Employers Start Using AI To Set Raises Based On "Predicted" Performance

Welcome to the future, citizen, where employees not replaced by robots will be granted raises based on their predicted performance by a computer.

And no, this is not some dystopian vision of the future: IBM is already doing just that - using their "Watson" artificial intelligence to look at an employee's "experiences and projects" and integrating data from their internal training system to assess whether an employee has gained new skills. 

According to Bloomberg, when performance review time rolls around at IBM, employees get judged not only on their past accomplishments (and failures) but also on how they will perform in the future.

How can IBM predict the future? In a word: Watson.

Managers then incorporate Watson's assessment into account when they make bonus, pay and promotion decisions. 

"Traditional models said if you were a strong performer in your current job that was the singular way that you got a promotion," said the IBM vice president for compensation and benefits, Nickle LaMoreaux, adding: "Well, we certainly still care about performance"... but there's a catch:

That now includes hypothetical future performance, too. IBM claims Watson has a 96 percent accuracy rate, as compared to IBM’s internal analysis with HR experts. The company spot-checks employee performance against its predictions.

Employers have traditionally relied on past accomplishments, performance improvements and other hindsight metrics to make compensation decisions. As LaMoreaux explains, "the half-life of skills is getting shorter and shorter," meaning that things employees could do yesterday doesn't matter as much compared to what they are predicted to do tomorrow. 

A survey of over 2,000 organizations by Willis Towers Watson found that over 40% of people are planning or considering changing the primary focus of performance management to include future potential and new skillsets. 

The shift to skills-based performance management is motivated, in part, by employers that say they’re struggling with a skills gap. A National Federation of Independent Business survey from June found that more than one-third of small businesses have open positions. The existence of a worker shortage is debatable. Employers claim they can’t find the right people, but they also haven’t raised wages much, indicating they’re not trying very hard. -Bloomberg

IBM says it can't find enough candidates with the right skills to fill seats - particularly in technology positions. In order to adapt to changing needs, IBM shows current employees what skills they require so that they can undertake corresponding training in order to "level up." As a result, IBM says their employees take an average of almost 60 hours of education per year.


NoDebt Tarzan Mon, 07/09/2018 - 20:18 Permalink

You know, there is a better than average chance that this software will be used to justify ditching a lot of the "affirmative action" hires that seem to hang around most workplaces forever.  

Goes like this:  If an employer decides to ditch a bunch of "minority" workers, it's automatically racism, even if those workers have clear, demonstrable sub-par performance or productivity.  But if an "unbiased" computer program decides who gets the axe, you've got legal cover to jettison the dead wood, since the decision was based only on "facts" and "data".

If you don't buy this argument, I'm not going to try to convince you.  I'm not sure I believe it myself.  Just trying to toss out an alternate theory here for consideration.


In reply to by Tarzan

NickelthroweR NoDebt Mon, 07/09/2018 - 22:31 Permalink


HR is where most companies have the biggest problems leading to the fat payout lawsuit.  Removing the human element is tantamount going forward plus the shareholders are going to demand it as it WILL increase profits as the diversity hires 100% find themselves unemployed.  

Job interviews are being done by narrow A.I. via video so all the applicant has to do now is answer some questions into his or her camera phone.  Because machine learning is how the narrow A.I. has taught itself to find the best candidate, no one knows how the A.I. is making the selection.  Hell, it could be looking at veins in your eyeball for all we know but it does a superior job to a human HR person.  

The machine has to teach itself so how do you code in a preference for race, gender or sexual orientation?  It wont be possible once there is competition for an A.I. HR department.  If someone makes a superior A.I. that can run your company and skyrocket profits and get rid of all the litigation and dead weight then Gattaca, here we come!  Everyone will jump on board.  Oh yeah, once blacklisted by any one of these Bots you will be blacklisted by all of them.

In reply to by NoDebt

Whoa Dammit Tarzan Mon, 07/09/2018 - 20:21 Permalink

If you're over a certain age, you will never get a raise because your future performance will inherently not have the same potential as someone younger.

Also, IBM is not an exemplar company for the development of anything cutting edge.

And finally has anyone ever learned anything of value from a "corporate training" class? 

In reply to by Tarzan

Antifaschistische DiotheDog Mon, 07/09/2018 - 21:03 Permalink

I've done a lot of hiring...can you add psychiatric medications into the model? 

...and percentage body fat.

...oh, and a muslim mans ability to ever work for a woman...holy shit has that one caused me some problems. 

I recently made another observation based on a sampling of 5.   A 50 year old lady who has never had her own biological children starts having all kinds of emo issues.   Is my sampling just too small?

In reply to by DiotheDog

UselessEater wee-weed up Mon, 07/09/2018 - 20:09 Permalink

The Equal Opportunity crowd will love it. It's going to devolve into totalitarianism on steroids, probably being discussed at Davos, Builderberg, masonic lodges and other secret societies. The test model is working well:


So glad we won WW2, it was very important in halting the expose of secret societies usurping our nations banking, economy, education, publishing, religious and family institutions. Now we can all become China bots or Congolese savages.



In reply to by wee-weed up

Krink26 Mon, 07/09/2018 - 19:39 Permalink

Im sure a liberal bias will not be in the AI a la FB, Twitter and Google. Then again, it isnt like HR departments are loaded with feminazis....

small axe Mon, 07/09/2018 - 19:49 Permalink

mandatory unpaid training and mandatory unpaid overtime will be required of all career-track employees...in other words, slavery is alive and well


XBroker1 Mon, 07/09/2018 - 19:49 Permalink

using their "Watson" artificial intelligence to look at an employee's "experiences and projects" and integrating data from their internal training system to assess whether an employee has gained new skills. 

Don't they have to also 'probe' your spouse to see if marital problems are in the offing &  potential future drinking binges? 😊



konadog Mon, 07/09/2018 - 19:53 Permalink

The phrase "garbage in, garbage out" comes to mind here. Mf'ing dystopian siht. I can just hear these worthless weasels that call themselves managers/leaders, "oh don't blame me, the computer decided."

Giant Meteor Mon, 07/09/2018 - 20:04 Permalink

Hal 9000, "I know I've made some very poor decisions recently, but I can give you my complete assurance that my work will be back to normal. I've still got the greatest enthusiasm and confidence in the mission. And I want to help you."

pocomotion Mon, 07/09/2018 - 21:34 Permalink

And it's 2 for me and 1 for you....  Are you not satisfied?  Replacements are lined around the corner.  If you give back the 1 then this will show us your loyalty to our institution.  Good job, employee #431....