our loved ones at home, Netflix provides content in an almost infinite set of variations. It seems so simple now. Of course we can watch House of Cards at any time on a train going 130 mph! But it feels seamless and frictionless because Netflix has thoughtfully adapted its entire system to the way we want to consume content. |
| Machine Learning | Netflix's recommendation engine—basically, a set of algorithms that connect us to content we want—is the most well-known element of the Netflix AI system, but the whole engine is actually much more than that. 15 The core of the Netflix experience connects us to content, distinguishes between family members, and processes billions of events a day related to movies, viewers, payments, and the like. More important, the platform improves over time. As we use it more, it learns about our tastes and serves up the best content available in a highly personalized way. It can distinguish between what we say we like and what we actually like. (Note, for example, that Adam Sandler's The Ridiculous Six is rated a middling three stars by Netflix users, but in January 2016 was the most watched Netflix program in history.) 16 The core of the Netflix system is a remarkable piece of software design and engineering, which, because it is so good, is nearly invisible. |
| Software Ecosystem | Netflix relies on connections to dozens of other systems to bring us Orphan Black and The Walking Dead (two of the most binge-watched shows). 17 The Netflix team emphasizes open-source software tools such as Java, MySQL, Hadoop, and others. Content distribution is supported by tools like Akamai, Limelight, and Level 3 Communications. It also relies heavily on Amazon's cloud systems for storage. 18 The point here is not the specific tools Netflix uses but that the company realized early on that it would need to build certain elements of its overall system and that it also could leverage other best-in-class systems to grow faster. |
| Sensors/Internet of Things | Netflix engages with us—both providing content and collecting data—via our devices, but it also learns about us via the many sensors and data associated with these devices. Other sensors are starting to matter to Netflix. For example, it is exploring ways to connect to Fitbits and even socks to monitor if we've fallen asleep. 19 |
| Data | The Netflix data warehouse stores about 10 petabytes of information. 20 (One petabyte is equivalent to 13.3 years of streaming HD-TV video. 21 ) With all that data flowing through the AI engine, Netflix knows you. We're sorry if that creeps you out, but it tracks the movies we watch, our searches, ratings, when we watch, where we watch, what devices we use, and more. 22 In addition to machine data, Netflix algorithms churn through massive amounts of movie data that is derived from large groups of trained movie taggers. (Netflix isn't saying, but the best guess is that it is applying more than 76,000 genres to categorize movies and TV shows. 23 ) |
| Systems of Record | There's no way to manage the vast reams of Netflix data without an absolutely top-notch architecture. Netflix started with Oracle, but has moved over to an open-source database called Cassandra. It uses Hadoop for data processing and Amazon S3 for storage. 24 It also links to back-end payment systems so nothing can interfere with our next purchase. |
| Infrastructure | Amazon inside! As of February 2016, all of the infrastructure services Netflix requires are provided by Amazon Web Services. 25 (Anyone still waiting for the cloud to “mature” is probably tripping over the stretchy cord on his or her phone at this point.) |
What Does “Good” Look Like? Attributes of a Successful System of Intelligence
There's a big difference between merely having all the necessary ingredients of the new machines and actually getting them to perform at a high level. A system of intelligence that can help you be the Usain Bolt of whatever race you're in will have all or most of these characteristics:
- Smart, not dumb: Effective new machines get better—more powerful and valuable—as they scale. One true test of good AI is that as data flows in, it's smarter tomorrow than today. In every case, the best systems of intelligence suck up data from a wide variety of sources, which helps achieve the data mass required to derive insights and create personalized experiences. Today's digital masters recognize that ultimately, “it's all about the data.”
- Open, not closed: Systems of intelligence that achieve their full potential are generally more open than closed. Think of Tesla giving away its patents and Uber with its open APIs; both of these policies have helped generate new solutions built on the companies' AI cogs and gears. One of the most striking current examples of this is the competition between Amazon and Walmart.com . In many respects, using the system anatomy we have outlined in this chapter, the two look quite similar. And the customer experience through their respective apps is, give or take 10%, pretty much the same. Yet viewed through the lens of APIs, the companies couldn't look more different. Amazon has more than 325 API mash-ups that open up its platform to others. Walmart, Target, Macy's, and Sears all had three or fewer as of September 2015. 9 Openness is necessary for digital success, because a fully fleshed-out ecosystem will be more powerful than a closed development environment.
- Smart hands, not just bots: Most successful ANI systems include people. One of the most common mistakes that we've observed on our travels occurs when companies try to completely dispense with the human element. 10 There are lots of things that machines simply cannot do as well as we can (at least within the timeframe of a meaningful business decision). Game-changing systems of intelligence are built around the integration of AI with humans by combining the best of what computers do with the best of what humans do. (More about AI “enhancement” is provided in Chapter 9 .) Some AI experiments, such as the robot hotel concierge, which attempt to mimic what humans do uniquely well, are a surefire way to waste time and money.
- Narrow, not broad: Trying to create one magic black box to solve all of the challenges you want AI to address might be enticing, but it's incredibly difficult to pull off. Even complex platforms like IBM's Watson or GE's Predix offer significant business value only when they are configured for a specific process or customer experience. 11 Focusing on a “moment” or a simple activity is one of the best indicators that a system of intelligence might have meaningful impact.
- One-to-one, not one-to-many: If you have ever worked with PeopleSoft or in an SAP environment, you know first hand that your user experience there was not designed for you personally. It was designed for the largest scale possible. This has been part of the architecture of almost every commercial software system since the punch card. While most existing enterprise systems are generally not designed to give us an individualized experience, systems of intelligence most definitely are. Alexa, for example, self-configures to your patterns. Amazon tunes its offerings to your virtual self. Old-school software requires you to conform to it; new-machine-age software conforms to you.
- Bespoke, not “off the rack”: It would be nice if we could access an iTunes store offering a range of systems of intelligence and pick up solutions there for our own organizations, but that capability doesn't exist yet. Over time, AI will become “productized”—and there are many machine-learning solutions that can be accessed on an “as-a-service” basis—but for the near future, you shouldn't imagine that your organization can simply buy a complete piece of software that will instantly get you into the heart of the new machine age. New systems of intelligence are emerging every day, so things are changing fast. This is the biggest area for innovation across virtually every industry. If you are looking for platforms aligned to claims processing, finance and accounting processes, or secondary education classes, you are in luck, because powerful systems already exist. However, in many if not most cases, there is not yet an existing product. You should keep a sharp eye out for new solutions that will matter to your business, but for now, many of these systems must be created or assembled from existing components.
From Vapor to Value
In looking in more detail at systems of intelligence, many profound, strategic, even existential questions are raised. What is a department store in an age of Amazon? What is a hotel in an age of Airbnb? What is car insurance in an age of self-driving cars?
Given just how large these questions are, it's tempting to think what we are calling a major trend is actually just a fad or the latest consulting theory. We hear it all the time: Maybe this is tomorrow's problem. Maybe we should wait until regulation or a disruptive event forces change on our industry.
We disagree. Every day that passes gives us more evidence and strengthens our conviction that the new machines that we've examined in this chapter are the engines of a Fourth Industrial Revolution. We are well past the theoretical phase. Individuals and companies that are realizing an early advantage are not all geniuses or legendary entrepreneurs. They are people like you, people who are using new technologies to solve major problems.
Remember, building your own new machine is becoming increasingly easy, even though, as we have pointed out, “some assembly is still required.” Increasingly you can take your credit card out and lease access to machine-learning code, infrastructure, and databases. The Google Cloud Platform gives you immediate access to a neural-net machine-learning platform. 12 Amazon Machine Learning gives you access to the same predictive analytics platform that provides recommendations to deep-pocket consultants and enterprises. 13 Only a few years ago, this would have cost companies millions and have taken many months to put into place.
The actual builders of the new machines all say the same thing: Making narrow AI the core of a system of intelligence is not a theoretical exercise; it's possible, and it's happening today. And what's really supercharging this explosion of activity is that the fuel for the new machines is all around us if you can see it, grab it, and use it, which is the focus of the next chapter.