Unintended Consequences of AI: What You Need To Know
I remember logging into my first AI ethics meeting and thinking, “What the hell am I doing here?” As we went around the virtual room, people offered their titles and backgrounds in simple introductions. Clearly, the organizers made some mistake inviting me. I mean, how the hell was a job post expert going to help in a room full of PhDs, AI experts, data scientists, and lawyers??
It is clear to me now that the diversity in education and experiences was exactly what made this group effective. There was always 1 person in the room who could flag an issue. The conversations we had changed the way I see the world - let alone AI and ethics.?
For example, we talked about recruiting nurses by geography. In this case study, the HR team reported that nurses were quitting because the commute was too long. The simple AI solution? Add a zip code and reject people who lived too far away. Seems straightforward enough, right??
Then our moderators layered on the demographics of the communities within the acceptance vs the rejection zone. The acceptance zone was over 90% non-hispanic whites and the rejection zone was a majority minority area. This company was asking for help with getting a more diverse talent pool while automatically rejecting the diverse talent pool based on 1 criteria. Ouch.?
It’s a scenario that could so easily be multiplied across HR and recruiting as we adopt AI. The catch? There’s not usually an AI ethics board in your company to ask the hard questions, just a pushy salesperson who’s trying to finish implementation before anyone asks a hard question. (PS: If you’re looking to build your own in-house AI ethics board, let’s talk .)
That’s why it’s critical that we have these conversations now about how AI really works. Recruiting and HR teams need to be empowered with a why, where data comes from, and the ethical considerations behind the data.?
That’s why I’m hosting this free webinar. I learned a lot in this ethics board. It’s about time I pay it forward by sharing it with you. Register here .??
Why should you attend?
Well, first of all: time with me (kidding). But really, here’s what I want to cover:?
领英推荐
--> Identifying common biases in AI-generated content and requirements,
--> Free tools for removing bias from any recruiting content you create,
--> Learn to create more effective prompts to generate quality results on the first try,
--> Create, test, and try free tools in a workshop format with me.?
Free registration here .*?
*Even if you can’t attend at that day and time. Registering is the only way you’ll get a copy of the recording (which will only be available for 30 days).
If AI isn’t your thing? Fine by me. Check out some of the other articles I wrote last month.?
Talent Acquisition ~ Heart for Family & Farm ~ People Person ~ Learner ~ Optimistic ~ Positive Thinker ~Human Resources - Recruitment & Employment~ RACRx~ RAAS ~ Reach Out & Read
3 个月Excellent presentation today, Kat! Thanks for sharing your expertise!!!
Recruiting top talent for Management & Engineering roles in chemicals, foods, plastics, & related industries
4 个月Bias is very strong with AI. Did you every play around DalieAI picture generation? Notice all men in ties ( was a chemical lab scene). Re-promts then gets females with long hair in skirts and heals. Like who's standing on those heals all day. Never a diverse looking team unless the prompts are very deliberate and specific. ( Just a pic for a proposal so I went in a different diection.)
Author, creative writer, thought leader and content marketer for HR technology and professional services, emphasis on job search, workplace equity, and social listening.
4 个月I have heard companies do this manually too. The thinking goes that people who sign on for a long commute are more likely to be a flight risk. I think this thinking is discriminatory but arguably not an illegal type of discrimination depending on the rationale, which makes it tough to criticize. Totally rejecting candidates that live outside a certain zipcode does seem unreasonably unfair to those candidates, though, and I could see how this could also be used in more nefarious ways.
?????Senior Recruiter (Startup->SMB->Enterprise Global F50->500 exp) - Career Strategist| Writer/Author (check out my articles!) Recruiting remote Solutions Engineers in N. America (US/Canada).
4 个月My recruiter self is going to push back on the zip code issue. It should not be up to the employer to determine if a commute is “too far”. Using zip codes to filter OUT candidates negatively impacts people of color. This has been a discussion point for decades.