Nevada will use Google's AI to help decide if people should receive unemployment benefits

[midian182]

A hot potato: In the latest instance of generative AI being used for something that AI probably shouldn't be used for, Nevada is set to become the first state to use the technology to make recommendations on unemployment benefit appeals. Since these systems are prone to errors, bias, and hallucinations, there are understandable concerns about the move. The fact that courts may not be able to overturn rulings made on the basis of an AI's mistake is compounding these fears.

Nevada's Google-powered generative AI system is set to launch within the next few months, reports Gizmodo. Its function will be to analyze transcripts of unemployment appeals hearings and issue recommendations to the humans that give a final verdict.

Nevada's Department of Employment, Training, and Rehabilitation (DETR) paid Google $1,383,838 for the AI tech, which uses a Google cloud service called Vertex AI Studio. It will generate recommendations based on transcripts of unemployment hearing appeals and rulings, comparing the information to previous cases.

The system is being introduced to expedite the appeals process. DETR's information technology administrator, Carl Stanfield, said the employees take about three hours on average to write a determination. The AI system will reduce this to just five minutes in some cases. That could be especially helpful in Nevada, which has a backlog of more than 40,000 appeals resulting from a rise in unemployment claims during the pandemic.

The downside, of course, is that generative AI remains notorious for spouting falsehoods and generally making things up. The DETR says recommendations made by its AI will be reviewed by at least one human. If the person agrees, the decision will be signed off. If they don't, the document will be revised and the agency will investigate what happened.

"There's no AI [written decisions] that are going out without having human interaction and that human review," said DETR director Christopher Sewell.

Google said it will work with Nevada to "identify and address any potential bias" and to "help them comply with federal and state requirements."

There are concerns that DETR employees will feel forced to rush their reviews of the AI's recommendations, thereby clearing the backlog faster. It's especially worrying as this could impact a claimant's ability to appeal a case.

"In cases that involve questions of fact, the district court cannot substitute its own judgment for the judgment of the appeal referee," Elizabeth Carmona, a senior attorney with Nevada Legal Services, told Gizmodo. This means a court might not be able to overturn a human decision based on an AI error. It would likely be difficult to pinpoint where an error originated, too.

Despite these legitimate concerns, don't be surprised to see more states start using AI to speed up processes by helping to make important decisions.

Permalink to story:

Nevada will use Google's AI to help decide if people should receive unemployment benefits

 

[yRaz]

>work for 15 years doing masonry
>get denied all benefits during covid
>see bums getting 2400/w on money going taking vacations
Unemployment is hell. Every year I file I spend the majority of my "time off" fighting with them and eventually get a check sometime in April after I get laid off in December. This is the main reason I want to move out of the North East US, so I don't get laid off in the winter and spend all my time and money fighting with Unemployment. You already never get to talk to a person, it's all online.

It should be as simple as:

"this is my job, this company told you that I was laid off because they have no more work for me"

"That sucks, here is your check"

Not:
"We see in our records that you only worked 32 hours the week of *insert random date* is that true?"

"I don't know"

" sorry, no Unemployment for you"

Then I have to spend the winter calling verious people to confirm what hours I have when and then resubmitting my Unemployment application. The worst part is, sometimes the company I worked with submitted my hours wrong to Unemployment so I have to call with them to resubmit my Unemployment information.

 

[ScottSoapbox]

Don’t worry! Not a single government bureaucrat will be laid off from the unemployment office after creating this new AI efficiency.

 

[brucek]

Doing this for the actual appeal is horrible. That's the step that's supposed to have the highest level of expertise to supervise and correct all the steps before it.

Where this could be more useful is if the first person / panel / judge doing the initial decision wrote his/her draft, and then had this AI to look for common mistakes and point out areas to maybe double check before issuing it.

 

[NumberNine]

I'm speculating that this all was discussed over a massive lunch paid for by the software vendor.

 

[Dr Roboto]

"There's no AI [written decisions] that are going out without having human interaction and that human review," said DETR director Christopher Sewell.

We all know darn well that the humans are just going to rubber stamp the AI suggested action. Does anyone really think that the human is going to read the persons claim "completely" and then read the AI's suggestion "completely" and then do an actual comparative analysis to see if the AI's suggestion is correct? BS

Google said it will work with Nevada to "identify and address any potential bias" and to "help them comply with federal and state requirements."

This is the more disturbing assumption. LLMs have been shown time and time again to have clear biases in their training data. Just wait a few years until a study on the results come out and we see a clear bias in whole gets approved or not. Yet, these people will have no recourse for action.

I support the use of LLMs, but common sense needs to used and they are NOT ready for use in a system this complicated and important. This is clearly a case of the uneducated looking for a quick fix to a problem and the company being more than willing to take the money.

 

[human7]

There's a reason they scrapped their "do no evil" mantra many years ago. Their leadership making choices like this only furthers that trend.

AI could have a place in making decisions like this someday, but for now people should be extremely skeptical.

 

[Jme Kerul]

...
Unemployment is hell.
...

Sometimes, I think "socialist" Europe is heaven compared to USofA.

 

[nitebird]

I'm surprised that a tech site is so techno-phobic. AI like any tool can be used to help or hurt depending on if it is used correctly or incorrectly. AI can not only offer a suggestion on how to resolve an appeal, it can also rate itself on how confident its recommendation is. If it is a common, open and shut case, which the AI can recognize and process in 1 second, that saves the case worker's time to process the more difficult edge cases that the AI rates its recommendation as low confidence. If we have AI review how the government spends our tax money, it can reduce fraud. Many parts of government use decades old technology that requires a lot of labor to perform simple tasks, if we automate more of the tedious stuff, it can mean less wait time, more fair judgments, lower taxes. Can it go wrong, sure! Does AI make mistakes more often or less often than people? If it means 50% lower mistakes, and 99% less wait time, would you pick AI to process your case? This is a tech site, not an Amish site LOL I expected to see a more professional assessment of pros and cons, risk and reward.

 

[yRaz]

Sometimes, I think "socialist" Europe is heaven compared to USofA.

There are pros and cons to both. The US is far from perfect but I'd rather be here than anywhere else.