eminent101

Ah! Sorry. I did miss that they said "you still need to do things like write an impact assessment".

So what's the best course of action? Remove comments feature entirely? Maybe that's what I should do. I wonder what everyone else's doing.

> Does anyone in your blog comments ever discuss circumvention of DRM?

No, they don't. My blog is not all that popular. It has got some programming puzzles, Linux HOW-TOs and stuff. Most of my audience is just my friends.

If I'm moderating all comments before they're published on the website, what's the problem? I mean, I've got a simple tech blog. I'm not going to publish random drive-by comments. Only comments that relate to my blog are ever going to be published. Am I making sense?

This is an honest question. Why does a blog need to shutdown? If they moderate every comment before it is published on the website, what's the problem? I ask because I've got a UK-based blog too. It has got comments feature. Wouldn't enabling moderation for all comments be enough?

> This article seems to be paywalled unfortunately.

I am no fan of Medium paywalled articles but if it helps you, here's the article on archive - https://archive.is/J53CE

I doubt there is any point in debating a complex topic with someone whose only responses are dismissive questions and "5 minutes of Googling"! I appreciate your thoughtful comments in this thread though!

> Trivial googling shows you ...

> This took me 5 minutes to Google.

If this is how you get your information, I doubt what you say can be taken seriously. Not to mention that the reference you quote seems like a random website nobody has ever heard of!

This was a very interesting topic and a serious discussion about this topic would at least include references to bonafide surveys or well established trustworthy sources. To find them takes much more than 5 minutes of Googling. Unfortunately I don't have the time to do that, so I requested that if you know something, you share it here with us.

Clearly you do not have the time to do your research either since all you have to present us is "5 minutes of Googling" that turns up a random source!

I have to say that this thread is very frustrating to read. I see @lazide is engaging with you in good faith and providing high effort, thoughtful answers. There's a lot of statistics and factors involved in a discussion like this. So I won't say @lazide's analysis is correct or flawed. But this is a good topic where a good discussion can be had and @lazide is holding up their end of the bargain.

But every response of yours is dismissive. And this makes this thread frustrating to read. You answer every reply with more questions and a tone of dismissal. If you know so much about this area, why don't you begin sharing some facts and enlighten us? Dismissing your co-commenter and answering their replies with more questions is not educating anyone of anything!

It would help if instead of answering a comment with questions, you share what you know. So how much is the cost of wiring, installation, programming and making greenhouses in the span of a year? How much copper is needed per capita? What do you know? Tell us!

But nobody in this thread thought it (simplifying assumptions) stopped. You seem to be making an assumption that someone thought that and then posting long explanations that nobody asked for. I read the "P.S." of grand-grand-parent comment as good humor. Nothing there implied that they really thought that simplifying assumptions would/should stop.

Imagine a world where every bit of humor is interpreted literally and then refuted pedantically! What kind of a world would that be?

If you take something as ordinary as “behavior tree”, 9+ consecutive words don’t surprise me. Many definitions across literature often look very alike.

> behavior trees, the system will often travel down the root of

How else would you write this? To be honest, I’d have written it exactly this way too. I’m sure many other researchers would write it this way too.

I mean this is such a weak example. If there were 9+ consecutive words matching each other in more meatier parts of the paper, then I would be very suspicious.

But the examples cited are about trivial sentences describing ordinary stuff where two people might describe the same thing in exactly the same way because there are only very few ways to describe them.

I looked at the cited examples but none of them look like “plagiarism”. They look like ordinary sentences and if I had to write them myself, I would come up with similar wordings too!

Like for example take the description of “behavior trees”. There are only so many ways you can describe a behavior tree. If 10 papers out there introduce behavior tree by describing it, some of those descriptions are bound to look very similar.

This is like arguing that my paper on prime numbers is plagiarised because the definition of prime numbers in my paper looks very similar to the definition of prime numbers in another paper.

With the amount of literature on these subjects out there, it is natural that some sentences I write would end up looking similar to some sentences in thousands of literature out there.

> Who are you to decide what style the author should use?

Who am I to decide? Well, who are you to ask me who I am to decide? Should we keep going in circles, or would you prefer we actually discuss the point at hand?

Well, seriously, to be honest, I don't care much about what style the author should use. The author should write in their style on their personal blog as much as they want. I care about it appearing on the HN main page and me having to click it and then being unhappy about my dear HN community upvoting a mundane article of questionable quality to HN main page. Who am I to do so? Well, I am the reader!

If I can't even critique an article that I read here, then what are we all even doing here?

> While the crowd here can maintain a domain

And domains can be lost too. Missed payments, error in administrating the domain, government takedowns. Many failure modes exist for domains too. Nothing in the digital world is permanent! That's why I find it disturbing that so much of our digital identities are tied to our email addresses.

But your parent comment said:

"Google is the Scorpion in the parable of the frog and the scorpion - it cannot help its nature."

The scorpion is not malicious in the scorpion and the frog fable. So what's your point?

Do you or someone else know if Fastmail is any better? If I buy a Fastmail email, what protections exist that would stop them from accidentally or arbitrarily locking me out?

And note, I don't live in the US. I'm wondering about this question from a global perspective.

How does that work in practice? I mean I may keep my data at multiple places. But my government, my hospitals, my utility accounts, they all want my email address to send me OTPs, password reset links and such things that are necessary to prove my digital identity.

How do I spread this risk and make it manageable? I have to give them some email address and I fear losing access to my email. And yes, I can lose my email address even if I have my email on my own domain. There are many failure modes for losing domain names. So how do I manage this risk?

> While you dont own a domain name, you have legals means to get it back.

Can you or someone else share more about this? Do these laws work across countries? Can someone in Bhutan exercise their legal right to get back their .com or .org domain name? Must someone in Bhutan always buy a .bt domain name? I'd like to learn more about how the legal framework works and protects the customer from loss of their domain names?

I'm sorry this happened. This is my digital nightmare scenario!

It is disturbing how much of our digital and physical life (utility accounts, medical insurance, etc. etc.) are tied to email addresses and these email addresses are something we can never ever truly own. If you are locked out of email, you are also locked out of at least half a dozen critical portals that send password resets, OTPs and all kinds of authentication fragments to your email address!

Most email addresses are on somebody else's domain and they can lock you out anytime. Even if you manage to set up your domain name, you are still renting the domain name from someone. One missed payment or you somehow mess up the admin work of your domain name or you lose your domain name for any reason (yes, it happens!), nobody in the world can reach your email address!

How did this happen? Weren't the old days of snailmail better? You could own a house or you could rent a house and get actual physical letters at your home. If you moved houses, you could have the new tenants of the old house forward mail to your new one until everything settled down.

Email addresses seem like good secondary mode of communication but I find it disturbing that all around the world, email addresses have become the primary mode of communcation and sometimes the only mode of communication!

Does anyone else feel extremely uncomfortable that so much of our critical digital and physical lives are tied to email addresses, things that we can never truly own and can be taken away from us anytime?

You may be right. I barely use the "flag" link myself so I could be wrong about its true purpose is.

> OpenAI hyping their latest model as more "thoughtful" and "safer" and whatever other bogus criteria they claim to boost their valuation.

If this is true, it's a valid criticism. I hear you. I think it is downright thoughtless to mislead the layman into thinking LLMs can be "thoughtful" by anthropomorphizing them! To all companies that are doing this, seriously, stop dressing up a statistical token generator as if it's some kind of sentient philosopher.

I'd strongly suggest addressing the points I've made rather than resorting to a lazy dismissal like I "sound like a generated account." It's a non-argument and contributes nothing.

For the record, I'm not affiliated with OpenAI. I use ChatGPT often. It solves my problem many times. Many times it fails miserably. I neither love it nor hate it. It is what it is.

My frustration stems from the sheer laziness of some posts and comments about LLMs here on HN. The ones that irritate me the most are those that expect LLMs to somehow possess magical abilities like being able to inherently distinguish between correct and incorrect outputs. Where does this expectation even come from? Have these people spent no time at all understanding how LLMs actually work? No? And yet they expect magic? It's baffling!

If you have a substantial critique of the points I raised, by all means, share it. If not, please spare me the hand-waving dismissals.

> ALL current LLMs are prone to confabulation. AFAIK, there is not a single model which is able to tell “I don’t know” instead of making things up.

I can't help but feel that people who keep parroting these lines haven't even bothered to take 30 days out of their lives to actually learning the basics of how LLMs work. LLMs are not "prone" to confabulation. They confabulate by design. That's how they work. They predict the next token based on probabilities derived from training data, not some magical ability to discern truth from falsehood.

The critical step people seem to ignore is that after the LLM generates a response, you need a filter to determine whether it's correct or not. But this is easier said than done! It can be done but in a very narrow range of cases, like generating code that passes a specific set of test cases or producing a Lean proof. The tests or proof verifiers act as that filter. But even here, you're far from guaranteed perfection. Test cases don't cover everything and proofs might be valid but still irrelevant to the problem.

Expecting an LLM to just say "I don’t know" fundamentally misunderstands what these models are. They're not epistemic agents. They don't "know" or "not know" anything. They just generate statistically plausible sequences. If that seems like a flaw to you, the problem isn't the LLM, it's your unrealistic expectations.

Maybe it did. But what's newsworthy about it? Is it really news that ChatGPT gets stuff wrong? If every person on this thread posts one article to HN describing how they asked a question to ChatGPT and it got it wrong, does that make this site better or worse?

There is no "downvote" for main stories but there is a "flag" option to flag articles that you think don't belong to the HN main page. Users with enough karmas can see the "flag" option.

This sounds fascinating! I know what a Sierpiński triangle triangle is but I'm having so me trouble seeing the connection from picking functions randomly to the triangle. Is there some graphics or animation somewhere on the web that someone can point me to visualize this better?

> trivial for a human expert is not barely understood by ChatGPT

I try to stay away from these arguments but it's a Saturday, so what the heck! I think this topic is becoming a tired topic now. One party would claim that ChatGPT doesn't "understand". Other party would claim that it's a large statistical model, of course it doesn't "understand". Do you expect your washing machine to "understand" that it's washing clothes? Then there is another party that would claim, rightfully, do we understand what "understanding" is? This has been rinsed and repeated over and over again ad nauseum. Is there anything more to learn in these arguments?

I mean, can we just give "ChatGPT got some things wrong, haha" some rest now?

Ok, so ChatGPT got some things wrong. If everyone began posting how ChatGPT got something wrong, we will be here all day. I can't see how this article is newsworthy.

And the article has constant literary interruptions like this:

"I set down my teacup and put my face in my hands."

"There was an almost infinitesimal pause before ChatGPT chirped"

"I closed my eyes and contemplated my life choices."

"I narrowed my eyes as ChatGPT brightly informed me"

I don't know what the author is trying here but there is a time and place to use literary pauses like these and I don't think, this article is that place. The article reads like a weird cross between a rant, technical article and a wannabe literary piece.

So many bold claims in this comment and little to no justification.

For what it's worth I've seen pretty much the opposite. I don't know about competent vs. incompetent engineers. But when it comes to experience, I've seen the inexperienced ones giving super low estimates and the experienced people giving larger estimates.