Safford Unified School District #1 v. Redding, 557 U.S. 364 (2009) (citing Thomas v. Roberts, 323 F.3d 950 (11th Cir. 2003)("This case involves a[n] ... action brought by thirteen elementary school students ... against Tracey Morgan, their teacher [and others].... [W]e affirm[] the district court's grant of qualified immunity to the individual defendants on the children's claims.")).
I feel like almost everyone using AI for support systems is utterly failing at the same incredibly obvious place.
The first job of any support system—both in terms of importance and chronologically—is triage. This is not a research issue and it's not an interaction issue. It's at root a classification problem and should be trained and implemented as such.
There are three broad categories of interaction: cranks, grandmas, and wtfs.
Cranks are the people opening a support chat to tell you they have vital missing information about the Kennedy Assassintion or they want your help suing the government for their exposure to Agent Orange when they were stationed at Minot. "Unfortunately I can't help with that. We are a website that sells wholesale frozen lemonade. Good luck!"
Grandma questions are the people who can't navigate your website. (This isn't meant to be derogatory, just vivid; I have grandma questions often enough myself.) They need to be pointed toward some resource: a help page, a kb article, a settings page, whatever. These are good tasks for a human or LLM agent with a script or guideline and excellent knowledge/training on the support knowledge base.
WTFs are everything else. Every weird undocumented behavior, every emergent circumstance, every invalid state, etc. These are your best customers and they should be escalated to a real human, preferably a smart one, as soon as realistically possible. They're your best customers because (a) they are investing time into fixing something that actually went wrong; (b) they will walk you through it in greater detail than a bug report, live, and help you figure it out; and (c) they are invested, which means you have an opportunity for real loyalty and word-of-mouth gains.
What most AI systems (whether LLMs or scripts) do wrong is that they treat WTFs like they're grandmas. They're spending significant money on building these systems just to destroy the value they get from the most intelligent and passionate people in their customer base doing in-depth production QC/QA.
This rings true. However I have used one AI automated support chat that didn't behave that way. I wish I could remember the vendor but I do remember being blown away when it said something like "that sounds like a real problem would you like me to open a support ticket for this?". Which it then did and subsequently a human addressed my issue.
The word "interstate" does not exist in the text of the Constitution.
There's arguably some merit to your position, but the argument that some case law is invalid because it doesn't meet the definition of a term defined in other case law is circular and incoherent.
Completely agree. This is really unique. Can you imagine if it were standard practice to be open to supply chain attacks like that, by blindly relying on hotlinked or unpinned dependencies?
Why imagine? Let's take a quick look at what's actually happening right now. We can check some widely used libraries and see what their instructions are teaching new developers.
Pay close attention, they are inviting the new developer to link not just to Bootstrap, but to Popper!
HTMX (code snippet from their quick start guide):
```
<script src="https://cdn.jsdelivr.net/npm/[email protected]/dist/htmx.min.js"></script>
<!-- have a button POST a click via AJAX -->
<button hx-post="/clicked" hx-swap="outerHTML">
Click Me
</button>
```
Fontawesome: A video quick start guide and instructions that recommends using the direct link to the kits via CDN for performance!
Look, I certainly don't think they should be used this way. But, to say that it's unique to the White House app? I definitely wouldn't say that. In fact, I think you've dangerously overestimated the status quo.
I was being sarcastic. Although hot linking is not particularly common, it's common enough; and unpinned dependencies are just as much if not more of a supply chain attack risk.
I'd bet something like 70+% of all JS apps are inadequately protected against the risk of a malicious actor gaining access to a dependency's repo.
Pearlclutching over this while ignoring the lessons of `left-pad` and `colors` is biased motivated reasoning at best.
I'm not sure I follow. How does an integrity check help when the source is compromised? The developer doesn't know that their repo is compromised. They continue posting legitimate hashes because the repo is legitimately compromised.
> This is particularly problematic given the ways that it could be abused by some of the more authoritarian governments in the EU.
> Yes, I'm thinking of Viktor Orbán of Hungary.
Lol what?
The UK leads [edit: in Europe overall, obviously not the EU] with approximately 18 per 100k prosecuted for online speech. Germany is at about 4 per 100k. Poland at about 0.8 per 100k. Hungary about 0.1 per 100K.
For any definition of authoritarian that relates to chat control, the UK is two base-10 orders of magnitude more authoritarian than Hungary (7 base-2 orders of magnitude).
This figure in the UK is unsourced and I'm fairly sure is not true (or at least not what you've labelled it), and has been quoted out of context by people trying to stir trouble not reasoned debate. I'll assume good faith here and say the start of the video lays out why the figure is not what you've labelled it to be
The issue isn't how much free speech online is being punished. It is how surveillance could be used to reinforce authoritarianism.
The UK does a lot of prosecuting people for having said nasty things online that someone else didn't like.
Hungary is far more inclined to surveil political opponents, put people in their network in jail without fair trial, surveil successful businesses whose bribes were insufficient, find excuses to punish those businesses.
Not only are there not similar reports about the UK, but its better position in international corruption rankings points to a culture that would be less likely to tolerate this.
Any further questions about why there should be concerns about how Hungary would be likely to abuse a law like this?
Germany and Poland are. Does the existence of a non-EU country in a data set about European countries detract from the fact that Hungary doesn't prosecute people for online speech to the same extent as other European (incl. EU) countries?
I'm quite sure they thought about the UK as well, given the practice of prosecuting for lawful speech, jailing or arresting for planning peaceful protests (or threatening to arrest a man with an EMPTY placard), jailing for opposing the genocide or voicing support for the unlawfully proscribed organisation.
reply