Well, his writing style is too good. The sentences flow too beautifully, he uses rich vocabulary and styling. It's unusual to see that style of writing online. I definitely don't poses that power.
I don't know the author of this article and so I don't know whether I should feel good or bad about this. LLMs produce better writing than most people can and so when someone writes this eloquently, then most people will assume that it's being produced by LLM. The ride in the closed horse carriage was so comfortable it felt like being in a car and so people assumed it was a car. Is that good? Is that bad?
Also note that LLMs are now much more than just "one ML model to predict the next character" - LLMs are now large systems with many iterations, many calls to other systems, databases, etc.
> LLMs produce better writing than most people can and so when someone writes this eloquently, then most people will assume that it's being produced by LLM.
I really don’t think that is what most normal people assume… And while LLMs can definitely produce more grammatically accurate prose with probably a wider vocabulary than the average person, that doesn’t necessarily mean it’s good writing…
I meant "good" in the formatting, grammar, vocabulary sense. I'm not arguing that LLMs are "good" in writing amazing prose.
I mean look at two of us - I have typos, I use half broken english, I'm not good in doing noun articles, my vocabulary is limited, I don't connect sentences well, you end sentences with "..." and then you start sentence with "And", etc. I very much believe you are a real person.
This once again brings up the point that while Meta and other "single company" social networks can easily exclude you, you can't get excluded on Nostr.
It's designed in a way that that's not even a thing. Anyone can create account locally on their computer or mobile phone (even completely offline) and that's it.
If you save & store your "notes" or "posts", you can always re-broadcast them later to different "relay" servers - and this is what your app can do for you anyway.
For larger amounts it makes sense to use the bitcoin rails for international transfers. I'm doing bank to bank international transfers and using bitcoin saves around 3% compared to Wise and you get the money immediately (or within 1hr, depending on what you use).
Few years ago I needed to transfer a big sum from a Scandinavian country into Euro. The official bank exchange rate plus fees was worse than Wise’s. But I asked the bank and the bank gave me an exchange rate that was like 0.1% better than one from wise.
Depending on the direction, but there are ways to actually make a little extra on top of the middle exchange rate (e.g. on the USD to EUR path), since there are many people that want to buy no-KYC bitcoin in Europe and they are willing to pay a couple % extra.
When you start transacting on Bitcoin Lightning network (which is essentially sending pre-signed bitcoin transactions in a smart way, without submitting them on the main chain), then you no longer see each transaction. Lightning introduces decent privacy, not perfect, but decent.
It does guarantee double spend protection, that's exactly what the lightning nodes do. If someone tries to double spend, they lose their bitcoin. I have a different definition for "using bitcoin" than you do.
Similarly I believe that HTTP Live Streaming is using Internet...
That's only true if they observe the channel being fraudulently closed in the right time window. You have to actively monitor the BTC chain to see if your Lightning network partner might steal from you. If you don't (e.g. There is some network or power issue and they take this opportunity to steal), tough luck.
Basically Lightning is like a tab that you open in a bar. You perform various transactions on the tab, and only settle later. No one would say that you're using Visa when you tell the bartender to put some drink on your tab, even though at the end of the day or week or whatever, the transaction will go through Visa.
Bar tab is so very bad analogy here that breaks for multiple reasons:
- At the bar is almost always a single direction: customer pays the bar. Lightning is both directions - sending money between friends and sometimes to shops.
- In this (bad) analogy bitcoin would not be Visa, but bitcoin would be "dollars". Both the bartender and customer would say that they are using dollars for the tab.
- The tab analogy doesn't really match the fact that you establish a "channel" (e.g. both you and the bartender would put some bitcoin into this channel and then you can pay your grandmother in another country with bitcoin in this channel... see the analogy doesn't work here), you can resize ("splice") the channel if you want, you can swap in and swap out...
In most solutions your wallet monitors the chain and it automatically resolves any of the dispute (e.g. Phoenix wallet, or Zeus). The time scale is also different, so even if power goes out for multiple days and you are running your own very private wallet without any associated service (LSP), then you still have on the order of weeks for your wallet to automatically resolve any issue.
I know I'll get hated for this on Hacker News, but this has been solved quite well on the bitcoin & Nostr side of things. It's easy to tip couple cents or whatever amounts and there are many apps / websites that support that.
The main difference is that using the legacy dollar rails is super annoying for small amounts, since there are multiple banks/companies on the path between you and the author you are trying to tip. And each of these intermediators needs their $$$ from you.
If you have a preference for this style of definition, then we could say that
Nostr is a protocol that's well suited for creating decentralized applications that need publicly verifiable identity, censorship resistance and event based communication.
For example, https://zapstore.dev/app is an Android AppStore that uses nostr to provide a decentralized way to verify the developers and remove "fake" apps.
I don't know the author of this article and so I don't know whether I should feel good or bad about this. LLMs produce better writing than most people can and so when someone writes this eloquently, then most people will assume that it's being produced by LLM. The ride in the closed horse carriage was so comfortable it felt like being in a car and so people assumed it was a car. Is that good? Is that bad?
Also note that LLMs are now much more than just "one ML model to predict the next character" - LLMs are now large systems with many iterations, many calls to other systems, databases, etc.
reply