Please tell the robot where you see yourself in five years

Note: I warned you on Monday - it is unofficially 'robot week' here on the blog. Bail out now if this is not your kind of thing. Don't worry, I won't know if you did. Probably.

Researchers and engineers have long identified manufacturing, warehouse operations, and even more 'advanced', complex, and interactive processes like patient and elder care as potentially fertile ground for the further automation and robotization of the economy and society. While each new encroachment of these increasingly better, smarter, cheaper, and more reliable pieces of technology, many of us pause to take stock of just how near or far we see our own roles and jobs from this impending and inexorable onslaught.

And also increasingly, the answer to the question of 'Just how close is my job to being replaced by a robot?' is 'Closer than you think.'

For the folks who read this blog, mostly HR and Talent management professionals, 'basic' kinds of automation have mostly made our jobs better, easier, simpler, and allowed us to spend more time on complex and higher order activities. Instead of endlessly keying and re-keying data on dumb terminals, we have our employees process their own transactions on their iPads. Instead of calling up references for soon-to-be-hired candidates, we send the references a link to an online survey and have some software send us any red flags or exceptions. You get the idea. 

Automation in HR has no doubt helped make our operations much more efficient, reduced errors, and with the latest batch of exciting new technologies, given us insight into our organizations that would have only a few years ago been impossible to see.

But will automation in HR ever go even further and reach into one of the 'essential' HR and Talent functions - the actual assessment of a candidate in the traditional interview setting? Some researchers at LaTrobe University Business School in Australia are betting that the answer to that question is 'Yes'.

Check this excerpt from a piece on the Australian Financial Review site, 'Interviewed for a Job by Sophie the Robot':

With big eyes, a feminine voice and some interesting dance moves, Sophie is rather cute but don’t let that fool you.

Sophie could soon be conducting your toughest-ever job interview, monitoring not just what you say but tiny twitches in your eyebrows that give clues about how you really feel.

Sophie and her fellow “human-like” robots Charles, Matilda, Betty and Jack plus two as yet unnamed robots are the product of a research joint venture between La Trobe University Business School in Melbourne and global electronics giant NEC Corporation in Japan.

NEC provided the robots and La Trobe is adapting them for use in recruitment, health care and as “emotionally engaging learning partners” in Australia. Rajiv Khosla, who has been driving the project since its inception, says the robots are a “world first in the area of recruitment”.

Sophie was already involved in trial interviews of candidates for sales jobs, asking 76 questions about selling.

“She captures their [candidates] cognitive verbal responses and captures their emotional responses by monitoring changes in their facial expression,“ Khosla says.

Khosla insists robots will not replace humans conducting later stage interviews or employers making final hiring decisions.

Sure, the interviewer robots are just here to help the process, that is all Mr. or Ms. HR Director. There's no way a robot for gosh sakes would be better at assessing the validity, truthfulness, accuracy, and the like of a candidate's responses.

There's no way a robot would be able to compare, in seconds, and with astonishing precision the information provided in the interview with a candidate's resume, LinkedIn profile, social web exhaust, old resume from five years ago they forgot was still on Monster.com, and so on.

There's no way a robot would be able to sense subtle eye movements, increase in respiration, body temperature changes, as different questions get asked and answered.

There's no way a robot could, in fractions of a second, compare and contrast dozens of candidates' answers to the same questions and produce detailed analyses on quality and accuracy, and perhaps truth across these answers.

There's no way a robot could also compare the new candidate's responses and reactions with the last persons hired into similar roles, and how the successful and not so successful hires reacted in similar circumstances.

There's no way a robot could conduct dozens and dozens of interviews across a high volume hiring period for retail or food service without getting tired, crabby, maybe even a little forgetful.

Nah, no way a robot can do all that. We need people for all those things.

Happy Wednesday.

When the robots realize what they don't know

Warning in advance to the good folks that check out the blog here, I sense another 'robot' kick coming on.

So bail out now if indeed the tales from the robotics frontier are not really your thing.

Today's dispatch from the robot wars - Robots Learning Better Ways to Ask Clueless Humans for Help, is from the IEEE site. Scientists and robotics researchers at MIT, (it is always MIT it seems), have taken the 'clueless' robot that traditionally can do only what is specifically programmed to do, and enhanced it with the ability to interpret where and when it needs assistance to carry out an assigned task, and then to communicate that need for help to an equally clueless human counterpart.

In the research conducted at MIT, scientists have taught the robots to ask for help to complete the assembly of a piece of furniture. They can make requests for assistance like 'Please flip the white table top over' and 'Please hand the blue robot the black table leg'. check out the embedded video below to see how this process and technology work (email and RSS subscribers will need to click through)

If you watched the video all the way through (you deserve a medal for that), you might have caught the most interesting line of all - that this kind of robot technology would eventually allow humans to supervise larger and larger groups of robot workers.

Robots are great at completing the majority of most simple tasks, but eventually there are one or two steps in whatever you want a robot to do where it's much more likely to fail. Giving robots the ability to recognize these failure points and then intelligently ask for assistance could open up many more tasks to at least partial automation, and it's likely to have the most impact in variable, unstructured environments.

You know, like the kinds of environments and types of jobs that we keep thinking are going to be safe from the eventual robot uprising.

Have a great week!

VIDEO: Unboxing the future

The 'unboxing' video format, (essentially, a video of someone unboxing a new gadget like a computer or a smartphone and describing the contents and packaging), has enjoyed a run of popularity amongst the geeky set.  I mean who doesn't want to watch a choppy video with bad sound and lighting of a random 15 year-old kid unpacking a new Xbox?

Yes, 'unboxing' videos are generally horrible, and it is with that horribleness in mind that I run the risk of alienating those readers that not only don't know or care about 'unboxing' but also are lacking my interest and fascination with advances in robotics.

Check the video below, (email and RSS subscribers will need to click through), of a team at MIT receiving and unboxing its Atlas robot (built by Boston Dynamics) to use in the DARPA Robotics Challenge. I will have a couple of (probably nonsensical) comments after the jump.

What is Atlas all about?  Here is a small description from the Boston Dynamics site:

Atlas is a high mobility, humanoid robot designed to negotiate outdoor, rough terrain. Atlas can walk bipedally leaving the upper limbs free to lift, carry, and manipulate the environment. In extremely challenging terrain, Atlas is strong and coordinated enough to climb using hands and feet, to pick its way through congested spaces.

The team at MIT will develop software to control and command Atlas to perform various actions in a disaster response situation - think things like defusing bombs, looking through tornado damage, potentially working in toxic waste spills, that kind of thing.

Why should you as an HR/Talent pro care about something like Atlas, and its capability and potential?

Because like lots of other technologies, these kinds of advanced robotics applications might start in research universities or government labs, but the best ones almost always become a part of the workplace.

Because at some point you as an HR pro will get asked a question from the CEO something along the lines of, 'Can't we find a way to automate that, instead of opening another assembly plant?' or 'Can you get me a cost/benefit breakdown of buying 10 new Baxters vs. hiring 50 new assembly workers?'

Because at some point someone you work for is going to see an 'unboxing' video like the one above from MIT and think, (perhaps erroneously, perhaps not), that pushing advanced automation further and farther into the business is getting easier and easier - not unlike how easy it is to set up that new Xbox.

Maybe I am completely off-base on this, and the time when the average HR pro really needs to concern themself with this kind of thing is decades away.

Or maybe I'm not wrong, and sooner than not you will have to add a 'person type' in your HRIS for new employees named Baxter or Atlas.

VIDEO: The robot would like a sidebar, please

How do you balance the demands of the modern workforce for flexibility around schedules, locations, and desire to not cut back on that white-knuckle ride on the daily commute, with many organizations desire to foster a collabortive and innovative environment that to many leaders only comes from workers 'physically being together?'

Meet your future colleague Ava 500, or rather, the robot that your future colleagues will be driving around the office or plant or store if the vision of the folks at iRobot and Cisco comes to pass.  Ava 500 combines the mobility and navigation capability from iRobot, with Cisco's teleprescence technology into a robot technology that can be used to teleport anyone in the organization regardless of their physical location to any other location that is equipped with an Ava 500.

Check the video below from iRobot to see Ava in action (Email and RSS subscribers may need to click through)

Pretty nifty, right? And did you catch that little feature with the 'robot' drags a couple of the meeting participants out of the room for a little private time? 

I think the long-term key for these kinds of telepresence robots to actually move past novelty and into more widespread use is that they have to seem less, well robotic, and more natural. They need to be able to move fluidly, be aware of their environment, and maybe have a little personality. 

The workers that teleport into Ava have to come across to their colleagues as close to 'normal' as is possible, and using the high-end Cisco telepresence tech is one way to try and achieve that. No one is going to want to interact with a person piloting an Ava 500 if the video feed resembles a dodgy Google Hangout from someone's dreary basement home office.

One thing the video didn't show, perhaps purposefully, is depict two different Ava 500's interacting with each other. In a way, if using a technology like Ava would be so fantastic for connecting one remote worker with their colleagues, then why not 2 or 3 or 20? 

Maybe the workplace of the future will be one that ends up being largely uninhabited by any people, but rather a fleet of telepresence robots that move from meeting to meeting while different workers take turns teleporting in from all over the world.

What do you think - is the Ava 500 coming soon to a workplace near you?

Virtual HR, or, 'Did you ask the HR chatbot?'

While I and many, many others have blogged, talked, and pontificated about how the ongoing advances in technology, automation, robotics, and artificial intelligence continue to 'hollow-out' or eliminate wide ranges of jobs formerly and traditionally that are done by humans, I also often think that many folks don't see these trends as all that interesting or potentially threatening. Most of the people who read this blog, I imagine, see themselves as knowledge workers that bring imagination, creativity, and perhaps most essentially, an understanding of subtle things like culture and attitude to their jobs and careers. Most of us, (admittedly me too), say of think things like 'My job is just too complex and ever-changing for it to even be outsourced to a less-expensive human (much less a robot).'

Possibly. But it also seems likely that given enough time, access to ever-improving technologies, and the right economic incentives, there are enterprising people and organizations that even if they couldn't completely automate or robot-icize everything you do, chances are a fair amount of even what we creative types do is already routine enough that the robots could do a passable, if not better (and cheaper and will less of a bad attitude), than we do.

But again, I know you don't really believe me, as you are (probably) and HR person that is reading this, and automation in HR has only meant changing how the transactional work of HR, (forms, time-tracking, payroll, etc), from paper-driven processes to computer-driven ones, (and often, initiated by employees and completed by managers with little direct involvement from HR). The important work in HR and in many other organizational functions still, and perhaps for a long time yet, remains the exclusive domain of humans - which humans (see the self-service HR example above), matters less today than it used to. 

But automation is coming - not just to manufacturing lines or driverless cars or better algorithms and assessments that can screen candidates much faster and more efficiently than you can. Perhaps in HR automation will have to look a little different than what we expect, since so much of the profession is about people - talking to them, understanding them, evaluating them, and motivating them - and ultimately helping, (or concluding that we can't), help them.

Those kind of interactions, even at a basic level, can't be automated yet. Right?

Well, maybe not yet, but that doesn't mean they won't be one day soon.

Don't believe me? Then ask Ivy her opinion.

Who is Ivy? The latest in automation - this time form one of the country's largest employers, and right in your area of interest - the delivery of HR services.  Check the details from the Jobs at Intel blog.

Okay, the newest thing we’ve launched is a “virtual HR agent”. What’s that, you ask? You know when you shop online, whether it’s for new gadgets or it’s for a plane ticket to go somewhere or maybe it’s just for odds and ends, some websites have a virtual agent that will answer FAQs for you and guide you through the process. Our new virtual HR agent, we named her Ivy, is set up to do the same thing, but for our employees at Intel (so this is an internal tool.) If employees have questions about their pay, stock, benefits, or other HR programs, they simply bring Ivy up on the intranet and type in a question. Ivy uses a combination of natural language processing, artificial intelligence and optimized search to find the answer to the question. Also, magic. Okay, well, it’s like magic to me, so…  As of today, Ivy has 4,331 possible responses. How do I know that number so exactly? I led the team that wrote all the responses. You can bet we’re excited for the launch after all that work!

 Catch all that?

Ivy, or the virtual HR chat agent, has over 4,000 possible responses to any employee question about pay or benefits or other HR programs, and using the same kind of intelligence we've seen in a consumer or retail environment, provides HR services to Intel employees. As it is an internal-only tool, I'm not able to test it out, but it stands to reason that with over 4K responses, and the ability to 'learn' and adapt, that over time a tool like Ivy would be able to do more that respond to simple questions, and provide more complex answers to more difficult questions.

Yes, Intel's HR team has to provide the 'intelligence' for Ivy to work, and that, as yet, is still a human job, but what if employees at Intel begin to prefer dealing with virtual HR over real HR? 

I'll leave you with more from the Jobs at Intel blog about Ivy:

Ivy’s no chatbot and she’s not backed by a human “behind the curtain”. She’s all software. We’ve got lots of metrics in place to monitor her performance and our employees can give a star rating to each interaction. Using the performance data and star ratings, we can tune Ivy to make her even better. Beyond that, what’s weird is that she learns. Seriously. Her artificial intelligence gets better as employees ask her questions. Amazing.

The money line in that?

We can tune Ivy to make her even better. Beyond that, what’s weird is that she learns.

Can you say the same about the people in your HR organization?

Happy Tuesday.