Hunter S. Thompson went to Las Vegas in March 1971 to write a glorified photo caption. Sports Illustrated wanted about 250 words on the Mint 400 desert motorcycle race.
He turned in 2,500 and the magazine rejected them. I know the feeling of overwriting.
A month later he went back for the National District Attorneys Association’s Conference on Narcotics and Dangerous Drugs.
Rolling Stone ran the result as a two-part serial in November of that year, and the book followed in 1972. Thompson said he put a fictional framework over what started as journalism, and part of that framework was squashing the two trips into one.
I’m doing a version of that, I guess.
He had a month between trips. I have five days. (Point: me)
His assignment was 250 words and it turned into a book about the American Dream. Mine is whether an HR agent works. (Point: Thompson)
If all goes to plan, you should be getting this in your inbox as I am in the air.
Rising was 17,000 people at one vendor’s conference. HR Tech is more than 7,200 people and 400 vendors who each want to tell you theirs is the one that works. Also, a shout out to the four of those sponsor 3Sixty Insights this year, so thanks to Insperity, Paylocity, Tavio®, and LYTIQS.
Both shows will give me the same story about AI in HR. I’m going in with four questions about what’s pushing it, and they’re for the whole floor.
Where are AI agents doing real HR work?
Every vendor will have an agent on a stage. Very few will have one that has handled real employee requests long enough to break something.
What’s pushing this is that employees don’t want another portal. They ask in Slack, Copilot or ChatGPT, and the vendors are following them there. Redii is launching a benefits agent that lives in Slack, Claude and ChatGPT, which is great for the employee.
My guess is that it’s a problem for anyone who charges per seat, because an employee asking a chatbot isn’t logging in to anything. The HRIS and payroll systems behind the chat window aren’t going anywhere, and they still hold the data. So I think the real fight at this show is over who owns the chat window and who just becomes the database behind it.
Are companies building talent or just buying and cutting it?
Junior hiring is down and AI is getting blamed for the layoffs. I don’t buy that AI is the main reason. A cut makes a stock price happy this quarter, and AI is a good story to tell about it.
Forrester expects half of AI-attributed layoffs to be reversed. Careerminds surveyed 600 HR professionals who had made layoffs and found that 55.1% never formally considered reskilling or redeployment. Then there’s IBM, whose CHRO says the company is tripling U.S. entry-level hiring after rewriting every job.
I want to find out whether IBM is the exception or just the one willing to say it out loud, and what it cost them in training and manager time.
How do you know the candidate, and the data about them, is real?
Gartner found that 39% of 3,290 candidates used AI during the application process, and it predicts that one in four candidate profiles will be fake by 2028.
AI took the cost of applying, and of faking an application, down to almost nothing. Volume goes up, recruiters can’t read it all, and so they buy tools and data to do the reading. Every recruiting vendor says its data is better than the next one’s. Both of those things can’t be true at once, and I plan to ask each side how it knows a person exists and what it does when the check fails.
Who owns pay now?
The cause here is a law. Pay ranges in job postings pulled pay out of the comp team and put it in front of recruiters and hiring managers. The EU’s pay transparency deadline was June 7, and only four of 27 countries made it.
A recruiter can’t post a range for a job family with inconsistent leveling. So somebody has to decide what the job is worth before the posting goes up, and I want to hear who that is now and who takes the blame when the number is wrong.
I am in early Tuesday and out late Wednesday. 36 hours in Vegas. What could go wrong?
This post was originally published on Beacon Turn.