> Cheaper & easier - not by much.

I spent a total of $150 for my entire homelab setup two years ago with not a single investment made since then. Used to pay $80 per month for streaming back then, bet it would be >$100 now

I would also assume the same for non-Chinese as well

The nightmare for Anthropic to be caught doing that combined with the temptation of their staff to virtue-signal by blowing the whistle...

I trust them to act in their own interest if nothing else.

Aren't they on the "we, the safe AI, must win at any cost" justifiers?

What's a little contract violation if the fate of humanity is at stake?

And what if they use your data tò generate syntethic data to train on?

That would be just as bad.

I would assume at this point that any SaaS product, LLM or not, where the data doesn't reside on your servers on your premises (or your own colo) is training on the entire corpus of your data, whether they'll admit to it or not.

Not for Enterprise. You can safely assume the trillion dollar companies would ban GPT/Claude from being used in house if that was a concern.

Or think it the other way, what if both A and O actually train with these data, then the enterprises found it(or maybe never), what's the other options? I'm not trusting these LLM companies because a. their model live with human generated data even if they claim to generate data with their own model, it's just nit human data, b. this is business not charity, eventually they live with customers' data, no exception

We're talking about a dozen companies with trillion dollar market caps in the US that would each point their army of lawyers at OpenAI/Anthropic. Neither would survive that litigation. That risk far outweighs any potential gains in training data. Just not worth it.

> That's total bullshit. Global music revenue fell by 50% because of piracy[1].

The article you linked does not provide any source and / or a methodology for how this was calculated and attributed specifically to piracy

It's unfathomably weird of you to frame an article as a source for your claim and then wait for someone to point out that the article does not support your claim before admitting that the source is actually a book that you haven't read.

Since the explanation for the claim is essentially pay-walled behind the book, I will just choose to ignore it

"People got used to free shit and stopped paying as much."

"Your claim is poorly supported."

"Well it came from a book. Check it out if you're interested."

"What? It's not free? Hell no I won't pay!"

(anecdotally) proving their point that pay walling something doesn't mean people will pay for it. Buying the book would have been hypocritical.

This whole thread does have me questioning my understanding of cited sources behind paywalls.

A contrived example I have in my head "My friend has more definitive proof in HIS book that <claim>, it only costs $1 trillion. Go read it yourself." For this particular thread its a bit disingenuous, but in a general case how does one go about understanding / disproving arguments made like this?

You decide whether or not the benefits from you obtaining the information outweigh the cost of obtaining the information. Or you convince Elon Musk to buy the book and leak the EPUB on twitter

> "Well it came from a book. Check it out if you're interested."

> "What? It's not free? Hell no I won't pay!"

I am also sceptical that my burning questions will be answered in the book, thus I choose to not buy it. Also, the Goodreads reviews don't look too good

> Public p2p sharing is pretty much dead in the West.

Orpheus and RED are going extremely strong right now, with very active userbases

That's private p2p, not public.

Ah, my mistake!

With all of the videos circulating online of the Tesla self-driving getting itself into extremely dangerous situations (like increasing speed once a child appears on the road), I'm not really sure that their self-driving tech is there yet

Absolutely. That's why their self driving will not be allowed to roll out in the EU and other places until it works well

> Progress takes time and there can be mistakes is basically the story of our civilization tho

Yeah, no one is doubting it. People are just asking to not be lied to about how much progress has been made

> why exactly is the US responsible for feeding people in random countries around the world?

It was not as in not responsible, since international agreements between countries are de facto non-enforceable. Such programs though did help the US in gaining influence and soft power in those parts of the world

There are lots of funding possibilities with Horizon Europe projects with new proposals being approved all the time

> I called siriusxm to get them to turn off their stupid advertising on my in-car infotainment

I know that ads on radio always were and still are a thing, but on actual screens as well? This just gives another reason not to upgrade from my 2010 car.

YES. it's absurd

The screen on my car displays navigation, climate/AC, & carplay. There is a pop-up dialog box that takes over the display and says either "special offer from sirius xm" or "Warning: extreme weather X miles away". The background of the popup is opaque so it blocks out the ENTIRE screen

I can't access navigation, radio, or climate controls until I dismiss the pop ups!

AND pop ups stack. I have a video in my phone of pressing 'dismiss' 23 times in a row before I am able to get back to navigation.

Imagine driving through a thunderstorm in an unfamiliar city and you have to dismiss 1-3 popups per mile to just see the map!

> Cursor has $4 billion annual revenue rate so $60b is 15 years of future cashflows.

This assumes that Cursor's annual revenue will be the same or higher for over a decade. It's not really like they don't have competitors

Yep, usually what people do is import the template into Overleaf and work on it from there. It's only when journals and conferences will provide typst templates is when we will see wider adoption for writing papers