Book Summary – What To Do When Machines Do Everything (How to Get Ahead in a World of AI, Algorithms, Bots, and Big Data)
Smart technologies like Siri, Nest and Alexa may already enhance your home and workplace. Many of your apps and software tools tap into self-learning AI systems. Their increasing sophistication is the power behind diagnosticians, vehicles, stock trading analyses, legal researchers that outperform human beings, and even game players such as Google’s AlphaGo.
“AI isn’t coming; it’s here. What this book attempts to do is show you there are things – many, many things – that you can do – must do – when machines do everything.”
These new machines – the “systems of intelligence” that enable this new industrial revolution – will transform businesses into super-informed “know-it-all” enterprises. Expect a rapid transition from “digital that’s fun,” like Twitter and Facebook, to “digital that matters.” Facebook, Amazon, Netflix and Google (the “FANG vendors”) represent the digital vanguard leading the parade before long-established corporations move in.
“In this new economy, we will witness an expansion of what is possible and move from machines that do to machines that appear to learn and think.”
“Stall,” “Boom” and “Churn”
On an individual and corporate level, you may feel stalled by relentless global competition, crashing start-ups, wealth inequality and loss of privacy. Today’s “stall zone” is an ongoing phase as the dot-com crash feeds into a historical pattern of great technological shifts. It heralds the “digital build-out,” a roughly 25-year leveling during which far-out ideas move to center stage.
A recent Oxford University study indicates that by 2025 the G7 nations will lose some 173 million jobs to automation. A more reasonable analytical consensus, however, suggests only a 12% job loss overall. However, new machine-facilitated jobs will offset the losses, emerging amid the churn that will result from job automation, enhancement and creation. Because manual labor and knowledge labor lend themselves to different kinds of automation, new tech will both destroy and create jobs. It tends to automate specific tasks – starting with the most routine and tedious chores – rather than entire jobs.
“Will some jobs be automated away by AI? Yes, of course. But far more will be enhanced, and in time, millions more new jobs will be discovered, driving future employment.”
In the First Industrial Revolution, 17th-century Luddites smashed mechanical looms. Loss of agricultural jobs in the Second Industrial Revolution and decreases in assembly-line work in the Third Industrial Revolution caused further upheaval. In the wake of these cataclysms, economist Carlota Perez sees a historical pattern of “bubble” wealth leveling out into what she calls “Golden Ages.” The current stall follows a large, unstable burst of computer and Internet innovation at the inflection point of a dramatic “S-curve” in GDP growth over time.
“Long-established companies are extraordinarily well-positioned for the digital build-out.”
Three main factors will drive the coming boom: “ubiquitech,” the presence of technology inside everything; the way current tech “stinks” compared to tech in 2030; and mass digitalization, wherein firms master the “Three M’s” of “raw materials, new machines” and “business models.”
Fears of chaotic joblessness recur with the emergence of each new kind of automation. But the advent of the new machine isn’t abrupt; AI creeps in slowly and accelerates after a tipping point. In the short-term, jobs change but don’t disappear. As rote elements of work evaporate, workers can turn to more rewarding activities. This enhances their effectiveness and develops their performance. The history of social progress shows repeatedly that “new abundance” generated by innovation increases the availability of worthwhile employment and boosts living standards by making quality goods and services affordable.
“Once you have automated, haloed and enhanced your company’s activities, the associated AI engines can be applied to innovation.”
“Three Subsets of AI”
YouTube and Uber run on comprehensible intelligence systems involving hardware, software, data, algorithms and human input. Their AI is in the “narrow” – or weak – category. AI labeled “general” – or strong – is equal to human-level, flexible intelligence. For now, it is a future vision like the “singularity” or some runaway “super AI” apocalypse. For now, for profitability, “a narrow view is best.”
“This is truly becoming an age of knowing it all. Yet, without the right business model to support your data-fueled new machines, you won’t get very far.”
Systems of intelligence differ from “systems of record, which use machine-learning, exponentially increasing processing power and vast troves of big data. These systems, with their complex logic processes and software ecosystems, still must engage with people, so their interfaces must frame interactions accessibly and perfectly. A service such as Netflix draws data points from the information it gathers about its users and its offerings, processes those data points in the background and works with them to generate a seamless user experience. Such systems need scalable smartness, openness (via application programming interfaces, or APIs), humans-in-the-mix, narrow AI and tailored suitability. New information systems are easy to build, and your enterprise can implement its own versions.
Data Beat Oil
Oil, coal, steel and electricity provided the power for previous industrial revolutions. Today’s prime raw material are data; their commodity value outstrips all past raw materials. The oil boom transformed society, as will the data boom. Data’s advantages over oil include the low cost of mining, inexpensive distribution, limitless supply, and its proprietary and exponential nature. Like oil, data flow best in an efficiently managed supply chain. Use business analytics tools and processes to extract and refine exploitable meaning from your information. Make any nonedible products “smart.” Adding instrumentation to your products is cheap and easy, and people will benefit. Large-scale, long-established companies tend to own more instrumentable assets than new software firms can offer. You may have that early advantage now, but if you dally, it will evaporate. Prioritize going smart; follow where the data lead.
“With sensors and instrumentation, it’s now possible to collect and analyze information about everything, to know everything about everything.”
The “AHEAD” Strategy
When Silicon Valley innovation disrupts your industry – and it will – can your products and services still turn a profit at a ruthlessly reduced “Google price”? To survive, firms must automate their processes. Determine how you can upgrade your business model to meet this challenge. Consider “hybrid” models, where firms move some processes to digital and keep others manual. The unwary are vulnerable to several pitfalls: taking the superficial route by “doing digital” instead of “being digital,” trying to copy the FANGs, going overboard with digital or denying its impact. To benefit from the new machine, use the AHEAD strategy:
- “Automate” by streamlining processes with AI and bots.
- “Halo” by instrumenting your products and resources to generate code.
- “Enhance” human endeavors with artificial intelligence.
- Promote “Abundance” by making your offerings cheap, plentiful and competitive.
- Pursue “Discovery” by applying AI to make your R&D more dynamic.
“Soon, sensors will be embedded within most if not all physical things in the world, and even within people.”
Creative Destruction and Automation
Automation already impinges on the way you bank, the way you book a vacation, and more. Increasingly, automation affects white collar work, contributing to a fall of up to 60% in operational costs. During times of automation-driven creative destruction, your firm may pay a high price for lagging. The sweeping change that upended journalism – bots now write news stories – already affects entire industries. Few outside those fields notice the impact.
You can automate back-office functions, like HR and administration, inconspicuously. Apply the “25%– 25% rule” to set priorities: Pursue a 25% cost reduction for a 25% productivity gain. Target specific processes for full automation. You may have to overcome resistance from the “brass wall,” experienced middle managers who are approaching retirement. Overcome their protests, and build at least one automated process that eliminates a manual process. Start low-key, but pursue clear aims. Prioritize removing a bottleneck. Prototype, test, analyze and repeat. Automation is the starting point, but your goal is revolutionized productivity.
“The key point is this: In thinking about digital solutions and artificial intelligence, we often focus on the impact of the technology on the world that we know.”
“Halos of Meaningful Data”
Digital tech enables you to place a “code halo” around anything to create its information-gathering, connected “digital twin.” You’ll note this trend in cars, the insurance industry, sports clothing, machinery, and more. Companies like Toyota, Boeing and Philips imbue their products with code-generating intelligence. Firms install “always-on” connectedness and, in the end, the code generated is more important than the product it halos. The “know-it-all” enterprise never stops gathering and analyzing data. For instance, GE Transport created a sensor-packed “intelligent locomotive” to help it transport goods more efficiently. You can add instrumentation to your fleet and your supply chain to glean back-office insight. You can gather data from your staff members. Source-data experts can turn such raw code into value. Let the code inspire your business model and give you the data to improve customer experiences.
“The ‘creative’ forces that machines unleash will be their real legacy.”
Boosting Human Potential with AI
Though accustomed to systems like GPS, many people hardly notice that they’re being personally enhanced by technology. In time, most people will choose tech-enhanced doctors, lawyers and teachers. AI assistants like Microsoft’s Cortana enhance your performance at work. Surgical systems like da Vinci enhance physicians. You can upgrade your firm’s capabilities to enhance its performance. Though Google’s AlphaGo program defeated Go champion Lee Sedol, he learned radical new moves from the AI. Automation enhances humans, and vice versa. Freed from humdrum tasks, humans think bigger and empathize more, which customers greatly appreciate.
Abundance occurs when prices drop significantly, thus boosting demand. Tech-driven efficiencies that squashed prices led to abundant cars, clothing, travel, refrigerators, and more. Systems of intelligence achieve such efficiencies in both the digital and physical domains, still reflecting the relentlessness of Moore’s law; now, the doubling computer processor power touches all industries. Keep a close eye on competing start-ups. Set up a focus group of young employees and ask which abundance tech product mightdestroy your firm. Embrace the “corner-shop” mentality of knowing your customers, and personalizing for them. Use the new machine to slash prices to breed abundance and prosperity.
“Look at your company’s most expensive, premium-level, differentiated products and services. Now imagine them at 10% of the market price.”
The Edwin Budding Effect
Your R&D discoveries gain a new dimension with AI’s help, but discovery still needs your participation. As an analogy, consider the manual lawn mower, invented by Edwin Budding in 1827.When these machines replaced the back-breaking scythe as a way to clear playing fields, it led to the clearing of more open spaces and to today’s $620 billion global sports industry. What “new lawn mower” will your firm create?
“The lesson from Apple is that making ‘things’ beautiful is not an esoteric exercise for those with nothing better to do; rather, it is the highest priority on your to-do list.”
To implement digital, apply an incremental kaizen (“change for better”) approach. Most R&D will miss the mark, but successes could more than compensate. Smart firms hedge their bets; for example, Toyota still builds conventional cars despite its investment in driverless vehicles.
In R&D, the new machine assists with “processes, frugality, sustainability,” and more. It iterates fast. Ray Kurzweil’s Law of Accelerating Returns describes the exponential rate of gains with leveraged AI. For example, blockchain tech may be a new, pervasive transactional paradigm.
“We are using the most powerful innovations since the introduction of alternating current to share cat videos, chat with Aunt Alice and hashtag political rants.”
Will AI threaten humankind with dystopia, as Stephen Hawking feared, or deify it into utopia, as Ray Kurzweil predicts? AI pragmatists move AHEAD with data as their raw material, systems of intelligence as their machines, and data-centric monetization as their business model.