Featured Video
This Week in Quality Digest Live
Innovation Features
Laurel Thoennes @ Quality Digest
One decision changes your entire reality
MIT News
Data scientists find that artificial data give the same results as real without compromising privacy
NIST
Understanding how RF platforms work in industrial settings
Dirk Dusharme @ Quality Digest
It’s not about jobs... believe me
Tom Scaletta
Timely, objective, patient-specific comments and data can help physicians enhance patient satisfaction

More Features

Innovation News
Brilliant inventions from the ‘American Idol for Nerds’
With 90% power savings, it could make speech recognition ubiquitous in electronics
Study finds law-breaking property that could lead to applications in thermoelectrics, window coatings
Three phases and challenges
By 2025, four levels of self-learning technology will be in play
Automaker’s decision marks reversal on plans to build new plant in Mexico
Mechanical engineer builds animal-like machines for use in disaster response

More News

Taran March @ Quality Digest

Innovation

AI and the Little Guy

Out of step and out of luck with Industry 4.0

Published: Thursday, February 16, 2017 - 13:03

Before my bosses get wind of the artificial intelligence (AI) platform Quill (“perfectly written, meaningful narratives indistinguishable from a human-written one”) and decide its 18-month ROI would be a great exchange for my pay and complaining, I’d like to present this human-centric survey of where things stand with AI and manufacturing. Call me old-fashioned, but I don’t consider it a “narrative.”

First off, this appears to be the year to invest in AI, the technology that enables machines and systems to imitate human perceptions and learning: “Why 2017 is the year to invest in artificial intelligence stocks” (The Motley Fool); “Top stocks to buy in artificial intelligence” (Fox Business); “Obama White House’s final tech recommendation: invest in AI” (Computerworld); “Artificial intelligence to drive China VC investments in 2017” (CNBC); “Ford to invest $1 billion in artificial intelligence startup” (NY Times).

These developments should not to be confused with AI in 2016, “The year artificial intelligence came of age” (the Guardian); or with AI in the breakthrough year 2015, “What robots and AI learned in 2015” (MIT Technology Review); or even with AI in 2014, when theoretical physicist Stephen Hawking said, “AI could spell the end of the human race” (BBC News). No, this is the year for elbowing out the competition and getting those deep-learning systems up and running.

Much of the flurry is coming from China, whose growing middle class, aging workforce, and improved infrastructure make AI a promising next step. The country’s central planners continue to promote factory automation, including dishing out financial incentives, and they have announced their intention to add China to the top 10 robotics nations, in terms of both producing and procuring, within the next three years. The five markets of China, South Korea, Japan, the United States, and Germany account for about 75 percent of global robot sales.


Automation looms in China’s manufacturing future.

Here in the United States, along with Ford’s billion-dollar bet on autonomous mobility, other high rollers are snapping up AI firms. Last year alone 40 private companies were bought out; Google leads the pack and now owns 11, with Intel and Apple tied for second place, and Twitter a surprising third. Though still out of reach for many companies, the technology is moving in the direction of affordability. The cost of industrial robots and software is expected to drop by 20 percent during the next 10 years, while their performance—both in agility and accuracy—will improve by 5 percent each year. Smarter... cheaper... it doesn’t take a self-driving car to see where this is going.

Factory automation continues to focus on real-time integration to improve brick-and-mortar investments, digitization, and processes. A study published late last year by Boston Consulting Group said that 67 percent of automotive executives expect that new technologies enabled by real-time integration will allow their teams to reach and exceed lean management and continuous improvement goals through 2030. Smart factories will soon be the new normal, since among their impressive benefits they can:
• Allow value chains to continually accelerate and improve
• Promote “multidirectional layouts, modular line setups, sustainable production, and the orchestration of smart and collaborative robotics...”
• Improve every area of production quality
• Allow flexibility in response to rapidly changing customer requirements

AI and robotics are also behind the trend to bring more manufacturing (not to be confused with jobs) back to the United States, part of a movement toward smart, local, drone-delivering factories.

“The U.S. is still the biggest consumer in the world, by a lot,” Charles Speelman, superintendent of the Tri-Rivers Career Center in Marion, Ohio, told AI Online. “Now with all the automation and robotics, the cost of this technology is the same anywhere in the world; thus, labor cost is not the highest cost. As a result, manufacturers will locate to where they have abundant shipping options and a trained workforce to operate, maintain, program, and install this new, highly technical automation equipment. That will be a huge advantage to the U.S.”

About that trained workforce

Those of us still bumbling along with unintelligent coffee makers and flesh-and-blood co-workers might feel a bit winded from these findings. The human habit of pondering change on the way to accepting it quite often now seems like an indulgence. Where technological progress is concerned, the “check” in the PDCA cycle has been squeezed aside to make room for an extra letter in the acronym: plan, do, check, act now.

A recent report by McKinsey Global Institute listed AI as one of the disruptive forces that will change life, business, and the global economy. Compared to the industrial revolution that began in the late 18th century, the report notes, the effects of Industry 4.0 are occurring 10 times faster, and at 300 times the scale—or roughly 3,000 times faster—than those of the industrial revolution. At that rate, only superheroes would have the speed and agility to rise up against the machines as the Luddites did, but what would be the sense? We can’t feign ignorance about their obvious improvement to, and relief from, repetitive factory tasks, whether on the assembly line or in a clerical cubicle.


Robots at a sorting task in the United States

Since the recession those jobs, mostly filled by high school-educated workers, have dwindled. There are nearly 1.5 million fewer clerical jobs now than there were before the recession; manufacturing jobs have also fallen by 1.5 million. “New middle class” jobs include healthcare work as X-ray technicians and phlebotomists, as well as computer-controlled manufacturing and some office occupations, like paralegals.

But despite some notable manufacturing apprentice programs, and a renewed, if scattered, interest in offering more technical and trade classes at high school and community college levels, labor experts say the U.S. educational system is failing to help young people acquire the skills needed for these. Technology’s pace and our educational response to it don’t match. As for the middle-aged among the middle class, their options come close to nonexistent. Those still working cling in place, staving off redundancy or retirement, whichever comes first; others, who fell out of the workforce during the recession, make do with temporary jobs, unemployment, or despair.

In the meantime, the great AI juggernaut rolls on.

“I worry about the analogies with the industrial revolution because the industrial revolution hurt a lot of people over a long period,” Kay Firth-Butterfield, an attorney, author, and public speaker on topics related to AI and ethics, told Enterprise Irregulars. “And yes, we came through it and we developed something better. But it looks as if this industrial revolution will be much faster, and we need to prepare not to hurt as many people very quickly.”

Also worried, or at least aware, are the creators, developers, and interested stakeholders of AI, who met quietly last month at the Asilomar conference center in California for a five-day conference titled “Beneficial AI 2017.”  During it they hashed out the opportunities and challenges related to AI’s future and possible steps to take to ensure that the technology brings more good than bad with its inevitable dominance.

One outcome was the “Asilomar AI Principles,” broad guidelines for tackling research and ethics issues. Among them were these lofty benchmarks: “AI systems should be designed and operated so as to be compatible with ideals of human dignity, rights, freedoms, and cultural diversity”; and, “Superintelligence should only be developed in the service of widely shared ethical ideals, and for the benefit of all humanity rather than one state or organization.”


Attendees ponder the future at Beneficial AI 2017.

There are a dozen or so video clips from the conference, but if you have time for only one, I recommend MIT economist Andrew McAfee’s 20-minute presentation, “A FAQ on tech, jobs, and wages.” Although he wryly characterizes it as “a quick drink from a firehose,” it explains much in a short time about what he terms the “hollowed-out middle class.”

Perhaps it’s best summarized by a quote from the French philosopher Voltaire, whom McAfee referenced: “Work spares us from three evils: boredom, vice, and need.” McAfee asserts that AI is currently capable of shielding us from only one of those:

In “It’s time for some messy, democratic discussions about the future of AI,” Jack Stilgoe and Andrew Maynard applaud the intentions of Beneficial AI 2017 but say they’re not enough, and that the conference particularly suffered from a lack of public input. “The new Asilomar principles are a starting point,” they say. “But within the broader context of a global society that is faced with living with the benefits and the perils of AI, they should be treated as hypotheses.... They now need to be democratically tested.”

Discuss

About The Author

Taran March @ Quality Digest’s picture

Taran March @ Quality Digest

Taran March is Quality Digest’s editorial director. A 30-year veteran of publishing, March has written and edited for newspapers, magazines, book publishers, and universities. When not plotting the course of QD with the team, she usually can be found clicking around the internet in search of news and clues to the human condition.

Comments

Mixed feelings

It’s amazing and terrifying at the same time. You look at the advancement and you are impressed, then the dominos start toppling. Let’s look at the self driving cars. If my car can get me from A to B and the steering wheel is a thing of the past…..why do I need a driver’s license? I’m not doing the driving so I am not going to pay for a piece of plastic with my photo on it! So now the only purpose to go to the DMV is for license plates and registration? This can be done in the mail, with 3D printers maybe you can soon go to a website and print the plates? Sounds awesome no more DMV!!! That’s a lot of jobs, your brain is thinking about the DMV. Why should a taxi cab service pay for a driver when the car is the employee? No more bus drivers, no more cab drivers, minimal if nonexistent DMV. If I were to take more time it could expand even more. It’s a slight comfort to see some people in the field are promoting caution, but I feel it will be just as effective as telling my children “No running in the house.”

 I will wrap this up with something that has stuck with me all these years. I grew up on what was a dairy farm many years before. What does this have to do with AI? I’m getting to it, in what remained of the milking control room where the milk was bottled. A dirty worn sign was placed above the many dusty controls and rusty breaker boxes that stated “Remember, you can always be replaced by a button.”

I always wondered if those button builders understood what they truly create.