> Gemini and Chat GPT with search are both perfectly capable of producing decent essays with accurate citations
OK, but this quote from your essay:
> The National Cancer Institute (NCI) reports that "cigarette smoking and exposure to tobacco smoke cause about 480,000 premature deaths each year in the United States."
...that citation is wrong. It's not from the NCI at all, the NCI cited that figure which came from another paper by the U.S. Department of Health and Human Services.
The essay doesn't have accurate citations, the model has regurgitated content and doesn't understand when that content is from a primary source or when it in turn has come from a different citation.
> It does seem to very much just be at the consumer level. We've still got servers chugging away in datacenters that are pushing 15 years old and still have several years of life left in them.
The datacenter is temperature controlled and the air is filtered. The server is locked in the rack and never moves. By comparison, your laptop or your phone go through hell on a daily basis.
> It looks like Visa debit fees are about $0.21 + 0.05%. Compare that to cash handling...
Debit fees are that low because of the Durbin Amendment, which legislated caps on debit card fees (amongst other things). Credit card fees are where the real money is made, and the meat of the complaint here.
>>> Credit card fees are where the real money is made, and the meat of the complaint here.
From the DOJ Press Release [0]:
"Justice Department Sues Visa for Monopolizing Debit Markets"
"Filed in the U.S. District Court for the Southern District of New York, the complaint alleges that Visa illegally maintains a monopoly over debit network markets by using its dominance to thwart the growth of its existing competitors and prevent others from developing new and innovative alternatives."
0 mentions of the word "credit"
27 mentions of the word "debit"
I'd love for this lawsuit to be about Credit Card fees, but it appears be limited in scope to debit card fees.
> when they do go into the office, they’re just having Zoom calls in the office instead. That’s what Amazon is trying to fix here.
Not really. At a company the size of Amazon teams are often in other buildings, if not other cities or countries entirely. The zoom calls continue unabated.
I'm not sure about Amazon, but in some of these large companies the team is still distributed even when they are "in-office". So everybody is still on zoom calls.
> Is "musicians who don't even know the genre they'll use professionally yet" a valid market in the first place?
That's not really Teenage Engineering's primary market, in the same way Rolex's primary market isn't "people who need to tell the time". Both T.E and Rolex products do their jobs really well, but the people buying them are buying more for the aesthetic than the function.
Teenage Engineering are primarily a design boutique, although musicians do use their products their main audience are collectors / audiophiles / graphic designers going through a mid-life crisis.
There are two big sides to the iPad market, the "spend more than >$1000 for a designer/pro tool" side and the "it's just a good <$500 tablet" side. The latter probably gets 5x-10x the amount of use per purchase, especially by younger audiences.
Because FBI CJIS requirements, adopted by state law enforcement bodies, require it. I support a Public Safety Answering Point (PSAP, aka a 911 call center) and I push back on as many of the inane requirements as I can with compensating controls.
Example: As of right now I am still required to expire passwords every 90 days. My state is considering the current guidance from NIST but FBI CJIS policy still mandates the expirations.
I don't know what CJIS requirements entail precisely, but at a first glance, they seem reasonable. But it's weird that people then think they can comply by installing a product with a disclaimer against their intended use. It's just a token acknowledgment: "Yeah, we've read it, but we don't really care."
If that's also the interpretation of the courts, then each company would be invidivually liable, at least towards the government.
Holy shit I cannot stand the password expiration requirements. Like you said, NIST literally recommends against it but so many regulations require it. So aggravating.
Because no endpoint protection software exists that doesn’t have the same disclaimer clause. So you install this one and accept the lack of vendor liability.
(If such a thing did exist, it would cost a lot more!)
The data entry endpoints in a 911 dispatch center should not be running a general purpose consumer OS. They should be single purpose machines much closer to a dumb VT100 terminal than a personal computer. Maybe something like a stripped down hardened Chromebook. No internet connection. No personal email, web, or other use allowed or even possible. A product like crowdstrike should not be needed because it should not be possible to run anything but the dispatching software on those machines.
That's what computer aided dispatch (CAD, in the industry) software was 30 years ago (my PSAP had an AS/400). The market has rejected it. Also, see my other comment re: FBI CJIS policy.
In the PSAP I support we have three dedicated PCs at each workstation to run the CAD, phones, and radio. Each of those has a dedicated VLAN, separate physical servers and storage, separate Active Directory forest for CAD (no AD for radios or phones-- standalone PCs), and default-deny ACLs for inbound and outbound traffic on the hosts and at the borders.
A fourth dedicated PC (VLAN, ACLs, physical servers, AD environment) does email, web browsing, etc. (All of it is shackled together with a nice KVM that supports a single keyboard and mouse controlling up to 5 PCs.)
Not every PSAP does this and I think that's insane. The law and fire agencies we interface with absolutely do put a single PC on a desk (or in a cruiser) and use it for everything (and we filter and monitor the traffic coming in from them over our VPN heavily and block access at the first sign of anomalous traffic). Often their budgets don't support the notion of using dedicated computers for task-oriented work. The marketers have pushed general purpose devices for this kind of application.
In the last 5 years all three "hardened" systems we use (all companies acquired by Motorola) have started requiring Internet access for various APIs they use, and for integration with third-party vendors (mapping, public information databases, and task instructions for telecommunications). I think it's ridiculous, but I don't get to decide the direction of the product roadmaps or what the business stakeholders want from a feature perspective.
Motorola (who makes the CAD software used by some of the largest US municipalities) is pushing for hosted CAD and integrating hosted features into on-prem systems. (Of course, they have a managed security product offering that they want to sell along side it.)
> I was really hoping Portal was going to catch on, I really liked it and found it to be a great home VC platform.
I worked on Portal and it came down to:
1. Cambridge Analytica. It is difficult to persuade people - and the media - you can be trusted to put a camera in their homes while you’re fighting a massive privacy scandal.
2. A leadership coup resulted in the project moving under Reality Labs, which never really wanted to be in the smart device business in the first place.
There were a bunch of other issues, but these two basically doomed the project from the start, which is a shame as it was well received by those who actually bought it.
OK, but this quote from your essay:
> The National Cancer Institute (NCI) reports that "cigarette smoking and exposure to tobacco smoke cause about 480,000 premature deaths each year in the United States."
...that citation is wrong. It's not from the NCI at all, the NCI cited that figure which came from another paper by the U.S. Department of Health and Human Services.
The essay doesn't have accurate citations, the model has regurgitated content and doesn't understand when that content is from a primary source or when it in turn has come from a different citation.