When I first used my own telescope to view Saturn I had a thought, "they can't fake that!" Photos in a magazine or on television could have been faked. And the moon landing truthers became unhinged about it. I had not appreciated that I subtly and unconsciously held a reservation about the truth of it all.

This is the modern epistemic crisis. And wait till Elon implants a brain computer interface in you. You won't even fully trust your eye looking through a telescope.

Such posts are identifiable and rare, disproving Dead Internet Theory (for now).

Every time I see people talk about 'dead internet' I think about what web sites looked like before Google dropped PageRank into the whole ecosystem and utterly killed the utility of putting huge blobs of black-on-black text at the bottom of a page for keyword-match purposes. This was then shortly followed by Google becoming so popular they could use user behavior itself as a signal for site quality (much to the chagrin of every new publisher ever who suddenly had a catch-22 of "Nobody visits my site because Google won't uprank it because nobody visits my site").

The Internet has never been dead. Or alive. Ever since it escaped its comfortable cage in the university / military / small-clique-of-corporations ecosystem and became a thing "anyone" can see and publish on, there has forever been a push-pull between "People wanting to use this to solve their problems" and "People wanting eyeballs on their content, no matter the reason." We're just in an interesting local minimum where the ability to auto-generate human-shaped content has momentarily overtaken the tools search engines (and people with their own brains) use to filter useful from useless, and nobody has yet come up with the PageRank-equivalent nuclear weapon to swing the equation back again.

I'm giving it time, and until it happens I'm using a smaller list of curated sites I mostly trust to get me answers or engage with people I know IRL (as well as Mastodon, which mostly escapes these effects by being bad at transiting novelty from server to server), because thanks to the domain name ownership model pedigree of site ownership still mostly matters.

> LLMs are just probabilistic next-token generators

How sick and tired I am of this take. Okay, people are just bags of bones plus slightly electrified boxes with fat and liquid.

Advertising is a cancer. Adtech is the delivery mechanism.

Paying creators is the dumbest and most consequential aspect of modern media. There is no reason to reward creators, zero. They should actually be paying Youtube for access to their audience. They actually would pay to be seen, paying them is both stupid and unnecessary. Kill the incentives and you kill the cancer.

I've used em dashes for decades — They're great.

You are absolutely right — Let me know if you want to read my personal anecdote on "Dead Internet Theory"...

Yeah, I especially hate how paranoid everyone is (but rightly so). I am constantly suspicious of others' perfectly original work being AI, and others are constantly suspicious of my work being AI.

What's wrong with using AI to write code?

What secret is hidden in the phrase “you are absolutely right”? Using Google's web browser translation yields the mixed Hindi and Korean sentence: “당신 말이 बिल्कुल 맞아요.”

> I don’t use social networks (and no, Hacker News is not a social network)

If you define social networks as a graph of connections, fair enough - there's no graph. It is social media though.

HN is Social in the sense that it relies on (mostly) humans considering what other humans would find interesting and posting/commenting for for the social value (and karma) that generates. Text and links are obviously media.

There seems to be an insinuation that HN isn't in the same category as other aggregators and algorithmic feeds. It's not always easy to detect but the bots are definitely among us. HN isn't immune to slop, its just fairly good at filtering the obvious stuff.