At one point in history, people believed that all you needed to run a business was an MBA. Actual knowledge of the product, processes, people, etc. was irrelevant.

A narrow focus on "managing" a team of AI chat bots suffers from a similar narrow-mindedness. Without actual knowledge of the areas that you're relying on the chat bots to manage, you have no way of determining if the work product you are getting is of any use.

Using a team of bots to produce code without domain knowledge or even general purpose computer science knowledge can only eventually end in tears. But I guess if you can generate the kind of mess a bureaucratic consultancy staffed with MBAs, at a fraction of the price... that's progress of sorts?

Actually, it doesn't even take 5 minute of manual work on Google.

Just checking to see if a given citation exists, nevermind the actual content, is a simple matching query. If I were a lawyer, the very first thing I would do would be to dump every public listing of caselaw I could get my hands on as a local searchable case index, in parallel with having access to a tool like lexis-nexis or PACER (with recap) for more in-depth research. It honestly should be one of the intermediate steps in an AI bot workflow - validate that the citations exist... and then verify that the citation actually bolsters the argument.

Suggested tools:

https://free.law/recap
https://case.law/
https://www.courtlistener.com/...

If you're going to turn an AI bot loose to generate arguments for you, the very least you can do is check its homework, same as you would do for any supervised clerk, paralegal, or lawyer in training working for you. That it can generate bullshit faster than you can is no excuse for shutting off your brain and signing off on it without even doing the bare minimum of due diligence.

The next level of work would be having an AI bot analyze each case and the citations used for the arguments in each, to generate a tree of citations that can be used to argue in one direction or another. Then depending on what arguments you want to bolster, you can selectively cite the cases that give more weight to your case, and prepare counterarguments in case the opposition has prepared an equivalent set of trees that will cherry pick case citations against your argument.

But forget robot lawyers generating bullshit cases. What I want to see is the robot trial judge (and the robot red team lawyer playing the part of the opposition) that can audit your case beforehand and pick it apart so you can be better prepared before you go to trial.

I wonder about that. But there's definitely a seeming drive for puffery on individual resumes, and a collective drive for puffery on the entire platform in order to drive mindshare. The same kind of short-term thinking that has people ripping out existing features to "improve" a product, so they can claim that they actually did something in their tenure there.

Instead of doing something to fix a hard problem, say the obscene memory consumption of tabs as part of the base browser, they do things to make Firefox more attractive to say... advertisers who want placement on Firefox's default home page.

This is my impression as a user - I have no window into the workings of the Mozilla team aside from depressing news bits like this one featured on Slashdot...

Easiest solution is to issue employees a corporate credit card that they are responsible for. All reimbursable expenses have to be correlated against the copy of the statement issued by the credit card company to the corporation.

But what about cash expenses, you ask? Issue a per-diem for travel, and a periodic "here's your budget for IT refresh, whatever you don't spend, you get to keep."

My question is, what kind of receipt fraud are we looking at? Invented expenses that they're using to defraud the company, or real expenses that normally wouldn't be reimbursed that they're disguising as reimbursable ones?

Also, wouldn't invoice fraud be a bigger threat? Fake suppliers sending you real looking invoices in the same of actual suppliers, but with the bank details modified to point to the scammer's accounts instead? Instead of hundreds of dollars in fraud perpetrated by dozens, maybe hundreds of insiders (employees), you get tens, if not hundreds of thousands dollars of fraud perpetrated by outsiders trying to pretend that they're trusted suppliers. Or worse... an insider at your supplier deciding to doctor the outgoing invoice so they can skim money off the top...

Learning COBOL is relatively easy (although there are some dialect differences between say, gnuCobol and the COBOL used on IBM mainframes). Learning Z/OS and JCL on the other hand...

There are resources, though... for example the Open Mainframe Project:

https://openmainframeproject.o...

As they say, if you want a job for the rest of your life, learn to work on mainframes...

You're assuming the remaining old dude (or dudette) is maintaining the code at this point. They could just be serving the role of whacking people with sticks to keep them from accidentally pulling on the thread that is keeping the whole system working, and making loud noises if some new boneheaded manager tries to kill the system without realizing that doing so would cripple the entire company.

Being the last person who understands what all the components are supposed to do and why they are important is not the same as knowing how to modify and maintain said components.

With that said... there's still going to be a crisis when this person announces their retirement...

A mainframe system that has survived multiple waves of "this is the new thing" has one of two things going for it:

1. It works fine and isn't broken, and handles way more thoughput than any alternative proposed without having to spend a stupid amount of money.

2. Nobody understands how the code and logic work, and nobody wants to touch it for fear of breaking it, and all the prior migration projects failed miserably, so much so that there's no money or will to try again. So even if they want to change how things work, they can't.

The two are not mutually exclusive, but #2 tends to be more the case for a system that desperately needs an upgrade, which I'm assuming all these unemployment systems are.

With that said... systems tend to reflect processes and culture, so I don't have high hopes for replacing just the mainframe, to fix the problems that they had/are having.

I mean, this should be basic recovery scenario one in any continutity and disaster recovery plan.

And it doesn't have to be AWS that has the oops. It could be an undetected corruption that borks the database during the next major engine/schema upgrade. A disgruntled employee wipes the encryption keys. A ransomware gang infiltrates your accounts, exfiltrates your customer data, and then locks you out.

Having AWS be the god-tier admin layer just creates one additional source of heartache if this happens on a global scale... outside of your immediate control.

Were your outages as an end-user or as a direct customer of AWS?

US-East-1 is normally the default region (and also historically the cheapest region) of all the alternatives.

https://technical.ly/entrepren...

As a result, even if as a consumer you think you'd be better served by a more local region, whatever service you're using probably has a number of critical components predominantly or wholly served out of US-East-1, usually for cost reasons.

There was a comment earlier wondering if the original article was real or made up. Here's a reference to the deployment of the remotely supervised stocking robots in a 2022 Nvidia blog post.

https://blogs.nvidia.com/blog/...

"Tokyo-based startup Telexistence this week announced it will deploy NVIDIA AI-powered robots to restock shelves at hundreds of FamilyMart convenience stores in Japan.

There are 56,000 convenience stores in Japan â" the third-highest density worldwide. Around 16,000 of them are run by FamilyMart. Telexistence aims to save time for these stores by offloading repetitive tasks like refilling shelves of beverages to a robot, allowing retail staff to tackle more complex tasks like interacting with customers..."

"...In the rare cases that the robot misjudges the placement of the beverage or a drink topples over, thereâ(TM)s no need for the retail staff to drop their task to get the robot back up and running. Instead, Telexistence has remote operators on standby, who can quickly address the situation by taking manual control through a VR system that uses NVIDIA GPUs for video streaming."

Maybe the perception of rampant crime is due to reported stories like this one (from Nov 15, 2023)?

https://cbs12.com/news/nation-...

"They got an arrest warrant for Green that contained even more charges than Dâ(TM)Antuono faces: 30 counts of robbery, burglary, firearms violations, and more."

Or this one from Sept 21, 2024:

https://ktla.com/news/local-ne...

"Just hours after about 50 juveniles ransacked a 7-Eleven store in Pico-Robertson, another 7-Eleven was robbed at gunpoint in a different L.A. neighborhood.

According to the Los Angeles Police Department, a call regarding a robbery at the convenience store, located on Oxnard Street in Valley Glen, came in at 2:15 a.m. Saturday. "

I don't know about you, but stuff like this would make me think twice about working the register at a convenience store...

Great real world test of services failing. I think this has less to do with AWS screwing up, and more with people putting all their eggs in US-East-1...

This was a temporary, resolvable issue. What happens if there is an irreversible issue that takes out a chunk of the servers sitting in US-East-1?

If you had warm standby systems synchronized and ready to go in a different region, this shouldn't have had much impact (other than maybe having to scale to handle a transfer of load from clients normally localized to US-East-1)... the fact that it did, is a giant red flag. If I was an insurer underwriting cyber loss policies, I'd be combing over the list of companies that had outages very carefully to see if any of them were my customers...

By this definition at least half of America (if not the world) is insane.

https://old.reddit.com/r/marke...

Also keep in mind that a lot of ads are now being served through sources other than web browsers, even if the backend is sending and receiving data over https. You'd have to implement DNS based filtering of ad traffic, if not outright MITM deep packet inspection to block traffic served via apps.

With that said, having recently fired up a browser that had not yet been configured with adblocking when provisioning a new machine... I don't know how people without adblocking can handle the ads.

People cloning posts, videos, and sites, and claiming it as their own work in order to frontrun/hijack social recognition and monetization.

SEO optimization which started out as hiring english majors to ghostwrite "articles" on topics in order to boost your business site (plumbing, roofing, auto repair) in the rankings, which now replaces the first two pages of search results on anything DIY.

Big corporations monetizing and then later discarding communities, and all of the information and relationships in them, when the new model of walled garden replaced the open internet. What happened to Yahoo Groups: https://en.wikipedia.org/wiki/...

Also, remember all the free work people did for the CD Database before it was hijacked and turned into a closed commercial project? https://www.eff.org/deeplinks/...

Yes, the internet was much better 20 years ago when you shared information online because it was a hobby you wanted to share with people. Not something you did in order to further a career as an "influencer". People largely shared stuff without fear of getting ripped off, because why would someone do extra work of copying your stuff without a benefit? People volunteered their time and effort because everybody was able to benefit.

Now a lot of that stuff happens in much smaller niche communities, or gated behind walled gardens, and the wider internet - the internet that we're able to access using search engines? A lot of it is just noise, and the remaining bits that aren't are really hard to find.

Aren't the smartphones the conduit to these addictive, mind-altering services? You could easily sue Apple and Google for developing the operating systems and UIs/app stores that enable predatory applications and ecosystems to thrive.

However, to my knowledge, with the exception of people on public assistance... you don't just get given a smartphone. Someone with disposable income has to buy it... then give it to the vulnerable youth in question.

This sounds very much like more posturing by politicians and lawyers if not an outright cash grab than an honest move to protect developing brains from manipulation...