Live data from Hacker News

Walmart, Delta, and Starbucks are using AI to monitor employee messages

cnbc.com

41–50 of 104 posts

Re: Walmart, Delta, and Starbucks are using AI to monitor employee messages

#42

Ok, exactly which communications does this AI monitor? This seems to be internal corporate coms (corporate email, internal chats) which are not subject to privacy. But the article is very unclear, leaving the reader to assume that they are monitoring social media or something else external to the organization. Internal company communcation channels are open to inspection by the company. Whether that company wants to…

> Internal company communcation channels are open to inspection by the company.

That is the current rule (at least in some places), but we're conflating the issue of 'is it legal?' with 'is it right?'.

> Whether that company wants to use AI or pay for agents to read ever internal email is up to them.

IMHO this argument needs to be buried once and for all, especially on HN. Automation by computer transforms the power of otherwise manual functions. For example, government could always surveil people on the street, but they had to do it manually. With computers, using cameras and facial recognition, they can automatically track the entire population (theoretically, and later if not now). It's not the same thing as manual surveillance of a suspect.

> Want your bosses not to hear you rant about a new corporate policy? Don't sent such comments via company email.

All that said, I think there are legitimate questions. The mass surveillance enables management to stay ahead of any employee organizing or other political (as in office politics) moves - management knows before the employees themselves realize that lots of people agree with each other. That includes unionization, for example.

Yes, if the company allows it then you can communicate through outside channels, but can you reach the whole company there? That is certainly a hinderance to any mass activity.

Re: Walmart, Delta, and Starbucks are using AI to monitor employee messages

#43
The real value if removing encryption is access for AI.

If we don’t have a comprehensive privacy legislation, something like a bill of rights for privacy, tied to individuals, AI will become a tool of totalitarianism and total information thought control. Full stop.

You can look at it though todays lens and argue this is just another application of machine learning, but it isn’t just a slippery slope but becomes a landslide as soon as you apply more sophisticated future digital intelligences to it.

Re: Walmart, Delta, and Starbucks are using AI to monitor employee messages

#44

Do we just call all of statistics "AI" now? I fail to see how or why applying basic machine learning techniques for data analysis is "artificial intelligence". This idea that intellection is nothing more than pattern matching is silly and needs to die. Conceptual understanding entails more than just pattern matching.

The AI part is the sentiment analysis of the chat message text and pictures. They're using statistics on the resulting aggregate data. This seems pretty reasonable to me. ML is roughly a gazillion times better than traditional NLP techniques for sentiment analysis. If you're doing sentiment analysis without ML in 2024 then you need to stop and reevaluate what you're doing. If you simply wanted to complain that they'r…

Did you mean to write LLMs instead of ML? Aren't traditional NLP techniques a subfield of ML?

Re: Walmart, Delta, and Starbucks are using AI to monitor employee messages

#45
post #40

As much as this turns my stomach, this is neither new nor unexpected, and honestly this is fairly benign to many of the things that companies do such as precision location tracking throughout the building, using company-issued phones, and using normal security surveillance to monitor "watercooler" talk and things like that. We had our chance to push back against company surveillance well over a decade ago, and nobody…

May I ask why privacy matters at work? Not trying to be a dick. But shouldn’t people at work be doing only the things they’re paid for on company time?

Privacy matters wherever you are.

Even at work, most people like to believe they are trusted to do what they've said they are going to do without needing to be constantly monitored and micromanaged. Anything else is unsettling and oppressive.

I would refuse to work at a company that expected me to give up my privacy during working hours.

Re: Walmart, Delta, and Starbucks are using AI to monitor employee messages

#46

>>>> Using the anonymized data in Aware’s analytics product, clients can see how employees of a certain age group or in a particular geography are responding to a new corporate policy or marketing campaign >>>> “It won’t have names of people, to protect the privacy,” Schumann said. Rather, he said, clients will see that “maybe the workforce over the age of 40 in this part of the United States is seeing the changes to…

In my experience, it’s always the over-40 group who are trying to implement these features to keep an eye on the young troublemakers.

It’s certainly not Gen Z kids in leadership positions implementing these AI snooping policies at these mega corporations.

Re: Walmart, Delta, and Starbucks are using AI to monitor employee messages

#47

>>>> Using the anonymized data in Aware’s analytics product, clients can see how employees of a certain age group or in a particular geography are responding to a new corporate policy or marketing campaign >>>> “It won’t have names of people, to protect the privacy,” Schumann said. Rather, he said, clients will see that “maybe the workforce over the age of 40 in this part of the United States is seeing the changes to…

>>> It's always the over-40 who are under suspicion. Because of our (over 40s) years of watching bad policy after bad policy destroying our work places, and speaking up.

There is a reason companies love hiring young college grads.

Re: Walmart, Delta, and Starbucks are using AI to monitor employee messages

#49
post #44

Earlier quoted context omitted.

The AI part is the sentiment analysis of the chat message text and pictures. They're using statistics on the resulting aggregate data. This seems pretty reasonable to me. ML is roughly a gazillion times better than traditional NLP techniques for sentiment analysis. If you're doing sentiment analysis without ML in 2024 then you need to stop and reevaluate what you're doing. If you simply wanted to complain that they'r…

Did you mean to write LLMs instead of ML? Aren't traditional NLP techniques a subfield of ML?

There's overlap, but many traditional NLP techniques are heuristics based. Here's an example: https://github.com/cjhutto/vaderSentiment

Re: Walmart, Delta, and Starbucks are using AI to monitor employee messages

#50
People often underestimate the effects of losing privacy. A lot of comments here seem to reflect the idea that “If I’m employed by CompanyY then I’m owned by them.” We are spending a majority of our waking hours working, often, and being trained to believe that an authority’s ownership of technical systems means that not only everything we do with other people via those systems belongs to an employer, but that our inadvertent inclusion of ourselves as humans and not simply task performers is “our own fault.” As humans, our psychological wellbeing rests on a sense of autonomy and these systems directly suppress that.
Post reply on HN