Monthly Archives: March 2017

It's no secret that during the second decade of the 21st century, governments are loathe to spend a cent more than they have to, and Centrelink is shaping up to be the touchstone for using automation as its salvation and failing badly at it.

One can easily imagine how the powers that be within Centrelink and its overarching Department of Human Services (DHS) ended up taking the decision it did.

According to Community and Public Sector Union (CPSU) National Secretary Nadine Flood, after years on the receiving end of efficiency dividends -- government-speak for reducing spending by a single digit percentage and expecting the same level of output and service -- DHS suffered a 10 percent cut in 18 months.

"It is not an exaggeration to say that the Department of Human Services is an agency is crisis, and it's not something I say lightly," Flood told the Senate Community Affairs References Committee on Wednesday.

Flood said DHS is unable to provide Australians with a basic level of service following a reduction of 5,000 permanent roles by governments of both stripes.

Given such a situation, it is hardly surprising that management decided to automate a decades-old process. But there was a catch. The process itself was not sound, as thousands of former Centrelink recipients found out over the Christmas break.

In the pantheon of decision-making, automating an already bad process is up there with drinking two pots of coffee back to back: It will allow you to do stupid things at a much faster rate.

"The department has been put in a position where it has made decisions, with the recent introduction of the automated debt recovery program, to remove or reduce the role of DHS staff in that crucial hands-on element of the work -- investigating suspected overpayments and advising on appropriate debt recovery actions," Flood said.

"This new approach which removes or reduces human oversight of suspected overpayments and reduces employees' roles at a range of elements of the system has been an absolute disaster for many Centrelink users, but also for the workers charged with implementing a system they know to be deeply flawed and unfair."

Copping some flack for its perceived involvement in the data-matching, the Australian Tax Office was at pains to distance itself from DHS, and said it merely provided annual payment summaries to DHS, as it had done for years. If there was any division by 26 in this process to miscalculate fortnight income and generate debt notices, the ATO was not the source of it.

With an environment focused on saving money, and a budget target of collecting AU$1.2 billion from former welfare recipients, it is disturbing but not surprising that DHS took its human process and ported it across to a machine.

"If we want to look at where robodebt has come from, it is a fairly obvious consequence of a department that no longer has the resources to provide effective services," Flood said.

"It has, of course, proven to be a classic false economy -- and has created costly reverse workflows where staff are taken offline to deal with complex and difficult disputes over incorrectly raised automated debts.

"Sadly, I would suggest in the last few years, one of the things DHS has become an expert in is band-aid solutions as it lurched from one crisis to the next -- this is simply the largest of those."

In its defence, DHS told the committee a lot of the trouble was caused by people not engaging with the notices they were sent.

"I think what we underestimated was how many people would not clarify, and would not engage," DHS Secretary Kathryn Campbell said.

"If I was to sum up what the problem has been, it is that when we wrote those initial letters, that recipients and former recipients didn't engage."

36 million unanswered calls would suggest that when Australians engaged, DHS was wholly unable to cope with what it had unleashed.

Automation has been far from Centrelink's saviour; in fact, it has been a very naughty boy.

Read this article:

Centrelink robodebt resulted in automation false economy: CPSU - ZDNet

The current job description of an accounts payable clerk will disappear in possibly as little as 20 years. This may seem bleak, but the reality is that software advances, developments in robotics, AI and machine learning are bringing a new age of automation one in which machines will be able to outperform humans in various work tasks.

According to McKinsey Global institutes January 2017 report on the future of automation, nearly half of the activities that people are paid to do in the global economy can be automated by adapting currently demonstrated technology. Activities most susceptible to this automation are repetitive, non-creative tasks such as data collection and processing. This puts at risk many jobs in customer service, sales, invoicing, account management and other data entry positions, not the least of which includes AP clerks.

However, these projections dont necessarily mean that the future is hopeless for those holding AP positions. In McKinseys words, People will need to continue working alongside machines to produce the growth in per capita GDP to which countries around the world aspire.

Skilled employees will work alongside software automation and RPA (robotic process automation) to approve data analyzation, guide software in the right direction and even perform tasks that we may not know exist yet. This will require some new skills-based learning, but it is also an opportunity for AP department employees to step out from behind the curtain, develop their job descriptions and have more interesting and meaningful jobs. Employees will be able to focus on raising their profile, supporting the business with more meaningful work, providing good internal service, and in turn, be more motivated.

Reckon this is wishful thinking? Think again. Its been done before.

After all, the first computers wore skirts. In the early decades of the 1900s, mathematical and technical calculations were made manually rather than by machine. This work required a large workforce to compute all the information. With the industrial boom brought on by WWII, organizations like NASA began recruiting women for this work, who they called computers. It has even been said that the first computers wore skirts.

Eventually, as the machines we know today as computers began to develop, many of these manual tasks were automated. Rather than discarding the women that had previously done this job, NASA and other organizations simply retrained employees to work alongside these machines and perform less menial tasks. This conscious step allowed the women who had been the quiet backbone of the organization to make themselves and their work known.

One example recently made popular by the book and award-winning film Hidden Figures is that of African-American physicist and mathematician Katherine Johnson and her team. Johnson worked as a computer on NASAs early team from 1953-1958, where she analyzed topics such as gust alleviation for aircrafts. When NASA used electronic computers for the first time to calculate John Glenns first orbit around the earth, officials asked Johnson to verify the computers numbers and her reputation for accuracy helped establish confidence in the new technology. Johnson herself went on to use these new computers to aid in calculations until her retirement in 1986. Similarly, the value of AP clerks and other accounting professionals will shift as they become valuable as human analysts and strategists, vital in the role of validating a machines processes.

These kinds of shifts can be seen throughout history, like in the move away from agriculture and decreases in manufacturing share of employment in the United States, both of which were accompanied by the creation of new types of work not foreseen at the time.

We can expect a similar response to automation in the accounts payable department. As AP software becomes more advanced, clerks and controllers will evolve to work with it, not be replaced by it. The important work of AP clerks will no longer be in the shadows. The job will be transformed from paper pusher to vital business asset.

See original here:

Voices The hidden figures behind automation - Accounting Today

The Age of Automation

Much has been said about how automation will affect employment and the economy. In almost every conversation, the looming threat of job displacement is focused on a very specific sector: the blue-collar job market.

One frequently cited study published back in 2013 by Oxford University and the Oxford Martin School says that 47 percent of jobs in the US will be automated in the next 20 years. In Canada, a study conducted by the Brookfield Institute for Innovation + Entrepreneurship says that 40 percent of jobs in the country will be taken over by machines in the next decade or two. In the UK, theyre predicting that 850,000 jobs will be automated by 2030. And in Southeast Asia, an estimated 137 million workers are in danger of losing their jobs in the next 20 years.

These predictions are premised on the fact that machines are now more than capable of completing repetitive jobs that most blue-collar human workers are handling today. But technology isnt going to stop there. Artificial intelligence(AI)is getting more sophisticated, implying that its not only the jobs defined by formulaic processes that are in danger, but also creative, service and knowledge-based professions.

We are starting to see in fields like medicine, law, investment banking, dramatic increases in the ability of computers to think as well or better than humans. And thats really the game-changer here. Because thats something that we have never seen before, says Sunil Johal, a public policy expert forCBC News.

Granted, the implications of more intelligent automation on white collar jobs are all speculative at this point. Theres little data to support how much automation will affect that job market, mostly because experts believe its impactwill be far more subtle than inblue- collar industries. In white-collar industries, theres more opportunity to shuffle employees around, or slowly phase out jobs, which means the threat of automation wont be as dramatic. That being said, it willchange things.

Johal believes that to keep up, one must actively develop new skills that will adapt to the changing needs of the job market.

If Canada doesnt take this seriously, we are going to see many Canadians left on the sidelines of the labour market, he adds. They are not going to be able to get back into the job force.

View post:

The Automation Upheaval Won't Be Limited to Blue-Collar Jobs - Futurism

"We cannot solve our problems with the same kind of thinking that we used when we created them," Albert Einstein supposedly said. One of his namesakes, an artificial intelligence called Einstein that Salesforce is incorporating into all of its products, embodies that sentiment: Salesforce is betting that human cognition won't drive its next wave of commercial growth. Rather, machine learning will push Salesforce's products deeper into its clients' businesses, and help Salesforce penetrate new companies, by augmenting human decision-making.

If Einstein is anywhere near as useful as Salesforce claims, the technology will supplant some human workers -- maybe a lot of them. Salesforce wants to make sales and marketing more efficient, which means that fewer people will be needed accomplish the same tasks. CEO Marc Benioff once wrote, "The only constant in the technology industry is change." Automation has hit factory workers hard, and soon members of the information economy will feel the same pain. The deadline may arrive before most knowledge workers, or the societies they occupy, are prepared.

On Tuesday, Salesforce held a "customer kickoff" event. It executives and partners discussed product development, with a heavy focus on Einstein's artificial intelligence capabilities. IBM CEO Ginni Rometty spoke briefly, apropos the two companies' recently announced partnership. "2017 is the year that AI enters the world at scale," she said. "The cognitive era is just beginning."

Of course, humans have engaged in cognition as long as the species has existed. What Rometty meant is we won't keep our monopoly on thought for long.

AI has been subject to enough hype cycles that default skepticism is warranted. However, recent technological breakthroughs suggest that the frothy press coverage and equally frothy corporate salivation may indicate something real this time. Google's DeepMind division programmed an AI that beat a complex strategy game before anyone expected it to be able to. Andrew Ng, renowned machine learning expert and chief scientist at Baidu, told the Wall Street Journal, "I think we're in the phase where AI will change pretty much every major industry."

Ng added, "Things may change in the future, but one rule of thumb today is that almost anything that a typical person can do with less than one second of mental thought we can either now or in the very near future automate with AI." Perhaps human workers should be frightened, since "there are a lot of jobs that can be accomplished by stringing together many one-second tasks."

Salesforce's Einstein is already doing some of this. In practical terms, the AI will not provoke an immediate revolution -- it's a version of what many SaaS clients expect from the applications they pay for. You could even argue that innovation hides in plain sight. Einstein enables prosaic but immensely useful functions like automated lead-scoring, based on information pulled into Salesforce's system. Einstein can aggregate signals such as whether a contact has looked at marketing materials (e.g. a webinar or whitepaper), and what their role is within the target organization.

"In a world powered by AI, signals are important," chief product officer Alex Dayon noted at the customer kickoff event. Einstein is able to interpret a variety of data inputs without much setup. In fact, most of Einstein's capabilities are available out of the box. Other features like Einstein Vision have to be integrated by developers. Salesforce is bringing automatic customization to all of its customers, while enabling those with greater resources to craft a deeper layer of specialized processes.

Some levels of customization are designed to be leveraged by non-technical users. "I can create a lead-management process with clicks, not code," product marketing EVP Stephanie Buscemi said. Salespeople who use Einstein are "not just smarter, they're also more productive," according to Buscemi.

Today's productivity gains are tomorrow's layoffs. It's not that productivity gains are bad. Rather, increased efficiency simply means that fewer inputs produce greater outputs. As it stands, labor is among the most significant inputs for a majority of businesses. The acceleration of what artificial intelligence can do is poised to multiply the impact of individual human workers -- while obsoleting others.

Go here to see the original:

Salesforce's Big Bet on AI Shows How Automation Will Affect Knowledge Workers - Inc.com

The same technology that makes self-driving cars possible could also transform wheelchairs.

For someone with advanced Lou Gehrig's disease or severe paralysis, a motorized wheelchair can be very hard to use: If you can't move a joystick with your hand, you have to use a switch embedded in the headset or a "sip-and-puff" device controlled with the breath.

The devices only allow you to control one thing at a time: You can adjust your speed or the direction that your wheelchair is pointing, but not both simultaneously.

[Photo: argallab]

"This basically makes it so the operation of the wheelchair is much more challenging, especially when you're dealing with tightly constrained spaces," says Brenna Argall, a professor at Northwestern University who is developing technology for autonomous wheelchairs in her Assistive and Rehabilitation Robotics Laboratory. "What this means, in practice, is that for many people it's a burden or very fatiguing mentally and physically to operate the wheelchair," she tells Co.Exist.

And there's the fact that some people don't have enough motor control to be prescribed a wheelchair. Children who can't easily use a wheelchair, similarly, may not be allowed to bring it to school.

Autonomy can change that by outfitting wheelchairs with sensors to avoid obstacles in much the same way a self-driving car does. Several researchers are working on variations of the wheelchair-adapted technology. At Oregon State University, a team is developing a low-cost kit that could be added to existing wheelchairs. At MIT, a team is developing a self-driving wheelchair that could be used in nursing homes or hospitals.

[Photo: C. Jason Brown]

"There is a lot we can borrow from the field of autonomous robots and what mobile robots have been able to do on their own for decades now," Argall says.

Her focus is on developing a wheelchair that can leave users with as much control as possible. "We don't want to take autonomy away from peoplewe only want to add autonomy," she says. The system balances human and robotic control; if the wheelchair senses that it's going to run into something if it continues on a path set by a human, it makes only minute adjustments instead of taking over and driving on its own. The final device will likely give people options for how much control they want to retain.

The automated wheelchair may still be in development for another five years, as the team continues to refine the technology. Argall's lab, which is part of the Rehabilitation Institute of Chicago, is moving into a new research hospital this month called the Shirley Ryan AbilityLab, which puts researchers directly next to therapy spaces for patientssomething that could help aspects of the technology's design, like the lab's wheelchair or robotic prosthetics.

"We're going to now directly see people interacting with their therapists in ways that maybe we wouldn't have seen in our lab, which might cause us to think of different considerations in the technology we're building," Argall says. "And vice versa, they will see what our technology is capable of or not capable of, and they might have ideas for how it can be changed or something new that we could develop."

If the wheelchair can be commercialized, it could change lives, potentially giving someone the ability to go to a job by themselves, or do daily tasks that would have required a caregiver in the past. "It could have big implications for costs in terms of caregiving," she says. "It's also been shown that independence matters a lot for people's mental health . . . part of that is because of your loss of autonomy, and this can help with that as well."

Slideshow Credits: 01 / Photo: C. Jason Brown; 02 / Photo: Sally Ryan; 03 / Photo: Sally Ryan; 04 / Photo: argallab; 05 / Photo: argallab; 06 / Photo: C. Jason Brown; 07 / Photo: C. Jason Brown;

Read the original post:

The Next Frontier In Automation: Self-Driving Wheelchairs - Co.Exist