The Great AI Debate: Believers, Skeptics, and the Undecided

I'll jump in and start.

I'm happy to start this off.

I think that my view is that, I mean, if you say from a, from a starting perspective of should they be able to use it without AI safe from the place? I don't think anybody should be doing that.

I mean, I think it's just fundamentally, if you, if you look at, Hey, do you want to be doing this appropriately and properly? I think you really need to design a program where when you're using artificial intelligence, you are using it in a safe and meaningful way.

Um, You know, I think my general view is from an anti-discrimination perspective is we also already have anti-discrimination laws in place on the books across the country.

We don't need to create specific artificial intelligence anti-discrimination laws, right? So we handle them through the normal methodology by which you'd handle any sort of anti-discrimination claim or something similar and go through that traditional process in the laws and regulations that come in place.

We don't create these structures for any other tool.

email anti-discrimination laws or other technology that you have.

So why are we doing something specific or unique for artificial intelligence? That probably is really my starting point of where we come from, from the anti-discrimination angle.

I don't think that that probably is where we're actually going to end up seeing it go.

But at least in my mind, that's probably if you would have me and give me the control over it, that's how I design it and say, hey, follow the rules that we already have in place.

But everyone should really be looking at it overall of establishing their overall safeguards.

I agree completely.

I think there's a lot of anxiety with AI because a lot is unknown.

And so people assume that we're putting a lot of decisions, impactful decisions in this gray box and just spit something back out and we just hope it's right.

But there's those safeguards to make it more tangible.

It's just called an EVA.

It's something that's been around for a long time.

It's an evaluation of the model, of the logic and of the output.

And so if you have a large staffing organization, theoretically, If you have access to your data, you could go back and reverse engineer the same evaluations against humans that you would with AI.

That was done in this article that says, how do people make decisions? Do they have bias in that decision-making process? And you can measure it.

And I think what you'll find with a lot of these AI models because of training, if you get detail-oriented, you get hands-on, you can beat your human metrics effectively.

So you may find more bias on the human side than you may on the AI side.

But that doesn't take away from that core that you need to have some sort of safeguard, some sort of evaluation, and you should record it and keep those records as the legal space develops.

Oh, I was just going to speak from a liability standpoint, specifically being an operator and a staffing company where we use several different tools and resources from, you know, third party vendors, as well as actual staffing HR technology companies who provide a service or a product to our company.